Timo Elliott's Blog | Business Analytics & Digital Business
This blog tracks innovation in analytics and social media, including topics such as big data, collaborative decision-making, and social analytics. Timo Elliott is an innovation evangelist and international conference speaker who has presented to business and IT audiences in over fifty countries around the world.
In a couple of weeks, I’ll be heading off to Nairobi, Kenya, to keynote the SAP Innovation Day on July 18th. I’ll be explaining how organizations can optimize every experience that matters using the latest technologies, as part of their Intelligent Enterprise digital transformations.
Innovation is moving at the speed of light, and our job is to show you how to move with it, how to arm your business for a brighter future with new levels of agility and resilience, and how to accelerate your journey to becoming an intelligent enterprise and uncover a smarter approach to generating value.
Professor Bitange Ndemo will deliver a keynote unpacking how the Digital Economy is a key driver of the world economy. Discover how the deep incorporation of intelligent technologies like artificial intelligence, predictive analytics, digital assistants, and conversational interfaces into applications and end-to-end processes, will change the way companies do business in the future.
Here’s a quick video I put together for the top four analytics trends I talked about at the recent VNSG (SAP Netherlands User Group) Analytics event in Utrecht’s railway museum: Embedded Analytics, Augmented Analytics, Experience Analytics, and… (as ever and always) DATA!
4 Top SAP analytics trends, from VNSG Analytics Day 2019 - YouTube
Timo's top 10 innovations- a conversation with @TimoElliott - YouTube
Here’s a summary of some of the points we covered in the video above:
1. We have a golden opportunity to make the world a better place
Technology is accelerating, and it’s amazing opportunity for win-win-win opportunities. We can improve customer outcomes, increase productivity and profits AND make the world a better place (come listen to one of my presentation in order to hear some real-life examples!). Our challenge is to ensure that the benefits are shared so that everybody is better off, not just a small global elite.
2. Weaponized algorithmic addiction
Some of the most talented, brightest people in the world are helping make the world a worse place because of bad business models. Maximizing engagement is great for ad clicks, but it’s an awful metric for society.
Casinos use analytics to find out exactly what it takes to get you back to the gaming table after losses — so that you can lose more. Social media and gaming companies are doing the same thing, at massive scale, to make their platforms as sticky — i.e. addictive — as possible.
More engaging does not mean “better”. For example, hate is a virus and we’re spreading it more efficiently than ever before thanks to modern algorithmic targeting.
Fake news is also very engaging — by definition, it’s more interesting than the mundane nuances of reality. And as Jonathan Swift remarked over 200 years ago, even if the lies are eventually debunked, the damage is already done:
“Falsehood flies, and the Truth comes limping after it; so that when Men come to be undeceiv’d, it is too late; the Jest is over, and the Tale has had its Effect…”
Throughout history, unethical people haven’t hesitated to use our human failings against us — but never before have been able to do so at such a massive scale.
Given the low cost and efficiency of modern disinformation campaigns, any group wanting to destabilize another country would be crazy not to use targeted astroturf social movements to increase societal divisions, whether it’s about immigration in the US, the Yellow Vests protests in France, Brexit in the UK, or Catalonian independence in Spain. All the research shows that we’re becoming more polarized than ever before, with technologists providing the weapons to increase the gaps.
How do we fix this? I don’t know — but I’m convinced that a big part of it is business models based on ads and clicks. It’s easier and more tempting than ever to get rich by faking authenticity, putting morals firmly to one side and lying with abandon.
4. Predictive maintenance for people
This is THE biggest outcome of modern technology: collectively, we’re going to live longer, happier lives. Wearable devices and analytics techniques that can detect subtle signs among dozens of different data sources mean personalized treatments that can head off the worst of heart attacks, mental health problems, and more. Science and technology has always helped improve healthcare, but with always-on monitoring we can find out what really works and what doesn’t, for each individual, on a massive scale.
5. AI ethics
First, it’s important to point out that it would clearly be incredibly unethical not to use technologies like machine learning and artificial intelligence, because of the massive efficiency benefits they can bring. Researchers such as McKinsey believe that up to 70% of some business processes can be automated using the latest technology, making better use of the world’s limited resources.
