Every web-browsing human being reads a product description, almost every day. Most product descriptions are eye-bleeding horrors of lousy copy and unclear information. That means some of the most-read digital content is some of the worst.
We can do better.
This is my sort-of-system for better product descriptions. Use as you see fit.
Two Kinds Of Descriptions
You’re writing two product descriptions:
The SERP snippet, to improve rankings and generate clicks
The product description page, to generate sales
Both impact rankings. But writing just for rankings will kill sales, and vice-versa. You have to find the right balance.
I beg of you, please don’t go and rewrite 10,000 product descriptions to the exact formula I outline here. It’s a starting point. Be creative.
The Product Description Page
This is the classic “product description.” Folks read it when they’re making their buying decision. They’re looking at two things that your writing can impact:
If UX is solid and the product is good, a great description will explain features and establish value so well that the customer clicks buy.
I focus on three elements of the product description page:
The on-page title
The Product Description Page
The Product Description Page
If you don’t have bullets, may I suggest adding them?
There’s lots of other stuff: Images, call to action, price, for example. I’m not writing about those here. I know my limitations.
Check For Duplication
If you’re rewriting an existing description, check for duplication.
Copy two sentences from the blurb
Put them in quotes
Paste them into Google
Do these sentences appear on other sites? That might be OK. But for SEO, duplicate content is a problem. Also, ask yourself: If your product description matches one or more other sites, what reason do folks have to buy from you, instead of them? If you can’t answer, you need to rewrite.
Gahhh! Duplicates. This is a problem.
Gahhh! Duplicates. This is a problem.
If you need to rewrite the description, don’t worry. Follow the rest of these recommendations, and it’ll happen naturally.
The On-Page Title
Note: Writing titles for Amazon is an entirely different discipline. Start with these recommendations, but you’ll need to include more product detail. It’s annoying, I know.
Your on-page product title starts as the product name:
Fast Roller TX 1000
But it must pass the Blank Sheet of Paper Test: The title, written on a blank sheet of paper, should make sense to a knowledgeable stranger. The Fast Roller is a road bicycle tire. Try this:
Fast Roller TX 1000 Road Bicycle Tire
That sounds like an SEO wrote it. When you’re wearing your SEO hat, though, you don’t write copy. You optimize it. Never optimize while you write.
I want something tighter. Remember, the blank sheet of paper test says a knowledgeable stranger. So this will work:
Fast Roller TX 1000 Road Tire
Maybe there are two TX 1000s, though: One for each valve type. Then I end up with:
Fast Roller TX 1000 Road Tire: Presta; and
Fast Roller TX 1000 Road Tire: Schrader
You want higher rankings, though, so you’re tempted to write a fifty-word title. Use your judgment. An overweight title won’t pass the blank sheet of paper test:
Fast Roller TX 1000 Road Bicycle Tire Flat Proof Presta 700C 150TPI Bike Rolling Thingie With A Valve And Tube And Stuff
The knowledgeable stranger will give up. Think before you start keyword stuffing.
The Blurb: Write An Appeal
Fill the blanks:
If [thing or need] then this is a perfect [product].
"If you [want puncture-resistance] then this is a perfect [road bike tire].”
That’s your appeal. It’s not the only way, but it’s a robust introduction.
You can combine multiple appeals:
“If you want puncture resistance and great handling, then this road bike tire is perfect.”
“The TX 1000 provides puncture resistance without sacrificing weight, for a tire that delivers great handling and low rolling resistance.”
Again, I plead. I beg. I implore. Don’t use this as a formula.
The Blurb: Point Out Results
Something about this product makes it uniquely valuable. I hope.
Tell me how you outperform:
“In testing, the TX 1000 showed greater flat-resistance than all major competitors.”
Describe unique features. Get specific!!!
“The TX 1000 is the only tire with an unobtainium valve for greater durability and easier inflation.”
I won’t call this the USP because the term’s so overused it makes me ill.
The Blurb: Find The Unnoticed Obvious
Find one important unnoticed feature related to the appeal. For example threads per inch (TPI) affect a road cycling tire’s puncture-resistance and handling. If no competitors talk about TPI, we should:
“150 TPI means a supple, flat-resistant sidewall.”
Now, I have:
“If you want to balance puncture resistance and performance, then this is a perfect road bike tire. 150 TPI means a really supple, flat-resistant sidewall that doesn’t sacrifice handling or increase rolling resistance. The TX 1000 is also the only tire with an unobtanium valve for greater durability and easier inflation.”
The Blurb: Remove Words That Should Never Be Spoken
Plague words. Ew. Additionally, really, indeed, obviously. Shudder. Dump them all.
I’ve got a whole list of plague words right here. If you use ’em, delete ’em.
“If you want puncture resistance, then this is a perfect road bike tire. 150 TPI means a really supple, flat-resistant sidewall that doesn’t sacrifice handling or increase rolling resistance. The TX 1000 is also the only tire with an unobtanium valve for greater durability and easier inflation.”
I also dislike overuse of unrivaled, unmatched, best, fantastic, or any other phrase that doesn’t apply to your product or your category of product. If you’re Rolex, maybe you can say unrivaled. If you sell shoelaces, stop it.
The Blurb: Remove The Breathless
Avoid the painfully obvious. Without bicycle tires, I get sparks and hemorrhoids. And only an idiot wants a tire that combines high rolling resistance with vulnerability to sharp objects. I get it.
Don’t tell me I’ll love this product, either. You’re already implying that. Saying it out loud seems needy and pushes me away.
“In cycling, tires are important. Performance and flat resistance matter. If you want to balance puncture resistance and performance, then this is a perfect road bike tire. 150 TPI means a really supple, flat-resistant sidewall that doesn’t sacrifice handling or increase rolling resistance. The TX 1000 is also the only tire with an unobtanium valve for greater durability and easier inflation. You’ll love this tire!”
I know I need a bicycle tire. Tell me why I need this bicycle tire.
The Blurb, Resplendent
Here’s what we’ve got:
“If you want to balance puncture resistance and performance, then this is a perfect road bike tire. 150 TPI means a supple, flat-resistant sidewall that doesn’t sacrifice handling or increase rolling resistance. The TX 1000 is also the only tire with an unobtanium valve for greater durability and easier inflation.”
On to the bullets.
The Bullets: Find The Questions (And Answer Them)
Bullets are punchy little bits of information. Readers scan for them. Use them to dispel concerns and answer questions. Finding good bullet content is easy:
Go to Amazon.com. Search for your product, or a relevant one. Scroll down to “questions.”
Answers to questions make great bullets
Answers to questions make great bullets
If there are any, find the five most-read and most-asked. Write a brief response to each one. Keep those answers handy.
Do the same on other sites: Walmart, Jet, and vertical-specific sellers all have “questions” sections.
Those will become bullets in your product description. You might even repeat items from the blurb, like “150 TPI.” Use your judgment.
I found many questions about tire weight, tube versus tubeless, and sidewall color. So my bullets could be:
10 grams (a guy can wish)
Requires a tube
You can skip bullets if you want. Maybe you don’t need them. That’s fine. I mean, who needs to answer all those pesky customer questions, right? That’s my over annoying parental way of saying you need bullets. Bribe the developer. Get the branding team drunk and ask for written approval. Whatever you have to do.
Destroy the FAQ
The FAQ is where copywriters go to die. If there are frequently asked questions about a product, write brief answers for those, too. Add them to your bullets, or the product description. Or add a separate section for related FAQ.
If you’re reselling someone else’s product, look at the manufacturer’s FAQ.
Those can become more bullets or part of the blurb.
Everyone wants to know if we guarantee this tire. We provided this answer, so I turned it into a bullet:
10 grams (a guy can wish)
Requires a tube
Guaranteed rim fit and flat resistance against normal debris. No rampaging hippos
Product specifications? Your call. I don’t think you need specs for a capybara plushie. You might if you sell roofing shingles.
Product ingredients? I like to put them after the bullets, but it depends on the product. A bag of popsicle sticks doesn’t need a list of ingredients. I hope.
Just Get To The SEO, Ian
If you did all of the above, you’ve got an optimized product description page. If you want to take it further:
Make sure you don’t avoid your keyword. You sell bicycle tires, not inflatable wheel support.
Start with the important stuff. Ingredients rarely belong in the first paragraph. Your appeal does.
If you’re not sure what a meta description tag is, read this.
The meta description has zero direct impact on rankings. It does, however, impact clickthru. These are a few things I try to do:
First: Include the features for which you believe people will search. Those get bolded. Their presence will reinforce that this is the right product. If someone searches for “rolling resistance,” “presta,” and “schrader,” and I have that word in my meta description, the search snippet will look like this:
A SERP snippet with bolded words
A SERP snippet with bolded words
The searcher is more likely to click.
Second: Use the highest-performing ad text. A few years ago, Wil Reynolds made this recommendation. It blew my mind: Use the highest-performing PPC ad text as your description tag. You’ve already tested that text. You know it gets high clickthru from a SERP. Blew. My. Mind.
I don’t recommend doing this for the on-page, visible product description. Ad text is optimized for search results, not a product page.
Finally: If it’s relevant, include differentiators: Shipping time, available colors/sizes, genuine original, etc. Anything that matters to your audience. On the other hand, don’t tell me you have genuine original socks. I care exactly not at all.
