In the world of SaaS, you need to practice what you preach. As an AI SaaS provider, we preach the value of our product day-in, day-out. But can we find value in it ourselves? We recently started building our help center, for a variety of reasons including: 1. Everyone else is doing it, so... it has to be the cool thing to do, right? 2. We’re great at supporting our customers (so they tell us), but really they expect to be able to use our solution without having to ask us for help all the time. 3. The more users find answers to their questions, the more they feel we understand them. Even if we are only explaining an unexpected error message. This in turns drives them into using our solution more. 4. Writing help content forces us to face the fact that, honestly, sometimes we just didn’t do the right thing. The harder it is to explain in a Help article, the more screwed up our UX approach was. After trying out a few solutions, we opted for Intercom’s Article, which is a simple but really quick solution to build a customer-fronting Knowledge Base. We already use Intercom for our own support and sales engagements, so it made sense to keep using this platform for our Help Centre. But whatever KB solution you pick, you still need to build your articles. And this is where the real fun begins. Our first approach was to document every single feature of our great and brilliant SaaS solution. And as we have a lot of features, we managed to generate a lot of articles. 550+ articles about 3,627 AMAZING features and 17 cups of coffee later we had a comprehensive database of every single corner of our solution. Instant success, right? And we know our product and features well, so it didn’t take us too long to write these articles. And it didn’t take too long for our customers to read them either. Because they didn’t. Our limited brains slowly came to the realisation that a Help Centre isn’t there to help you do a sales demo of your solution, it’s here to help customers, by answering their queries. So the question arose: what are they actually asking us? Most of my colleagues remember their last 40-50 intercom support conversations (personally I remember 8 or 9…). So we could have started making a list of what we remembered. Hoping we remember it correctly. But we are not that dumb (honestly, once the caffeine kicks in we can actually pass as clever). Actually one of the many reasons our customers use our AI solution is because it is supposed to help them do exactly this: understand what their customers are contacting them about. I know this for sure because it’s written on our web page. And also because they tell us. We just didn’t think of this use case as being beneficial to ourselves, simply because a few years ago, we were so small I could actually remember all my support conversations. So we went ahead. We plugged our own AI solution (Cx MOMENTS) into our Intercom instance and started analysing those customer support conversations. And it worked! Within minutes, we discovered that customers were asking us about stuff we thought was already there or that we thought was the most obvious thing in the world. And sometimes it was stuff that had nothing to do with the app itself, but how to get a web demo from us, the difference between paid plans and free trial, error messages, data updates, and so on. Connecting Intercom to Cx MOMENTS took just a few clicks (Pheww! this is what we promise on our website. So I can keep writing this article.) Importing pre-defined topics into our Cx MOMENTS dashboard was actually good fun. Our Library feature enables users to decide, in a few clicks, which of our 250+ pre-defined/automated AI topics they want to bring into their dashboard. As we did this, we realised that some of the most frequent queries were about error messages. We didn't invest much in generating eloquent and engaging error messages, and it definitely showed in our customer's queries. Our customer-facing team’s first move was to ask our dev team for a list and description of these error messages, that seemed to be randomly appearing. We didn't the They got a short answer. Very short. So they asked the question differently. They used Cx MOMENTS to list all the terms associated with “errors”, which gave them a good idea of what these specific messages were. Then the marketing team re-formulated their question to be much more actionable, referring to these specific messages, and supported by evidence from our customer conversations. The answer from Dev came back and was a much better one. We can now document these messages in our Help Centre… or get them corrected if they were not right. This is the beauty of using customer feedback to write FAQs. They provide “facts” about what people are asking and volume trends that help prioritise them. The discussion between Cx teams and Product Development is then quite different. It becomes specific, actionable and evidence-based, and much, much easier to progress to a positive conclusion. Another benefit of analysing this customer feedback is that it helps you formulate your FAQ in a way that your customers can actually find them. For example, customers don’t always ask to “reset their passwords”, sometimes they say “my account is blocked”, or “I entered my login details too many times”. By writing your FAQ Articles with the same words and formulation that you find in your customer's questions, you increase the chances that they will find the right FAQ when searching your Help Center. And this is what we did and will continue to do in Cx MOMENTS: - Find the top 20-25 topics our customers are contacting us about - List the sentences they use to formulate their questions - Start writing Articles to answer these questions - Discover and trend the next 25 topics and do the same.