But it would also be unethical to ignore the potential downsides. The benefits and problems are not shared equally. For example, while machine learning will probably increase employment overall, it’s clear that it will eliminate some jobs that exist today, causing personal misery and hardship.
And algorithms are psychopaths — that is, they can behave “intelligently” and make decisions, but they have no awareness of the human consequences of those decisions. They are perfectly happy to replicate existing human biases and mistakes, making the wrong decisions faster and more efficiently than ever before.
Algorithms do what you say, not what you mean — and can be tricked. We’re used to humans making simple but consequential mistakes because of misunderstandings — just wait until you see what algorithms can do!
AI needs to remain firmly under human control, with clear responsibility for any automated decisions that are being made. Organizations need to work with ethicists and other experts like never before, and adopt the same principles as medical doctors since Hippocrates: “first, do no harm”.
6. Personal data sovereignty
Data is more important than ever. But who “owns” or controls the data that I generate as an individual? How do I stop it from being abused?
Recent legislation such as the European GDPR framework tries to help — but it’s clear that the implementations are far from perfect, and that we ultimately need a much larger approach such as “self-sovereign identity“. But can such approaches work in the real world?
In many ways, it’s about getting back to the levels of privacy we take for granted in the non-digital world. For example, I can stroll into a store and pay cash for a newspaper without the bank, the newspaper, or the vendor knowing who I am. Will I ever be able to subscribe to a newspaper online with that transaction remaining anonymous to both my bank and the content provider? Why can’t an ID service vouch that I am, say, an adult, or a citizen, or that I don’t have a criminal record, without having to also leak all of my other personal information?
7. Surveillance society
Throughout history, every new method of control and surveillance has been abused by authorities, until the abuse was discovered and eventually legislated against — from habeas corpus onward: mail, telegraph, telephones, cell phones, internet, GPS on cars, and more. This is unfortunately bound to continue — and potentially get much worse — with AI-powered technologies such as face recognition.
The ability to bring together many disparate information sources to get a single view is a huge boon to businesses — but can be a societal nightmare in the wrong hands.
Attitudes such as “I don’t care because I have nothing to hide” or “Great, it means that criminals will be caught more easily” are dangerously naive. For example, I guarantee — based on the weight of historical precedent — that there are currently senior political and business leaders around the world that are being blackmailed and coerced using evidence of crimes or embarrassing personal information that hackers or state actors have dug up on them. We just won’t find out who, and what the consequences are, for decades to come.
And today’s secrets won’t necessarily stay that way. Almost every aspect of your entire life is stored in various databases around the globe, just waiting for somebody with bad intentions to misuse it. It’s rumored that governments are sucking up large quantities of encrypted data, confident that new quantum-computer approaches will be able to crack existing factorial-based encryption methods in the next decade.
In the meantime, researchers are trying to find new approaches to protect sensitive data, such as lattice-based cryptography — but so far none of these have been proven to be safe from the discovery of new mathematical shortcuts in the future.
8. Team AI
Algorithms are powerful. Teams of algorithms will be much, much more powerful.
A lot of the recent developments in AI involve generating lots of different algorithms and then let them fight against each other to find the most effective approach. For example, the generative adversarial networks (GANs) used to create “deep fake” images use one algorithm that creates fake faces, battling against another algorithm that try to detect fake faces. The result, after many generations, is startlingly realistic images.
Today machine learning is typically used in business to automate complex, repetitive decisions in a single process, such as automated invoice matching. But most human activities involve teamwork, and teams of algorithms working together may turn out to be better at teamwork than we are.
For example, a team of (simulated) jet fighters using genetic fuzzy algorithms thoroughly outclassed human pilots, using tactics such as deliberately firing missing shots in order to induce the human pilots to take evasive maneuvers that ultimately puts them in untenable positions — literally, playing 3D chess with rockets.
And teams of AIs working with each other can lead to problems — for example, different automatic price algorithms in an industry can end up behaving like and illegal cartel — in ways that might be even harder to spot than when humans do it.