Try to use all available characters. As I write this, the accepted maximum is 300 characters. It changes all the time. Do your research.
Done! Or Not.
You’ve written a great product description page. You’ve got a great SERP snippet. Nice!
Keep an eye on page performance. Look at clickthru rates. Revise. Keep trying to..
Recently, the team from Portent attended the latest installment of Seattle Interactive Conference, a fantastic digital conference just down the block from our shiny new home office. Thousands of attendees flocked to downtown Seattle and listened to dozens of speakers share their insights about the digital marketing landscape. (Portent’s own Tim Mehta spoke at SIC about friction in UX and how it affects a user’s journey.) As with any conference, some talks were riveting and others just missed the mark. I attended 11 talks during the two-day event. Here are three content strategy insights I took away from those SIC discussions.
Emphasize Data Transparency (Multiple Speakers)
In a GDPR world, we are the product for many global companies, like Facebook, Twitter and Google. Old news to digital marketers, but still very much worth repeating. Knowing what data these companies collect about us and how it’s used is of monumental importance. Look at the Cambridge Analytica scandal as an example of how data gets abused when users are kept in the dark or mislead.
Nathan Kinch, the author of “Designing for Trust: The Data Transparency Playbook,” says data transparency is crucial because user trust is at an all-time low. His solution is for brands to tell users exactly how they intend to access and process their data. This includes discussing who else (external companies, partners, etc.) has access to the data and what benefits users gain from this data exchange.
Among companies that do provide this transparency, many of them don’t present the information in a user-friendly fashion.
Ask yourself this question: Is any “average user” going to scroll through 15 pages of legalese that they can barely understand, let alone use in an actionable way? I highly doubt it.
Data transparency must be easy-to-digest information that involves users from the get-go. Not just checkboxes and an “I Agree” button.
So what does that interactive process look like?
If users create a new account with your brand, hold their hand and guide them through visual and interactive steps that discuss exactly what happens to their data. If at all possible, give users the option to opt out of data collection entirely. If you do give users this option, provide a caveat that teaches them what services and benefits they’re going to forfeit for withholding the data exchange. An important reminder for this step, teach your users in plain language what they’re giving up but don’t guilt-trip or shame them into giving data access.
This interactive process builds trust and allows users to create a positive connection with the brand. The opt-in, opt-out method allows users to become partners instead of just a data piggy-bank.
Unlike a bad navigation or slow website, data collection is not an in-your-face UX concern for many users. But unlike aesthetic UX snafus, if a brand misuses, loses or abuses data, users lose trust in the brand as a whole. Ultimately, user experience is all about trust, communication and ease-of-use. And data transparency is among the most important pieces of that puzzle.
Congratulate Your Users (Multiple Speakers)
Receiving a high-five feels good, right? It’s a celebration that you did a great job or achieved some task worthy of congratulations. But when is the last time a website or app gave you a high-five for visiting, making a purchase or being a returning customer?
Unfortunately, you probably don’t get to experience high-five moments online very often.
A high-five moment in UX, also known as a “delighter,” occurs whenever you provide users with a fun, congratulatory experience for completing a given task.
Ultimately, high-five moments provide users with a more human experience and allows them to feel more connected to a brand, product or website.
If high-five moments are so beneficial for a user’s psyche, why don’t more websites use the tactic?
The simple answer is that properly executing a high-five moment is really damn hard. And there is no “look at their elbow” cheat in UX.
As UX Planet writes, the MailChimp high-five GIF works as a delighter because the underlying MailChimp product does a great job of fulfilling user needs. MailChimp has fine-tuned its service so it’s useful and easy to use.
The totality of product functionality and the service’s UX is what makes-or-breaks a delighter.
“In order for your delighter to have a positive effect, you must first meet or exceed the user’s basic expectations,” the UX Planet article states. “Otherwise that moment will likely add a layer of cheese on top of the original disappointment.”
There are countless examples of cheesy delighters, but one high-five moment gone horribly wrong stands atop the garbage heap: Clippy, your “favorite” helper from Microsoft Word. Clippy is obtrusive and annoying when it congratulates you for writing in a Word document. Instead of providing a helpful service or making you feel accomplished, Clippy creates aggravation and diminishes Microsoft Word as a whole.
Too often, designers and UX’ers alike focus on ensuring designs are aesthetic and practical. We forget that users love delightful experiences, too (just don’t go overboard and create something like Clippy). So the next time you’re designing or evaluating an experience, ask yourself three questions:
Does this product/experience/website/etc. meet my users’ basic expectations?
Do I think the product/etc. can be innovated or improved enough to provide a performance payoff?
What change to the product/etc. can make the intended innovation more delightful for users? The answer becomes your “delighter.”
If numbers 1 and 2 are a “yes,” then consider how to incorporate number 3 into your design or recommendations. After all, who doesn’t enjoy a well-earned high-five?
Use Authentic, Emotional Stories to Inspire Action – Paul Norris
Storytelling is humankind’s natural form of communication. Stories teach us morality, entertain us, and even help our societies evolve. They’re also crucial components in any marketing strategy because an excellent, authentic story compels users to satisfy whatever emotions or desires the story inspires.
So how to do you create a compelling story? The best place to start to place a modern twist on Aristotle’s classic storytelling framework.
At SIC, Paul Norris spoke about emotional storytelling in UX. Below is his modern twist on Aristotle’s formula.
Vision: Present your users with a scenario. Help them imagine how life is in the world.
Problem: Teach your users about what’s wrong in the world you’ve presented.
Inspiration: Show users how the problem has changed the world in this story, and give them a glimpse of hope for how these villainous circumstances can be repaired.
Involvement: Present users with the solution you hinted about earlier.
Fascination: Allude to (or show) how the world has changed or can change now that users are involved and adopting the solution.
As you can see, within 30 seconds the video tells a compelling story that successfully follows this formula.
Importantly, viewers are left with one defining message: travel keeps the world moving forward and you should participate. With a posteriori knowledge of Airbnb, the argument that Airbnb helps people travel is unstated but readily apparent, too.
This type of storytelling is powerful because it empowers users to see the story as aspirational and shows how they can have a positive effect in the world by getting involved and traveling.
This formula works because it triggers three distinct responses in users:
Inherent Response: an immediate feeling that influences your instinctual perceptions about a scenario.
Reflective Response: how you believe this scenario affects your life or the lives of people you care about.
Behavioral Response: the actions you take based on the outcome of your reflective response.
Tapping into these three responses helps guide users toward a specific behavior. It’s a stellar method to use while developing your user journey map.
If you missed SIC this year but you’d like a taste of what it was like to be in the room for a standing-room only presentation, you can watch Portent’s own Tim Mehta give his full talk on finding and fixing friction in user experience here.
Expanding into a new international market is an incredibly exciting moment for a business, and can open up fantastic opportunities for new growth. If you’ve done the research and are confident that this is the best next step for your business, let me start by congratulating you.
And when it comes to international SEO as a necessary part of global expansion, it’s important to keep in mind that what you’ve learned in your home country won’t necessarily translate to every new market. (See what we did there?) While the use case for your product or service remains the same, simply translating your existing marketing copy may throw a wrench into the way that information is perceived by your future customers and search engines.
Lost in translation – A lesson about localization
I came to the U.S. in fifth grade. To this day, I have a very vivid memory of a “self-esteem building” exercise we did as a class. For the exercise, all students taped a blank sheet of notebook paper to the top of their desks. Then, everyone had to go around the room and leave positive comments about each one of their classmates.
At the time, I barely spoke any English. Aside from knowing how to say “hello” and “how are you?”, I relied on my pocket translator to help make sense of the world around me.
After everyone completed the task, we went back to our seats to review. The comments were anonymous. To me, so were the words.
I reached for my pocket translator and began translating every word, from English to Russian.
One of my classmates wrote, “you rock!” — A common phrase, usually interpreted in a positive regard. But think about the direct translation… “you stone!”
Here’s where my foreign brain immediately took me:
“Am I dull and boring?”
Since then, I learned that telling someone they rocked was a compliment. I also learned not be so trusting of my pocket translator. There’s a great and not-so-subtle lesson here for businesses looking to expand to new markets and new languages. Your good intentions won’t translate without a little extra effort.
Why Google Translate or translation chop-shops won’t cut it
Though technological advances in natural language processing have come a long way, when it comes to marketing your business in a foreign language, it’s best to avoid using direct computer translation without a heavy dose of human help from someone living in the market who knows your industry.
While programs like Google Translate may be sufficient to help you get directions or to order a meal in 100 different languages, pure translation services should not be trusted with the core messaging of your business. At least not yet.
The example from my childhood is harmless. However, consider how a direct translation of the word “rock” could impact your customer’s understanding of your product.
Say, for instance, you run a business that sells top of the line toothbrushes and you want to expand your market to Russia. Why toothbrushes? Because I’m a strong believer in dental hygiene (just ask my dentist).
Now imagine, somewhere in your marketing copy, you quote a happy customer: “this toothbrush rocks!”
Russian translation: “Эта зубная щетка камни!”
Direct meaning: “This toothbrush stones!”
I don’t know about you but something about a stone-like toothbrush doesn’t sound pleasant. If I came across that quote, I’d pass.
So what can you do instead? Hire a professional translator to both translate and localize your content. If you find a good one, they’ll be able to provide a far more accurate and engaging translation of your copy—which plays a major role in the way your customers will find you.
Going further with localization
Now that your content doesn’t read like a direct output from Google Translate, your next job is to make it feel as though it belongs in the targeted region. This means finding a way to fully localize it.
I think of localization as travel research. Before traveling abroad, I research cultural customs, learn a few basic communication phrases, and of course, get a good understanding of the local currency. Why? Because I don’t want to look like a complete idiot, especially not while wearing my obnoxiously bright, 60-liter travel backpack. #tourist
Beyond not wanting to look like a total goon, I also do that research because I want to avoid being immediately judged as the outsider. For a business, avoiding this moment of mistrust is key to growth and development in new markets.
Localizing your business means taking the extra step in getting to know the foreign market and the individuals who live there. It’s also a critical step in establishing your business as authentic to the new country and communicating your message without friction.
Beyond things like changing dollar symbols into rubles, you’ll want to ensure that other parts of your content are customized for things such as regional variations and cultural norms.
Even if you’re expanding from an English-speaking home base to another English market, you need to be mindful of regional differences. For example, if you’re marketing a U.S. based product to Brits, at the very least, you’ll want to have a good understanding of the most common regional variations that exist between the two.
Keep in mind, regional differences aren’t only international. If you take a look at the U.S. market, you’ll find search differences for “soda” vs. “pop”. While search engines are generally smart enough to understand the user’s search intent, the user is still likely to choose the option that is more familiar to them—this is based on the mere exposure effect.
Having a solid understanding of regional variations will improve your search results. After all, if you don’t know how your prospective customers are referencing your product or service, they’ll never find you.
Cultural norms can mean a lot of different things. A useful guide to the dimensions of cultural difference as you’re looking at international expansion is Hofstede’s Cultural Dimensions framework. Cultural norms vary region by region, city by city, and even home by home, but this model looks primarily at differences between countries. One of the best examples and most important as you’re planning your marketing strategy and brand positioning is in the difference between individualistic and collectivistic cultures.
If you’re marketing to an individualistic culture, you’ll want your message to focus more on the “I”. It is common for people of such a culture to have goals and desires that help them be independent. Countries like the United States, Canada, United Kindom, and Australia are well known for their individualistic views and practices.
Alternatively, if you’re marketing to a collectivistic culture, you’ll want to focus more on the “we”, highlighting groupthink mentality and focusing on solving goals that will benefit the well-being of a community, not just a single individual. Countries like China, Russia, and the Philipines are great examples of markets that fit into the collectivistic culture norms.
To state the obvious, this is the kind of foundational positioning and marketing work that would never, ever come back from a translation engine or an inexpensive translation service.
Tying it back to SEO
If you read our post on Content-First SEO, you already have a clear understanding of the content/SEO relationship. When it comes to the international expansion and SEO, the core lesson from that post still applies: SEO is about content.
Once your content is properly translated and you have a solid understanding of your localized target keywords, then (and only then) will you be able to make technical SEO changes that will help you show up in search for the exact phrases that your customers use to find a business like yours.
Whether you found a new way of referencing your product or service, or you identified new keyword opportunities, you’ll need to optimize your website with all of those new or refined targets in mind.
Where to next?
Expanding to a foreign market is incredibly complicated. We’re only scratching the surface of the cross-cultural marketing considerations here, and there are a huge number of fundamental marketing challenges from core positioning to brand or line extensions that you’ll need to work through before you ever get to international SEO and localization. But, beyond ensuring that your content is simply coherent and optimized for search, I hope you take this charge to create content that connects with the people in every market you serve.
I’m going to be brazen and say something extreme: you can’t do outreach without tools. At least, you can’t do good outreach. And you can’t do great outreach without great tools. You get it. A great digital marketing strategy is like a huge rock band—each player has a part to play, but they wouldn’t be anywhere without their instruments (picture a group of grown men and women making silly gestures in the air—no result, and I’d probably be a little embarrassed for them).
I’m sure tools vary from company to company, but after years at this, and going rounds with the good ol’ school of hard knocks and trial and error, these are my lucky 13. To give more tangible examples, I’m going to be pitching my fake travel site EuropeforaDollar.com (hey—a girl can dream).
Streak Chrome Gmail Extension
Streak is one of those tools that I don’t know how I ever lived without. There are lots of ways you can use Streak but for what it’s worth I only use the email tracking.
I can see if an editor has opened my email and how many times. Then, when they don’t reply, I can send them an innocent email asking if they had received my email when I know that they opened it TWELVE times. By the way, this tool can also drive you to the brink of insanity.
Twitter is a great tool for outreach, because, according to some, journalists, editors, and PR people make up a huge percentage of its users. Many editors openly call for pitches and quite a few of them make their email addresses public.
Just using a simple search for “pitch me” yields a lot of great results. The caveat here is that these editors get a LOT of email, so make that subject line pop and be sure to follow up. Twitter is also great for maintaining and building relationships with your contacts. Some editors spend way more time on Twitter than they do answering emails but I will not name names. You know who you are.
Google Advanced Search Queries
You just can’t beat a good, well-executed advanced Google search when you’re looking for guest post opportunities, but Google search does have a language all its own. If you’re less familiar, aHrefs has a pretty good list of queries you can use in Google along with explanations and a few examples. If you’re a search nerd and an outreach specialist, using combinations of these queries virtually guarantees you’ll never find the end of the internet when you’re looking for guest post opportunities.
For example, sometimes I’ll cast my net wide with something like “inurl:travel “write for us”” or just “inurl:blog travel.” As another example, if I were looking for niche Canadian travel sites that talk about European cruises I can search for “intitle:European Cruises inurl:.ca inurl:travel.” But really, you shouldn’t get THIS specific. In this example, I only got one really targeted result which isn’t super helpful, but you get the point: you can get really granular.
No one wants to spend a ton of time writing a guest post for a site that has a domain authority (DA) of 10 unless it’s super niche, super relevant, or it’s a new site that looks like it’s going to gain authority fast. For my fictional travel site, a publication called European Cruise Savings would be a jackpot, even if it only had a DA of 18.
Granted, anyone who has tried using the Moz Chrome extension for long periods of time knows it can be a little buggy. Take this for the wonderful free tool that it is, and press on. The Moz Bar is indispensable and I highly recommend it. You can see where it tells you the domain authority of a site as you search.
Buzzstream is the single best tool for promoting content. It makes prospecting for journalists and media outlets a breeze, not to mention sending emails with handcrafted and personalized templates. It’s even more valuable if you have a team doing the outreach, as you can see who has had the most success with different contacts and who has had none.
For the promotion of my content piece, an evergreen calculator called “European Cruise Calculator Tool” I’d create a project for each country and subprojects for each vertical, like “News,” “Lifestyle,” “Travel,” etc. and begin searching for great outlets to cover my free tool.
Buzzsumo is a powerful content analysis tool. It allows you to easily research who is writing about a particular topic (and you can then add those journalists to a project in Buzzstream). You can set up alerts to track your client’s brand mentions and links, and you can get a ton of helpful ideas when brainstorming for new content promotion campaigns.
For example, I used Buzzsumo to come up with my calculator tool idea above, and will use it to get alerts every time a media outlet covers it.. When I search for “Take a European Cruise” I can see that the writer, Annette White, is someone I might want to reach out to because she has an interest in cruise travel.
Ahrefs is such deep tool that it will take years to figure out all that it can do. It has its own ranking system similar to Moz’s, called domain rank (DR). You can see a complete history of referring domains, track new links coming in, and do a competitive analysis on your client’s competitors. The Link Intersect tool in particular will help you find links that all your competitors are getting and you aren’t, which is a goldmine. Ahrefs does have a Chrome extension but it’s not the most reliable, so I wouldn’t suggest it.
How can anyone do outreach without Moz? Not only do they provide a ton of content for learning about link building, but their tools are amazing. Moz’s tools are widely used by SEO specialists, but it is quite useful to the Outreach Specialist. The “Link Explorer” tool will give you an overview of a site’s backlink profile, who’s linking the most, the DA of those sites, and much, much more. It’s great for competitive analysis, or to track your own client’s inbound links.
Similar to Ahrefs, SEMRush is a very agile tool. I like the user interface a little bit better than Ahrefs, so I use them interchangeably. Occasionally SEMRush doesn’t have information that Ahrefs does, so I toggle between them.
I especially love the geo information that SEMRush has, which is especially helpful when targeting certain countries in your outreach campaign. For instance, if I was just targeting Canadians for my cruise calculator tool, I could find publications with the most Canadian traffic.
Like many of these tools, you’ll need a subscription to use it, but boy what a tool Cision is for outreach. Thousands and thousands of media outlets and contacts are listed in this search tool, and most of them are gold: email addresses, interests, role, and actual beat that they cover.
For instance, I could find the travel editor on most major publications and pitch them specifically for my calculator tool. I could also export a .csv of media outlets and contacts and import it into a project in Buzzstream. Get this tool.
RocketReach is my backup email finder. If I can’t find an editor or journalist’s email address in Cision, I use RocketReach. It’s pretty straightforward.