9. A new golden age for knowledge workers
We’re used to the notion that machines can help scale up physical labor. A single person with a modern tractor can work fields that previously required hundreds of people.
Over the last few decades, computers, spreadsheets, and other applications have also helped knowledge workers to be more productive. But the new generation of on-demand cloud infrastructures means that individuals now have the same computing power at their fingertips as massive organizations or governments. It means a single person can operate a multi-million dollar company, and a company with around a dozen employees can be worth billions.
It means that people with smart ideas can take them further, and benefit more from them personally. It’s the start of a new golden age for knowledge workers.
10. Disparity and diversity
The profits from technology are flowing to the segments of society that are already the best educated, richest, and most powerful, increasing the disparity between the haves and the have-nots. Big inequalities bad for us all. We’re wasting some of the best minds on the planet because they don’t have access to basic resources. We need to find new ways to spread the riches that technology brings.
Is technology good or bad? It depends! — it depends on us making the right choices as a society.
I was recently interviewed on the future of analytics by Tamara McCleary, CEO of Thulium, a social media analytics and consulting agency, for her new “Tech Unknown” podcast. You can read her blog post summary or listen to the podcast directly below.
We explored the current state of the art for AI/ML-driven analytics and talked about what business leaders can do prepare for what’s to come:
Ways AI and ML-driven analytics drive better business outcomes
The most significant business opportunities for AI
Best practices for leading a team through digital transformation
Remarkable feats that companies have accomplished through smarter analytics
Some selected quotes:
“Technology is the closest thing that we have to magic in the modern world.”
“Obviously, it’s about better business and better customer outcomes. But I strongly believe that we can also make the world a better place at the same time.”
You can see the list of other episodes in the series here, which include interviews with other analytics and data science experts such as Tammy Powlas and Kirk Borne.
Over 50 tonnes of plastic waste were generated across the European Union last year, according to a study by the World Wildlife Fund, and consumers are increasingly concerned about the potential environmental impacts vividly illustrated in programs such as the BBC’s Blue Planet.
A message from Sir David Attenborough - Blue Planet II: Episode 7 - BBC One - YouTube
Stephen Jamieson, Head of SAP Leonardo in the UK, saw an opportunity to innovate with purpose. Working with design expert David Kester, he put together a series of Plastic Challenge workshops bringing together consumers and retail industry experts to think about ways to alleviate the problem.
Fixing Real-World Environmental Problems: SAP Leonardo and the UK Plastics Challenge - YouTube
Using the Design Thinking methodology, the group came up with innovative, win-win opportunities, including rewarding consumers for better buying decisions, and arming consumers with better information about how they should be recycling waste.
At the SAP UKI & Ireland User Conference, Stephen explained his five big takeaways from the process:
Inspiration action means start with “why” rather than “what” — per Simon Sinek. Rather than starting with what the group should do, it started with a discussion of the groups’ purpose and core beliefs.
It’s important to humanize the technology. Having twenty-four representative consumers in the group, and taking the time to understand their feedback, was the “rocket fuel” of the project.
Defining a safe space for collaboration was key. The room was full of brands that are infamous competitors. But the team set clear rules of engagement, with a commitment to transparency and openness.
Iteratively prototype in order to fail early and cheaply. For example, the Reward for Change app got all the way through the first clickable prototypes, but then proved “not quite right” — but the group was able to quickly regroup and iterate new versions for testing.
Follow a good design process. The group applied best practice design thinking concepts, without jumping any of the steps.
As a result of the workshops, SAP has launched Plastics Cloud, that uses machine learning, big data, and analytics to bring together information currently stored in different silos across the supply chain. The goal is to help reduce the environmental impact of plastics waste and finding new ways for industries players to collaborate. For more information about the project, including a 10-minute documentary about how the Plastic Challenge was organized, visit the official site.
It’s a great example of how organizations can innovate at scale using the approaches behind the SAP Leonardo digital innovation system. In the modern world, doing good is good business, and organizations are increasingly realizing that being the change is a game-changer.
So what’s your purpose? How can you work with others across your industry to come up with innovative solutions to help make the world a better place?