MailTester is a free tool where you can guess and test an email to see if it works. If I know that Bob Smith who is a film editor at Travel & Leisure and his email is bsmith@t&l.com, but I want to contact Joe Johnson, I could try jjohnson@t&l.com. MailTester will tell you if it’s a valid email address. Make sense? Simple tool, but helpful.
Similar Web Chrome Extension
SimilarWeb is a great extension in my nav bar. I can see geo traffic for every site I’m on, as well as monthly site visits for most sites. I then report to my client how much visibility a particular placement earned them. For instance, if I were to get coverage for my calculator tool on Travel & Leisure, I could report a whopping 9.1 million monthly site visits. Mic drop.
Where to from here?
The success of your outreach efforts is in your hands. Use some of these tools and your outreach rock band will be able to successfully play “Mary Had a Little Lamb.” Use them all and you’re looking at mastering “Bohemian Rhapsody.” Your choice—but I say, go big or go home.
TL;DR We publish digital content. We write and create for the internet and screens. But our content workflow dates back to the days of desktop publishing. This post suggests a better way that uses Github and Markdown. It gets away from word processors and folders, helping you create content that’s ready for reuse. It uses readily available tools that won’t bloat your work or lock you into proprietary formats. Read on for a step-by-step walkthru. With a few stops for ranting and raving.
A couple years ago I measured how I spend time when writing. I wanted to see where/if I could tweak my content workflow. The result was so depressing I wore the same pair of cycling socks for a week:
It takes me about 90 minutes to write a 750-word blog post. In that time, I spend:
45 minutes writing
15 minutes editing
30 minutes doing… what exactly?
Wait. What? That can’t be right. I spend a third of my time on a writing project not writing.
What. The. Actual. Fucklebucket.*
So I checked my notes.
33% of writing time is sucked out of me by time-wasting stupidity:
Flailing around, trying to clean up a formatting problem in a blog post
Troubleshooting software problems
Digging up the second version of my intro, which I think I liked better than the fifth and sixth, but I honestly can’t remember because I usually just delete and type over stuff
Trying to figure out how the hell I create a smart quote
Reformatting a whole Word doc because the editor wants it in Google Docs format
Recapturing images because they’re all embedded in the document, and they look like poo
All these examples fall into two groups: Formatting and version management. Both are our jobs. Even if we have our own editorial and design team
HAHAHAHA DID I JUST SAY THAT BECAUSE YEAH SURE I HAVE 3 PEOPLE AT MY BECK AND CALL JUST TO MAKE SURE MY STUFF LOOKS GOOD HAHAHAHAHAAAAAAA cough sniff
we’re responsible for the first pass. Formatting and version management is part of our job.
And we suck at them.
We worry about flinging the next blog post over the wall, instead of creating versatile content that we can publish anywhere and reuse anytime.
We’ve Got Problems
We think about decoration first, structure second.
We worry about flinging the next blog post over the wall, instead of creating versatile content that we can publish anywhere and reuse anytime.
We save five minutes now and cost ourselves hours later.
We spend our time on writing, editing, and publishing. But we ignore the stuff that moves content between those steps: Workflow. So our workflow is riddled with potholes and broken glass. We still have to move from writing to editing to publishing. But instead of helping, the workflow slows us down.
An Outdated Workflow For Outdated Tools
We’re working on the internet, but we use tools created for desktop publishing in the 1980s. Yes, even you fresh-faced youngsters with healthy backs and no ear hair are using the software equivalent of antiques.
I’m talking about Word Processors. Word processors weren’t designed for writing. They were designed for typing and decorating. Word, Google Docs, and their devil-spawn relatives were designed to bring bold type and Comic Sans to the masses. They are not designed to help writers create, manage, and publish structured content.
I’m also talking about storage. Files saved in folders—online or off—are a metaphor for paper and filing cabinets. They’re about as useful as the corded phone I now use as a doorstop. They were created so that one person (maybe two) could work on one piece of content at a time.
Sure, there are tools out there that claim to update all of this, streamline workflow, etc. They all become bloated, expensive versions of the word processors and dated storage metaphors they try to replace. Fancier potholes. Prettier broken glass.
So poof. A third of my time gone because I’m spending my time in Dropbox and plinking away in Microsoft Word and Google Docs.
We need a better way to do this.
Find A Better Way
This is what I’ve worked out: A workflow that creates structured, reusable content. It lets me manage versions across devices and multiple writers. And it works well for 750-word blog posts as well as chapter books.
Most important, it lets me spend more time writing.
There’s a nice benefit to all this, too: Your writing looks better on the web. Let’s face it: Most of us write for the web. We’re publishing blog posts day in, day out. Use this workflow and your writing is web-ready (I hate that phrase, but it works here). It looks better, loads faster, and gets view source-ing HTML geeks like me nodding with approval.
First, a successful workflow has to follow these four rules:
Structure > Format And Design
Structure matters more than format or design: The fact that you have say, a level one heading is more important than exactly how that level one heading looks, because it may look different on various platforms.
You can pour well-structured content into lots of different formats, from Microsoft Word to HTML. You can easily change the design of well-structured content using CSS or any other templates, including Microsoft Word and Google Docs.
You can also easily repurpose well-structured content, adapting it from prose to presentation to script.
So a good content workflow has to support structure first, format and design second.
Proprietary Tools Are Traps
Tools—word processors, for example—trap us with proprietary formats. They also trap us on specific devices.
Yes, I can open a Google Doc or Word file on my phone. Yes, I could edit it. But I won’t, because it’s awful.
I need a cross-platform process that’s as tool-independent as possible.
Repurposing Is Inevitable
Smart content creators re-use their work.
E-books become blog series
Blog posts become slide decks
Transcripts become blog posts
Word documents become Google Docs become HTML
I’m going to repurpose content. My workflow has to support that.
Good Writers Are Orderly Hoarders
I always seem to delete an old file moments before I need it. Or I forget why I made an edit. Or I edit in circles, ending up back where I started.
We need to keep everything for every project:
All revisions and drafts
All image files
All other supporting files
The full change history for all of the above
And we need to keep it in an organized format.
Setting Up Your Tools
You only have to do this work once. If you’ve never used these tools, there’s a learning curve that feels pretty short if you’ve been using desktop software.
Still, it takes a little patience to install new stuff. Take the extra time now, and you’ll save hours later.
Install Markdown Tools
First, you’re going to work in Markdown. I’ve already written about why, so I won’t waste too many words on it here. Markdown is a super-simple markup language that’s perfect for writers. Everything you’re reading right now started as Markdown.
There’s also a great tutorial here and a syntax cheatsheet here.
You can write Markdown using a plain text editor, then preview and export it to other formats, like HTML.
But there are a few Markdown-focused editors that make it easier. They still use plain text—no proprietary formatting—but they flatten the learning curve. My favorite is Typora.
Typora runs on OS X, Windows and Linux. Type in Markdown and Typora gives you an instant preview. So if I type:
It instantly turns that into a level one heading “Typora.”
And my longtime favorite, atom.io. Atom is a pure text editor, so it’s not as friendly, but it’s fast as heck and easy to customize. If you want to get nerdier, that’s your best choice.
You can use a standalone preview and export tool like Marked 2, or automate advanced templating and batching with Pandoc (very steep learning curve, very powerful). But for 99% of your work, Typora’s export features will work just fine.
Set Up Version Control With Github
Rule of Content Workflow Number Four: We need to be orderly hoarders. To do that, use Github.
Developers built it to support developers. But what do developers do? They type stuff into files. They save the files. They manage versions of those files while lots of other developers edit and contribute.
So this tool they built is perfect for writers, too. And while it looks scary, it isn’t that hard to learn.
It’s where I keep the shortcuts I use to create smart quotes and such. They’re handy if you’re in a hurry or can’t remember the HTML entity for a smart right quote (like me).
What You’ve Got
You’re ready to follow all four rules of content workflow:
Markdown and Typora help you create structure
You’re not using any proprietary formats
The resulting content will be easily repurposed from one format to another
Thanks to Github, you’ll be an orderly hoarder
Hopefully, this whole process didn’t take more than fifteen minutes. If it did, I apologize. Please throw empty beverage cans at me the next time you see me. Empty ones. It’s not like I stole your car or something.
Getting Ready To Write
I’ve got lots of steps here, but this takes about five minutes, start to finish:
Step 1: Create Your Work Folder And File
Your work needs a home. Create one in advance. I’m writing this to prep for Learn Inbound. I’m going to have images and some supporting files.
Now, create your markdown file.
Create a folder where your work will live. I’m writing this to prep for Learn Inbound. I’m going to have images and some supporting files. I create folders accordingly
Click File >> New
Save the file in your work folder. You can name it whatever you want. I’m not feeling all that creative, so I’ll name it “post.”
You’ll end up with this:
For bonus points, in Typora, click View >> Outline and you’ll see a nice heading-based outline of your work as you write. It’s a structural view of your content. And structure is what we’re all about, right?