I was honored yesterday to be one of the panelists in one of a series of SAP Radio 2019 prediction shows, hosted by Bonnie D. Graham. You can catch a recording of the session Voice America site or below.
I’m on at 21′ in — here’s a rough transcript of what I talked about:
“My prediction for 2019 is that we’re at the start of a new golden age for human intelligence.
I’ve been working in data and analytics for around 30 years. And that whole time, it’s been perfectly clear what business people want — easy, intuitive access to all the data they need to run the business.
But the technology just hasn’t been there. In the late 1990s and early 2000s, I helped launch a whole series of products — BusinessMiner, predictive analytics for the masses, Intelligent Query that allowed you to ask questions in everyday language, and even a tool for querying using voice. But it turned out that none of these were viable or usable in real-world situations.
But we’ve reached a tipping point, and the technology has finally caught up with our aspirations. These technologies are transforming every aspect of business and being adopted at phenomenal speeds. For example, Gartner just revealed that machine learning implementations tripled last year last year alone.
For analytics, it means that we can finally deliver the kinds of dashboards that business people have been dreaming about for over forty years — easy, real-time access to information about every aspect of their business, literally at their fingertips.
And just as the mechanical era allowed a single farmer to plough hundreds of acres using a single tractor, I think we’re poised to see the same sort of thing, but with algorithms augmenting human intelligence.
I truly believe that knowledge workers will be able to do vastly more, in less time, than ever before, and so it’s going to be golden age for humans, and human intelligence. “
2019 is the year that analytics technology starts delivering what users have been dreaming about for over forty years — easy, natural access to reliable business information.
1. Machine learning everywhere. We’ve reached the third great wave of analytics, after semantic-layer business intelligence platforms in the 90s and data discovery in the 2000s. Augmented analytics platforms based on cloud technology and machine learning are breaking down the longest-standing barriers to analytics success. They bring insights to users rather than forcing users to unearth elusive trends, and provide more intuitive interfaces that make it easier to get the data people need to do their jobs.
2. Embedded analytics accelerates. The historical line between operational applications and analytics continues to blur. Thanks to advances in machine learning, prescriptive “intelligent applications” have become a reality. These data-driven, self-learning business processes improve automatically over time and as people use them.
3. Cloud analytics adoption skyrockets. Cloud brings agility and faster innovation to analytics. As business applications move to the cloud, and external data becomes more important, cloud analytics becomes a natural part of enterprise architectures.
The advantages are particularly important for smaller organizations: the cloud offers affordable, on-demand access to analytical and data processing power that was previously reserved for much larger organizations with dedicated analytics teams.
However, some data will never move to the cloud — a nuanced approach is required, leveraging existing analytics investments while moving to hybrid on-premise / cloud analytics architectures.
4. Better user experience drives greater adoption. Advances in speech and text recognition mean users can finally ask business questions using everyday language. AI-assisted data discovery can automatically mine data for insights and propose appropriate views of what’s new, exceptional, or different.
Chat bots and personal assistants provide seamless access to the basic numbers used to run the business. And using real-time systems as a foundation, managers finally get dashboards with all the information they need to run every aspect of the business, in real time, at their fingertips.
5. Compliance drives true data platform adoption, supported by more flexible data management. As it has been for the last forty years, data collection, preparation, and standardization remain the most challenging aspects of analytics. The rise of compliance and privacy concerns are driving the adoption of more standardized approaches — for example, reducing the attractiveness of data discovery architectures that extract and manipulate data separately from core systems.
Real-time processing, data catalogues and new “data orchestration” systems allow organizations to retain a coherent view of data across the organization without having to physically store it in a single place.
6. Data literacy will continue to be a big problem. The biggest barrier to analytics success has never been technology. Giving somebody the best pencil in the world will not make them Picasso.
Analytics culture, skills, and organization continue to be the biggest barriers to turning information into lower costs or increased profits. Organizations must invest as much time and money in analytic skills and incentives as they do in technology.
7. There will be an increasing number of AI fails. Like any powerful technology, AI brings new dangers. Algorithms are sociopaths: they have no knowledge of what they are doing. AI brings amazing opportunities for improved productivity and augmented human intelligence. But it magnifies any existing problems with data quality and data bias and poses unprecedented challenges to privacy and ethics.