Outline View In Typora
Outline View In Typora
Step 2: Create Your Github Repository
Git is pretty intimidating until you break it down into steps:
Open Github Desktop
Click File >> Add Local Repository
Choose your work folder
If Github Desktop asks, click create a repository here
Create a new repository in Github Desktop
Create a new repository in Github Desktop
Click Create Repository or Add Repository as relevant
Github does its thing. If you look in your folder and can see hidden files, you may see some stuff like this. They belong there. Let them go on with their lives:
You’ll also see the repository appear on the left, much like this:
Did you know there are 300-500 new web pages created every minute of every day? If that's not cause for reflection on the value of content marketing, I don't know what is.
Confession time. There’ve been a few times where I’ve found myself asking (or exclaiming):
Does content marketing actually work?
It’s possible this existential crisis was related to a full week spent writing blog posts attempting to espouse the need for the latest organizational software. But there was and is a part of me serious about trying to understand “my place” in the marketing world. Does the work I do as a content marketer really support and impact the overall success of a business, or am I shouting with a chorus of others into the proverbial void?
If you’re looking for some compelling or reassuring stats, our friends over at Content Marketing Institute have a great set of quant-data to share with your boss. Maybe even follow up with some of the strategies from our post on how to sell content strategy to your boss. So by the numbers, that’s a “yes, content marketing does work.” But if we already have the numbers as proof, how could I still have lingering questions about the value of adding more and more content on behalf of great brands?
One powerful statistic from this round-up that jumped out at me is attributed to Neil Patel:
Content marketing leaders experience 7.8 times more site traffic than non-leaders
Whenever I read a statistic like this, I start thinking backward to what it actually took to become part of the “leaders” group. Successful brands that are also purveyors of amazing content makes me wonder: which came first, the success or the content? The answer to that quandary is beyond the scope of this post, but it’s worth observing that they’re most certainly correlated, though causation is often tough to prove.
To me and to the team at Portent, “doing content marketing correctly” means connecting with your target audience through empathic storytelling, while keeping in mind what a successful interaction means to your organization. For some, a “like” on Social or a reTweet is an indication that their content marketing strategy is working, while others are looking for a larger return on investment (like a sale or subscription confirmation). Again, if you’re looking for help thinking through how and what to measure for content, start here.
chart of KPIs
Courtesy of Cate McGehee
For me (and part of what gets me excited about coming to work at Portent in the morning), effective content marketing makes readers think and feel something new. By extension, I’m a huge believer in the impact of storytelling (or D&D style World Building, if you’re Ian Lurie) as a means to meaningfully and impactfully engaging with an audience. To become one of the coveted “content marketing leaders”.
For example, my favorite parts of the Olympics (besides the obviously impressive feats of athletic strength) are the P&G tribute ads to the mothers of the Olympians. My heart surges and my eyes well with tears watching the moms make sacrifices for their children to pursue their dreams of Olympic glory. I’m not responding to Proctor and Gamble, purveyor of consumer goods; I’m responding to the reminder of the everyday struggle families (mothers in particular) weather to support their kids. That Proctor and Gamble happens to manufacture dozens of common household goods is second to the feelings I now associate with their ads.
Why do we connect with stories?
On a biological level, our brains are literally wired to respond to stories and narratives. Viewing or experiencing emotional stories triggers the production of two specific hormones: cortisol and oxytocin. Cortisol is a stress hormone and grabs the attention of the viewer, while the release of oxytocin promotes feelings of empathy and connect, which researchers say may make viewers “more generous and trusting.” MRIs have shown that our brains respond the same way to hearing a story as they do to actually experiencing it, which is good news for content marketers. Bottom line: storytelling in content marketing is effective because our brains crave stories.
Why use storytelling?
Stories are how we figure out what’s going on, why we feel the way we do, and what to do with the information we’re given. Consciously or not, when we read or interact with marketing content, we’re looking to connect with information.
Good stories give us “all the feelings” and said feelings persuade us to form an opinion about a topic, or by extension, a brand. Connecting to a story bigger than oneself is a powerful feeling, which is why emotional responses are a leading indicator of the consumer’s intent to buy.
At the end of 2016, Google released a two minute video compilation of the major highs and lows of the year, as determined by the most popular searches of 2016. The film begins with scenes of chaos, the result of terrorist attacks and military conflict. Even if we didn’t personally live through the events, we are reminded how tumultuous 2016 really was. The video transitions into shots of athletic triumph, of humanity coming together, and support for causes greater than ourselves.
The story this Google video tells us is that through tragedy, pain, and violence there are still messages of hope, reasons to move forward, and the overall presence of love.
If you present a problem with a solution—your solution—the audience is more likely to see themselves (or their company) as the subject within your narrative. Even better news is that they’ll most likely respond to it.
And it doesn’t matter the scale or the topic, storytelling works. Whether you’re shopping for camping gear or a new apartment, you are searching for an element to help complete the story: “becoming one with nature” or “transforming into the sophisticated downtown socialite.”
Beyond the obvious bottom-line goal of convincing viewers to use Comcast, the ad creates a story arc in which anyone can find unique ways to connect with their world. And if they’d like a little help from Comcast, that’s there whenever they’d like it. The message in the story is that Comcast has something for everyone and becoming a Comcast member can impact more than just your entertainment preferences.
When we use content marketing, we work to sell an idea or product without coming right out and screaming “BUY, BUY, BUY!“ Remember the first stat from the Content Marketing Institute article: 200 million people now use ad blockers.”
The ads or marketing campaigns that work best are not those that interrupt or pester us with unrelatable messages. The ads we remember and respond to are the ones where we can see ourselves as the hero, or at the very least, the co-star. Give the reader an opportunity to see themselves in the narrative, and you’ve gained a customer or at very least, a fan.
The reasons the two examples above, Google and Comcast, are so compelling is that they present simple concepts:
Love conquers hate
Perceived limitations don’t have to be limiting
Nowhere in the ad does Google encourage viewers to use the search engine, they just tell us the story of a hectic 2016, reinforced with examples. The same with Comcast: Emily may not have her sight, but she has her imagination. When we respond to ads like these, we are responding to the message that the brand puts forth, and when we buy from or use these brands we’re buying into the message they represent.
How to tell if the stories you’re telling are making an impact
The other half of doing content marketing the right way is knowing when and deliberately measuring how your content is successful. That means identifying and tracking metrics to determine when your content is achieving the desired impact.
For example, I’m a compulsive reader: I will read anything. Biographies, novels, comment sections, instructional manuals, the back of a shampoo bottle. If copy is involved, I am likely to consume it.
As an obsessive reader, you can gauge my successful interaction with a piece of content with two questions:
Did I get to the end of the piece?
Did I learn something?
If I answer yes to one or both of these questions, I consider that a successful piece of content.
The challenge for marketers is to create a content strategy that drives us to produce the right pieces, distribute them to the right people, and to define what we consider to be a successful interaction with content. Beyond editorial calendars, your content strategy needs to include clear definitions for your KPIs or content measurement criteria, so that you know you’re not stepping back in front of that proverbial void, preaching to no one.
In addition to KPIs, A/B testing content can be very enlightening when determining whether a particular piece of content is achieving desired results. Savvy content marketers will take the time to prepare multiple versions or test updated content against existing content to hone in on which pieces of content audiences are responding to. It’s important to remember that you should have more than one KPI. And, as your business grows, so should your metrics.
Easy ways to utilize storytelling in your content
There are endless ways to tell stories through marketing outside of the traditional marketing content, like long-form blogs and white papers. Good stories sell us more than soft drinks—they sell us ideas and concepts.
Here are a few areas of your content where you can inject storytelling, no matter how small:
An impactful headline gets to the point and attracts the reader’s attention. It might also be the only thing the reader remembers, so don’t waste the opportunity to make an impact. NPR has a great checklist for putting together headlines, and includes advice like making your headlines specific, easy to understand, and capture the spirit of the story.
If ever there were a place to tell a story or give the reader aspirational feelings about what you sell, a product description is that place. Of course, some products and service lend themselves more readily to action-packed narratives. For example, REI’s website is not trying to sell outdoor gear. Rather, it’s an overarching, how-to guide for adventure seekers, regardless of their experience levels. Consumers browsing through outdoor equipment are looking to fulfill a story they have in their mind: one where they’re the star of an action-packed trip. Storytelling in product descriptions can be as simple as leading the reader to conclude that they, too, can conquer any mountain…if they have the right hiking boots. Product descriptions don’t sell a product, they sell the idea of what one can do with the product.
Perhaps not the most intuitive part of storytelling via content, but hear me out.
A list can tell a very succinct, organized story.
Using lists to your advantage means presenting ideas and concepts to a consumer and allowing them to draw their own conclusions about what comes next in the story. Additionally, lists are cost effective and require less time to both write and consume. With the advent of the internet, readers have a shorter attention spans, so a list with catchy copy and supporting images cuts right to the point. Plus, a good list gives you plenty of opportunities to link to other parts of your site to continue crafting a story. For examples of entertaining, engaging lists used in content marketing, look no further than the impact of the BuzzFeed listicle. Readers respond to this type of content because it gives them the freedom to attach meaning to the parts of the list that resonate most with them.
Giving away a free bit of advice or sample of what your service does shows prospective clients that you know what you’re talking about, and you’re willing to back it up with evidence. A guide or template that walks the consumer through the steps needed to achieve desired results shows that you know what you’re talking about. Organizational software company Lucidchart offers free diagram templates on its website. This allows users to experience the product before they commit to anything long-term.
Start telling a story, any story.
The best storytellers have honed their craft with years of practice. And while some may seem like natural-born public speakers, or authors, or designers, the impactful ones know that practice makes perfect. They also know that hitting the nail on the head the first time is unlikely, which is why they constantly test their material. For every decent headline you read I can personally assure you there were somewhere precisely between 3 and 300 throw-away headlines that came first.
Not every story will be a winner, propelling you to the rarified air of “content marketing leader” of your category (We should all be so lucky.) But having the courage to press on and keep telling stories will dictate whether or not (and how) your content influences audiences. The stories you develop and share in your content marketing pieces don’t have to be everything to everyone. They only have to resonate enough with your audience to persuade them to like, share, follow, or eventually buy what you’re selling.
The battle for the voice of the internet has begun. In one corner, we have computer programs fortified by algorithms, Artificial Intelligence, Natural Language Processing, and other sexy STEM buzzwords. In the other corner, we have millions of copywriters armed with the only marketable skill a liberal arts education can provide: communication. Who will lol the last lol?
Writers, your jobs are probably safe for a long time. And content teams stand to gain more than they stand to lose.
I remember the day someone told me a computer had written a best-selling novel in Russia. My first thought? “I need to get the hell out of content marketing.”
The book was called True Love—an ambitious topic for an algorithm. It was published in 2008 and “authored” by Alexander Prokopovich, chief editor of the Russian publishing house Astrel-SPb. It combines the story of Leo Tolstoy’s Anna Karenina and the style of Japanese author Haruki Murakami, and draws influence from 17 other major works.
Frankly, that sounds like it’d make for a pretty good book. It also sounds a lot like how brands create their digital marketing strategies.
Today, every brand is a publisher. Whether you’re a multi-billion-dollar technology company or a family-run hot sauce manufacturer, content rules your digital presence. Maybe this means web guides, blog posts, or help centers. Maybe it means a robust social media presence or personalized chatbot dialogue. Maybe you feel the need to “publish or perish,” and provide value and engagement in a scalable way.
Brands require a constant influx of written language to engage with customers and maintain search authority. And in a way, all the content they require is based on 26 letters and a few rules of syntax. Why couldn’t a machine do it?
In the time since I first heard about True Love, I’ve moved from content writing to content strategy and UX, trying to stay one step ahead of the algorithms. But AI in general and Natural Language Processing in particular are only gaining momentum, and I find myself wondering more and more often what they’ll mean for digital marketing.
This essay will endeavor to answer that question through conversations with experts and my own composite research.
Portent’s Matthew Henry Talks Common Sense
“The Analytical Engine has no pretensions to originate anything. It can do whatever we know how to order it to perform.”
-Lady Ada Lovelace, 1842, as quoted by Alan Turing (her italics)
Lady Lovelace might have been the first person to contend that computers will only ever know as much as they’re told. But today’s white-hot field of machine learning and Artificial Intelligence (AI) hinges on computers making inferences and synthesizing data in combinations they were never “ordered to perform.”
One application of this Machine Learning and AI technology is Natural Language Processing (NLP), which involves the machine parsing of spoken or written human language. A division of NLP is Natural Language Generation (NLG), which involves producing human language. NLP is kind of like teaching computers to read; NLG is like teaching them to write.
I asked Portent’s Development Architect Matthew Henry what he thinks about the possibilities for NLP and content marketing. Matthew has spent over a decade developing Portent’s library of proprietary software and tools, including a crawler that mimics Google’s own. Google is one of the leading research laboratories for NLP and AI, so it makes sense that our resident search engine genius might know what the industry’s in for.
I half expected to hear that he’s already cooking up an NLP tool for us. Instead, I learned he’s pretty dubious that NLP will be replacing content writers any time soon.
“No computer can truly understand natural language like a human being can,” says Matthew. “Even a ten year old child can do better than a computer.”
“A computer can add a million numbers in a few seconds,” he continues, “which is a really hard job for a human being. But if a cash register computer sees that a packet of gum costs $13,000, it won’t even blink. A human being will instantly say Oh, that’s obviously wrong. And that’s the part that’s really hard to program.”
“Knowing that something is obviously wrong is something we do all the time without thinking about it, but it’s an extremely hard thing for a computer to do. Not impossible—to extend my analogy, you could program a computer to recognize when prices are implausible, but it would be a giant project, whereas for a human being, it’s trivial.”
It’s not news that there are things computers are really good at that humans are bad at, and some things humans are really good at that computers can’t seem to manage. That’s why Amazon’s Mechanical Turk exists. As they say,
“Amazon Mechanical Turk is based on the idea that there are still many things that human beings can do much more effectively than computers, such as identifying objects in a photo or video, performing data de-duplication, transcribing audio recordings, or researching data details.”
Amazon calls the work humans do through Mechanical Turk “Human Intelligence Tasks,” or HITs. Companies pay humans small sums of money to perform these HITs. (A made-up example might be identifying pictures where someone looks “sad” for 10 cents a pop.)
Matthew might instead call these HITs, “Common Sense Tasks,” like knowing a pack of gum shouldn’t cost $13,000.
“People underestimate the power of common sense,” Matthew says. “No one has ever made a computer program that truly has common sense, and I don’t think we’re even close to that.”
And here’s the real quantum leap for not only NLP but Artificial Intelligence: right now, computers only know what they’ve been told. Common sense is knowing something without being told.
It sounds cheesy to say that our imaginations are what separate us from the machines, but imagination isn’t just about being creative. Today, computers can write poetry and paint like Rembrandt. Google made a splash in 2015 when the neural networks they’d trained on millions of images were able to generate pictures from images of random noise, something they called neural net “dreams.” And in 2016, they announced Project Magenta, which uses Google Brain to “create compelling art and music.”
So it’s not “imagination” in any artistic terms. It’s imagination in the simplest, truest form: knowing something you haven’t been told. Whether it’s Shakespeare inventing 1,700 words for the English language, or realizing that kimchi would be really good in your quesadilla, that’s the basis of invention. That’s also the basis of common sense and of original thought, and it’s how we achieve understanding.
To explain what computers can’t do, let’s dig a little deeper into one of the original Common Sense Tasks: understanding language.
Defining “Understanding” for Natural Language
NLP wasn’t always called NLP. The field was originally known as ”Natural Language Understanding” (NLU) during the 1960s and ‘70s. Folks moved away from that term when they realized that what they were really trying to do was get a computer to process language, not understand it, which is more than just turning input into output.
Semblances of NLU do exist today, perhaps most notably in Google search and the Hummingbird algorithm that enables semantic inferences. Google understands that when you ask, “How’s the weather?” you probably mean, “How is the weather in my current location today?” It can also correct your syntax intuitively:
And it can also anticipate searches based on previous searches. If you search “Seattle” and follow it with a search for, “what is the population,” the suggested search results are relevant to your last search:
This is semantic indexing, and it’s one of the closest things out there to true Natural Language Understanding because it knows things without being told. But you still need to tell it a lot.
“[Google’s algorithm] Hummingbird can find some patterns that can give it important clues as to what a text is about,” says Matthew, “but it can’t understand it the way a human can understand it. It can’t do that, because no one’s done that, because that would be huge news. That would basically be Skynet.”
What is Skynet
In case you don’t know what Skynet is and you’re also too embarrassed to ask Matthew, too, here’s the Knowledge Graph.
Expert Opinion: NLP Scholar Dr. Yannis Constas on Why Language is So Freaking Hard to Synthesize
To find out what makes natural language so difficult to synthesize, I spoke with NLP expert Dr. Yannis Constas, a postdoctoral research fellow at the University of Washington, about the possibilities and limitations for the field. 
There are a lot. Of both. But especially limitations.
[Note: If you don’t want a deep dive into the difficulties of an NLP researcher, you might want to skip this section.]
“There are errors at every level,” says Yannis.
“It can be ungrammatical, you can have syntactic mistakes, you can get the semantics wrong, you can have referent problems, and you might even miss the pragmatics. What’s the discourse? How does one sentence entail from the previous sentence? How does one paragraph entail from the previous paragraph?”
One of the first difficulties Yannis tells me about is how much data it takes to train an effective NLP model. This “training” involves taking strings of natural language that have been labeled (by a human) according to their parts of speech and feeding those sentences into an algorithm, which learns to identify those parts of speech and their patterns.
Unfortunately, it takes an almost inconceivable amount of data to “train” a good algorithm, and sometimes there just isn’t enough input material in the world to make an accurate model.
“When we’re talking about a generic language model to train on, we’re talking about hundreds of millions of sentences,” he says. “That’s how many you might need to make a system speak good English with a wide vocabulary. However, you cannot go and get hundreds of millions of branded content sentences because they don’t exist out there.”
Yannis says he once tried to make an NLP model that could write technical troubleshooting guides, which might be a popular application for something like corporate support chatbots. He only had 120 documents to train it on. It didn’t work very well.
Right now, his research team is trying to figure out a way to combine corpuses of language to overcome the twin pitfalls of meager input:
Output that doesn’t make much linguistic sense
Output that all sounds pretty much the same
“We tried to take existing math book problems targeted at 4th graders and make them sound more interesting by using language from a comic book or Star Wars movie,” says Yannis. “That was specific to that domain, but you can imagine taking this to a marketing company and saying, ‘Look, we can generate your product descriptions using language from your own domain.’”