Comprehensive governance and data transparency policies are essential. Sadly, things will probably get worse before they get better — organizations must implement ethics processes, councils and external advisors before high-profile disasters hit the headlines.
8. End-to-end decision-making platforms emerge. Analytics has traditionally only paid attention to a small part of the end-to-end “data journey.” Information and insight is useless unless something actually changes in the business.
More holistic, end-to-end approaches to analytics are emerging, not only because of the combination of operations with analytics, but also with more proactive and seamless approaches to converting user analysis into concrete actions. This depends fundamentally on human judgement, consensus, and creativity, but it must be supported by better integration between analytics, traditional business planning activities, and social collaboration platforms.
9. AI and machine learning makes analytics more human. More powerful augmented analytics will eliminate a lot of the work around collecting and processing data, and identifying areas for further investigation. But ultimately people are the most important “technology” required to turn data into business improvement.
In an era where basic decisions can be increasingly automated, more strategic choices rely on uniquely human skills such as creativity, understanding of context, and leadership.
10. New experience analytics. Understanding and optimizing the customer experience is the bedrock of successful digital transformation. Traditional analytics focused on structured data flowing from operational systems. Newer analytic platforms have blended more unstructured data such as text, images, and raw sensor readings into analytic workflows.
The next step is to expand analytics to both operational data and “experience data” — the unique, subjective experiences of individuals as they interact with products, brands, and internal business processes.
Looking forward to 2020 and beyond. With ever-more devices capturing more nuanced data, with technology capabilities accelerating, and powerful machine learning still in its infancy, analytics is poised for a new golden age.
Andy Bitterer and Timo Elliott talking about the top ten analytic trends for 2019
The first #AskSAP web seminar of the year was on the topic of the top trends in analytics for 2019. It was hosted by Avery Horzewski and featured SAP Analytic Evangelist Andy Bitterer and myself chatting about our top ten favorite themes for this year in the area of business intelligence, analytics, data platforms, big data, and artificial intelligence. I hope you enjoy it as much as we did!
Experience management is the process of monitoring every interaction people experience with a company in order to discover opportunities for improvement.
For decades, SAP has been the leader in business applications that generate and leverage operational data (O-data) — hard numbers like costs, accounting, and sales — to inform their business decisions.
But while operational data gives you an idea of what has happened, it only provides limited insights into why it happened — especially when talking about the customer experience.
To get the full picture, you need to understand customers’ thoughts and emotions as they interact with your company by gathering and analyzing so-called experience data (X-data).
For example, imagine that you are in charge of a roller-coaster ride. The operational data is all the hard numbers from ride itself, including the number of riders, and speed and acceleration at each point. But this gives a very limited view of what the customers ultimately care about: the subjective experience.
An illustration of the difference between objective operational data (speed, acceleration, etc.) and subjective experience data
In the pictures above, we can see two riders who are having a very different experience. While the person on the right seems to be enjoying the whole ride, the person on the left shows a range of emotions from elation to fear. Gathering this type of data is clearly essential to optimizing the roller-coaster experience.
Experience management is about combining O-data and X-data to ensure that every business decision is based on both hard facts and soft intangibles.
It’s a key step in the move from older customer relationship management (CRM) platforms to new customer engagement platforms that optimize the full end-to-end customer experience.
Doing this successfully requires new business models and moving beyond traditional standalone CRM systems. In order to provide a fully personalized experience, organizations require seamless coordination between the “front office” systems such as business-to-business ecommerce and the “back office” systems such as logistics, finance, billing, and workforce resources.
And of course, the overall customer experience also depends on the perceived quality of products sold and employee satisfaction. The Qualtrics platform allows organizations to manage the four core experiences of business—customer, product, employee and brand—on one platform.
Whatever business you are in, your prospects can now easily find out the true value of your products and services, from your existing customers, with just a few clicks.
In an era of digital transparency, it’s clearer than ever that “experience matters.”
The combination of Qualtrics expertise and SAP’s leadership in business applications promises be a huge step forward in experience management for companies around the world.