That’s the grail of NLP: language that is accurate to the domain yet diverse and engaging. Well, one of the grails. Another would be moving past the level of the sentence.
“80 to 90 percent of the focus of NLP has been on sentence processing,” says Yannis. “The state-of-the-art systems for doing semantic processes or syntactic processing are on a sentence level. If you go to the document level—for example, summarizing a document—there are just experimental little systems that haven’t been used very widely yet…The biggest challenge is figuring out how to put these phrases next to one another.”
It’s not that hard for an algorithm to compose a sentence that passes the Turing Test, or even hundreds of them. But language is greater than the sum of its parts, and that’s where NLP fails.
“When you break out of the sentence level, there is so much ambiguity,” says Yannis. “The models we have implemented now are still very rule-based, so they only cover a very small domain of what we think constitute referring expressions.”
“Referring expressions,” Yannis tells me, are those words that stand in for or reference another noun, like he, she, it, or these. He uses the example, “Cate is holding a book. She is holding it and it is black.” An NLP model would probably be at a loss for realizing that “she” is “Cate,” and “it” is “the book.”
“It’s something that sounds very simple to us,” says Yannis, “because we know how these things work because we’ve been exposed to these kinds of phenomena all our lives. But for a computer system in 2017, it’s still a significant problem.”
Models are also inherently biased by their input sources, Yannis tells me. For example, we’re discussing an AI researcher friend of his who combines neural networks and NLP to generate image descriptions. This seems like it would be an amazing way to generate alt tags for images, which is good for SEO but a very manual pain in the ass.
Yannis says that even this seemingly-generic image captioning model betrays bias. “Most images that show people cooking are of women,” he says. “People that use a saw to cut down a tree are mostly men. These kinds of biases occur even in the data sets that we think are unbiased. There’s 100,000 images—it should be unbiased. But somebody has taken these photos, so you’re actually annotating and collecting the biases.
“Similarly, if you were to generate something based on prior experience, the prior experience comes from text. Where do we get this text from? The text comes from things that humans have written…If you wanted to write an unbiased summary of the previous election cycle, if you were to use only one particular news domain, it would definitely be biased.”
Sometimes it’s right. Sometimes hilariously wrong.
Finally, perhaps one of the biggest hurdles for NLP is particular to machine learning. Interestingly, it sounds a lot like something Matthew said.
“One common source of error is lack of common sense knowledge,” says Yannis. “For example, ‘The earth rotates around the sun.’ Or even facts like, ‘a mug is a container for liquid.’ You’ve never seen that written anywhere, so if a model were to generate that it wouldn’t know how to do it. If it had knowledge of that kind of thing, it could make the inference that coffee is a liquid and so this mug could be a container of coffee. We are not there. Machines cannot do that unless you give them that specification.”
 Note: This interview was conducted in May of 2017. Quotes from Yannis only reflect his work, experience, and understanding at the point of this interview.
NLP is Hard. So is Programming. English is Harder.
Let’s start with an assumption: As a brilliant digital marketer, you know that understanding your prospects and customers intimately is the cornerstone for creating great content and a successful marketing program. But how can you understand them without asking them the right questions?
When you think of formal “user research,” words that come to mind might be “expensive, “time suck,” or “overwhelming.” If you’ve been at a bigger company, they might have had the luxury of an outside user research firm to handle things such as recruiting, session facilitation, or a card-sort. Today, we’re talking about the basics.
Just because you don’t have access to a user research firm who can do the recruiting for you, doesn’t mean you should be too intimidated or worried about finding participants yourself.
In this blog post, we’ll go through 6 methods for recruiting participants yourself, so you don’t spend your time and money trying to reach the wrong people.
And if you need a little more convincing on why user research is critical to creating great user experience which absolutely includes content, check out “UX Without User Research Is Not UX” from the Nielsen Norman group.
With around 234 million unique users and thousands of communities or “subreddits,” Reddit can be an excellent place to connect with people over a wide range of niche topics.
During a recent user research project, I needed to interview sustainability managers. Not surprisingly, I didn’t know anyone with this job title in my immediate circle of friends or colleagues. I asked the client if they could connect us with some of their contacts, but with a brand new product, all their prospects were very early in the sales process. They understandably weren’t ready to risk botching the sale.
Skipping user research with this target audience wasn’t a choice; we needed to find out what types of content they consume, what pain points they have at work, and what social media sites they frequent. I knew I had to get scrappy, fast.
I found a sustainability subreddit and posted the following message:
We received a couple of comments and a DM from this post. Although that might not sound like a lot, the DM turned into an insightful 30-minute phone interview. The person I interviewed was also interested in the research we were doing and when asked, sent an email out to his network asking if anyone else was interested in participating in our research.
Ask participants for referrals
This might sound incredibly obvious, but you’d be shocked how many companies forget to do it. Do not be afraid to ask the people you interviewed if they would be willing to help you find more participants.
If you’d prefer to be as unobtrusive as possible during the in-person session, I recommend always asking at the end of the interview whether or not it’s okay to reach out to them again with any additional questions. In your follow-up email, you can ask if there is anyone else in their network who might want to participate.
If you’re worried about the participant sending you an unqualified referral, remember that you should create screener questions to send to every potential participant before you interview them. Whether you decide to interview them is ultimately up to you.
In this case, after I sent this email the participant connected me with another sustainability manager who gave me another highly informational interview. So far, and without much effort, I was 2/2.
Organic Social Media
If you need to interview people with a particular job title who live in certain areas, try using LinkedIn. You’ll need to use their InMail feature, which will allow you to message people who aren’t in your network. You can sign up for a 30-day free trial of Inmail and then pay $29.99 a month if you need it for longer. The one downside is that (at the time of this writing) you only receive 5 InMail credits per month.
Once you’ve thoroughly combed through user profiles and decided that someone’s a good match, you can send them a message explaining the study and why you want to learn from them.
It’s important to tell them how interviewing them is going to help you or your organization. People like feeling useful.
Another reason why I like using LinkedIn for recruitment is that it shows mutual connections. If you find someone who looks like a good candidate and they’re connected to someone you know, you can and should mention that in your message. You can also reach out to the person you know and ask if they can help you start the conversation.
Pro-Tip: Because other people’s time is precious, I always offer some kind of incentive to thank them for their time. Depending on where you work, this could be a discount on your services, a gift card to your store, or an Amazon gift card.
Send out an email or Slack message to your company and see if they know anyone who fits the criteria for your study. If your company is large and diverse enough, you might be pleasantly surprised at how many people they can connect you with. And telling people in your company that you’re doing user research gives you an opportunity to educate other people in your office about the importance and impact of what you’re doing.
Paid Social Media
If you can’t find enough people through organic social, try putting a small ad spend behind your posts to boost their reach. On Facebook, you can use both demographic and psychographic targeting to find people who fit the criteria for your study.
For a recent site launch project, I needed to find people between the ages of 26-65 who watched documentaries on at least a monthly basis.
We segmented the ads into 4 different sets based on interest (science, history, general, and documentaries), so we could tailor our questions to each group. Then we ran the ads for seven days with a $150-$200 budget to reach our target number of participants.
Each ad drove traffic to a screener I created in Google Forms. Based on the data, I was able to determine who was qualified to move on to the round of interviews.
I haven’t yet experimented with recruiting participants via LinkedIn or Twitter ads, but if you try it, let me know how it goes!
If the scrappy or DIY approaches above seem like too much work or you still need more participants to round out your research, I recommend using Respondent. Respondent is a web-based platform that not only helps you find qualified participants, it schedules interviews and allows you to pay them all in one place.
You can also choose whether you want to conduct your research remotely or in-person, how much of an incentive you’re willing to give, and how much time you need to interview respondents.
Once you pick all your targeting criteria and write your screener questions, respondents will flood your inbox within the Respondent interface and you can decide whether or not you want to schedule an interview with them. There’s a per-person fee for each interview, but it’s relatively small. To give you a general idea, I recently conducted a 30-minute phone interview with a participant, paid them a $50.00 incentive and the recruitment fee was $17.50 (it’s always 35% of the incentive).
If you take nothing else away from this article…
Just because you don’t have a fancy recruitment agency or a mountain of cash, doesn’t mean you shouldn’t do user research. There is so much incredible knowledge to gain from talking to actual users about what they want and need, and we hope you’ll explore some of these approaches. Let us know which ones work for you or if you have any other recruitment tips!
They suck because they are wish lists, not sensible plans. They’re inflexible. They use terminology your team and clients don’t understand, like “10x” and “long form.”
They suck because we try to jam our marketing teams into calendars. That’s backward. We should build calendars based on our teams.
Instead of wrapping your team around a calendar, build a calendar around your team
You should create editorial calendars based on your team’s strengths and weaknesses: Subject matter expertise, writing specialties, efficiency, and skill. Sure, you hire based on what you want to create. But smart planners start by asking, “What resources do I know I can get? What additional resources do I need to complete the work? How will I close that gap??” That way you create an achievable calendar, not a thrill ride.
We should also have a single, consistent way to note what we’re going to need and when we’ll need it. Both your team and stakeholders must see the same information you do when they read the calendar. That way you create a plan everyone can follow.
My team has incorporated a structure like this into the planning process:
We classify using content types that reflect what we know we can produce
We categorize content using branding levels that reflect how sales-y a particular piece should be. That tells us what general skill set we’ll need (marketing copywriter versus tech writer, etc.)
Together, the structure creates an achievable content calendar and provides a common language to talk about that calendar. Here’s how it works:
Content types don’t show resources required. Resources come and go. They are based on what’s possible at any point in time. They indicate level of complexity. They call out the sophistication of the deliverable.
If you’re a skier or mountain biker, you will know these labels right away. If you’re not, it’s OK. They’re easy to remember:
Content Types Notation Shows What's Possible
Green circle content is whatever you can produce once per week. If your organization and resources let you create one blog post per week, then that’s your green circle. If the best you can do is one tweet per week, then that’s it.
Make it valuable to your customers. Make it polished. Make it consistent. Just be sure you actually make it.
What to expect: You want all content to generate business, of course. But keep expectations low. “Low” doesn’t mean giving up, or producing lousy content. It means a sustainable content set that creates momentum for prospective customers, for your team and for your stakeholders. Create green circle content to maintain engagement, build content hubs, and answer individual audience questions. Get readers to move “up” to blue square and black diamond. Typical green circle content rarely generates last-click leads or conversions.
Blue square content is whatever you can produce once per month. If you can write one blog post a month, that’s your blue square. If you can write a 3000-word interactive piece, that’s blue square. If you can build a sophisticated, public-facing toolset every month, that’s your blue square, and I envy you.
What to expect: If you can only create something once a month, it has to be a measurable business booster. It should generate shares and links, serve as the center of content hubs, and kick readers “up” to black diamond content. It may attract marketing qualified leads and other mid-level conversions. But remember, it’s whatever you can create once per month.
Black diamond content is the toughest and carries the highest expectations. You can produce it once per quarter, at most. It might be a blog post (I’ve seen it). More likely it’s a high-production-value, high-value piece. It might be interactive. It might be a tool. It could be a video series. It might be an award-winning long-form piece that requires designers, writers and developer time.
What to expect: At this level of effort, have lofty goals. It should attract folks who have never read anything by your organization. Aim for quality lead generation, high-quality organic link growth, and media visibility. Ideally, this content is evergreen: It should hold its value for a long time.
What, No Examples?
Nope. Remember, content types are about what’s possible for your organization, not the one that just won a Webby Award for Best Longform Piece of the Century.
I will say that Black Diamond content often requires a designer, research time on the part of the writer or a separate subject matter expert, and possibly a developer. It might be a tool or a fancy blog post. A single writer can create green circle content. Blue square sits in between.
Know Your Team!!!
Remember: Instead of trying to wrap a team around a calendar, you’re building a calendar around a team. You need to know your team, and know who you can add to that team. You should set a content plan that’s a stretch. But be aware of what’s possible. Creating an editorial calendar that requires impossible growth and eye-popping productivity isn’t inspiring. It’s soul-sucking.
Not all content is brandless or a sales pitch. We use three classifications:
Content branding levels show what you'll need
In a calendar, branding levels are better guidance than, say, word counts. Combined with content types, branding levels tell content teams what to create. You can provide details like word counts later.
Lightly branded content may have your logo or use your blog page template. That’s it. No sales pitch and no advice that requires your services. This is the stuff that anyone can use whether they use your services or not.
Social media posts and tweets that offer tips are lightly-branded. So are some blog posts, long-form content, videos, etc. Anything may be lightly-branded, as long as there is zero sales pitch. This blog post might be considered lightly-branded content.
To create this kind of content, you need an ambidextrous writer who can create high-value, zero-pitch stuff. That might be a great marketing copywriter who is also a great nonfiction writer, or a “blogger.” Our marketing director, for example, was a comparative literature major. Someone at Portent might even be an ex-lawyer. Cough.
Moderately branded content is lightly branded but has your product, company or service as the focus. It doesn’t say “buy this!” but it does say “here’s how you’d accomplish this with our product, but you could use someone else’s if you prefer.”
Will it Blend is classic moderately branded content. It provides entertainment value using the brand’s product.
Lea & Perrins publishes recipes using their products. You could use any other sauce, though. That’s great moderately branded content. Great. Now I’m hungry.
Moderately branded content is least-understood and most valuable because it bridges the gap between lightly-branded material and your ultimate call to action. Think carefully about how you can work it into your content calendar.
To create moderately-branded content, you need a writer used to writing both marketing copy and lightly branded stuff. This is the toughest assignment.
Great heavily-branded content requires a great marketing copywriter.
Somehow, heavily-branded content gets left out of most content plans. But we have to sell stuff, right? Include it, and plan for it. Marketing isn’t an existential experience. At some point, visitors must convert.
Why You Should Care About Branding Levels
Branding levels are easy shorthand for the kind of writers you’ll need: Marketing copywriters for heavily-branded, “bloggers” or whatever you want to call them for lightly-branded. Heavily-branded may require input from product managers. Moderately-branded will probably require all of the above, and demand your absolute best talent.
Branding levels are a strategic planning tool: They show proportion of resources assigned to lightly, moderately, and heavily-branded content. For example, on an e-commerce site with 100+ SKUs, I might go with 40/20/40 lightly/moderately/heavily branded content (don’t neglect heavily-branded content!!!!). On a site with only a few products, or a services site, I’ll recommend 60/20/20.
Regardless, I now know exactly what kind of writers, topics, and material I have, and those I need to gather.
Create The Content Structure
Now I can create a content structure. It tells me what type of content we’ll write, when we’ll write it, and the branding level of that content:
The content structure brings it all together
Yours may look completely different. Remember, I built this structure based on my team and what they can accomplish in a week, month, or longer. It’s based on my team’s specialties. If there are gaps, I can fill them.
Now, I can expand it to a full editorial calendar with topics, assignments, and titles. But I’ll always use the same content types and branding levels and base my plan on the structure I created above. The types and branding levels form a standard language everyone can understand. They clarify expectations and requirements for your team and other stakeholders.
You don’t have to use the vocabulary I’ve described. Just use something simple and consistent, and make sure it answers three questions:
What can your team produce?
How long will it take?
What levels of branding do you need?
Do that, and you’ll create editorial calendars that don’t suck.
If you have any hand in the content on your company’s website or your clients’ websites, it’s important that data informs your decisions. Otherwise, you risk creating content that doesn’t meet your audience’s needs. But knowing this will only get you so far–it’s even more important that you know how to find and interpret the right data.
For many of us content folks, the idea of analyzing data can be intimidating. I know it was to me. But after learning the basics of Google Analytics, I warmed to the idea. Using data made my client recommendations stronger, and I felt more confident standing behind those recommendations.
Having ideas is great. Having ideas backed by numbers is even better.
Thanks to our analytics team, I learned about 4 reports in Google Analytics that made me a better content strategist. If you don’t have an in-house analytics strategist or you just want to understand a few things to kickstart your journey to data-informed content, I’ll show you how, when, and why to use these reports.
1. Behavior Flow Report
What: The Behavior Flow Report shows the path users traveled from one page or event to the next. You can use it to see what content keeps users engaged with your site.
Where: You can find it under Behavior > Behavior Flow
How: Although it’s important to look at metrics that show us how people interact with our content, we also need to look at what happens after people read our content to determine whether or not it’s helping us reach our business goals. The Behavior Flow Report is an excellent way to figure this out.
The report will look something like this:
Let’s look at an example. Let’s say I want to know whether the Portent SERP Preview Tool is leading to any contact form submissions. I would right-click on “Portent SERP Preview Tool” and click “Highlight traffic through here.”
Next, I would look at 1st interaction and notice where people go after they click on the SERP tool. After clicking on the group details for (>100 pages), we can see that our “Contact” page received 3.4K sessions after people click on the SERP tool.
If our primary goal is to have more people fill out the “Contact Us” submission and we think 3K is too low of a number, we may want to determine a way to make the “Contact Us” CTA more prominent on the page. If you’re wondering what the drop-off rate is (I know I was), this is just a proxy for Exit % for people who leave from that page without doing anything else on the site. But how is it different from the Bounce Rate? The only difference is that users could have seen a page previous to the one they left on.
2. Assisted Conversions
What: The purpose of the content on your site isn’t only to educate your users, it’s also to build trust and get people to do business with you. One way to tell if your content succeeds in this is to use the Assisted Conversions Report.
Where: To get to this report, go to Conversions > Multi-Channel Funnels > Assisted Conversions
How: If you haven’t already, you’ll need to set up specific goals in GA, which could include objectives such as newsletter subscriptions, ebook downloads, or visits to a specific landing page.
Once you’ve set up your goals, you can look at the data to see which landing pages helped you achieve those goals. And once you have that knowledge, you can determine if you should create similar content or find ways to optimize poorly performing content.
This report can also help you identify which content types perform best on what channels.
Here’s what you’ll see when you click on the Assisted Conversions Report:
Next, click on “Other,” type in “Landing Page,” and click on “Landing Page URL.”
From here you can see a list of the top landing pages that assisted in conversions. Here’s an example from Portent’s Assisted Conversions Report:
To pare this report down to specific conversions, use the drop-down list under “Conversion” in the upper left-hand corner of the report.