We started assessing Google AutoML a good while ago, to see how it would complement our own internal ML, AI and Natural Language processing technologies. We were a little doubtful at start, but decided to give it a go, simply because Google so often does things really right. By the way, a little word of warning before you read any further: this is NOT an academic or scientific article about Machine Learning or AI. You will NOT find complicated words and concepts that expect the reader to have PhD in neural networks. If you are looking for this, you can find excellent articles in fastAI or Medium and many other places. What we are trying to do here is explain, in layman terms, how these new Machine Learning approaches, such as Google AutoML, can actually help end users achieve specific goals. In our case, our “end-user goal” for Google AutoML is to analyse large amounts of customer service interactions, in the form of Zendesk Tickets, Freshdesk tickets, Zopim Chats, or Intercom conversations, to determine and trend the reasons why customers are contacting the support team. This is an area we understand well, as all our customers are Cx managers who know they cannot rely on tags manually applied by agents, hoping to be able to run a report at the end of the week and see which customer issues are trending. Manual tags are invariably wrong, or not applied at all, and the reports are wrong. So what we are doing here is use Google AutoML to automatically classify, categorise Zendesk, Freshdesk, Zopim and Intercom interactions to determine why customers are contacting a the customer service team. Sounds simple enough? Well actually, it is. The process we went through is in 5 steps: 1. Select a large set of tickets or chats from Zendesk, Freshdesk, Zopim, Intercom. 2. Annotate them manually: what this means, in layman’s terms, is manually applying a label to each customer query and generate what we call a “training dataset”, to teach Google AutoML how to adapt its model to classify this dataset accordingly. For example, after applying the label “Order_not_received” to hundreds of sentences such as “Did not receive my order”, “Parcel never arrived”, “no delivery today”, we hope that the next time we show a sentence like “I did not get my parcel”, Google AutoML will recognise it and apply this label. Cx MOMENTS facilitates and accelerate this greatly, with our AI-assisted categorisation, which pre-filters candidate sentence for classification into a label. 3. Review the Training / Test results: Google AutoML selects about 10% of your training data as “test data”. This means it will train its model on 90% of your training dataset, and use the 10% remaining to verify that its model is right. You then get test statistics such as Recall and Precision. If you don’t know these words, “precision” means basically how accurately you classified each interaction, and “recall” means how many of these you managed to categorise. 4. Bring the model live: If you are happy with the training results, then you can bring the model live in production to automatically categorise new customer interactions. 5. Improve the model: This is a key part of Machine Learning. Once you think you have arrived, you are actually only starting! The model is never perfect to start with. And things change…. So as users start actually "using" Google AutoML classification in a production system, they will invariably find problems: this customer interaction is about something else, or that one was not classified at all. Google AutoML makes it relatively simple and very effective to apply these corrections, and constantly make the model better. All you need to do is manually tag these few examples with the right labels, and feed this back to Google AutoML as an incremental training dataset. The model will retrain, and start producing better results. We have gone through a few iterations like these, and results are very impressive. So what is the good and the bad about Google AutoML ? The good: - Google AutoML uses Transfer Learning. In layman’s terms, this means it is already trained on large amount of data, and only needs an incremental amount of your data to adapt its model to your use case. - The workflow is relatively simple to use: a UI lets you annotate manually, or load a training data via csv file. - API is well documented and once a model has been trained, makes it relatively easy to classify new sentences tickets in production. - Test results / scores are clear and well presented. The bad: - Google AutoML workflow is limited: training data should be injectable via API, not just via csv file. - The UI is limited: we actually used our own UI to annotate and generate the training dataset. - The test results & scores should be made available via an API, so they can be re-used in customer facing systems.
How often do Product teams look for customer feedback about a new product, a new feature… or about an old problem! Is our new product working? Is this old problem really solved? There are always new reasons to wonder about how such feature is perceived, or if this product is working correctly. For example, a manufacturer of audio equipment wants to know if a Bluetooth pairing function is working correctly, or if the user manual is actually included in every box. Or an eCommerce site selling tricycles with a customisable wheel wonders if the right wheel is always delivered. Or a B2B SaaS vendor wants to check if its new configuration widget is easy to use. Customer surveys would be the traditional way to collect customer feedback on what they think about your service or product. This is an essential tool, and everyone should use it at every touch point or support interaction. But there are better ways to gather and analyse customer feedback at scale, especially when it comes to the specifics of product feedback. There are 3 main challenges with customer surveys. 1. The relatively low response rate: 10% would be good, 5% to 6%. Is probably a more common response rate. 2. Customer surveys can only go out so often, and every customer will only be surveyed once or twice a year. This reduces further your scope for learning. 3. Finally, there are only so many questions you can ask in a survey. There is rarely room for, or success with, including detailed questions about the customisable wheel of a tricycle … assuming you even bought that product! This is where customer support interactions become really valuable to analyse. 1. You will often find that the volume of customer support interactions is tens to hundreds of times larger than your survey volume. And 100% of these interactions can be analysed. 2. Customers will contact your service team whenever they want, not just when you ask them. At the end, it is likely that 20% to 80% of them will do so, providing on the way valuable information that you can then analyse. This is much more than the subs-set of your customer base that you have surveyed, and of which only 4-6% will respond. 3. But most importantly, when contacting your support teams, customers will talk about what they want, and in as much details as they feel is relevant to explain their problem. This makes customer support interactions an invaluable source of insights for product teams: this green LED keeps flashing, the power won’t turn off, chicken meals are too salty, the widget won’t allow me to click, and so on. All spontaneous and real product experience. Now the questions is: how do you collect and analyse these feedback? And how do you share them with the product team? Collecting is easy: your support tickets or chat logs have done this already. Analysing is easy: Cx MOMENTS will automatically detect and trend queries, questions or complaints in your customer support interactions, and will let you add, in a few minutes, topics that can be very specific to your business. Sharing is easy: - you can export a report from Cx MOMENTS and send it to your product team - Or you can upload the same as a Trello card if your Product team uses Trello like we do. - You can invite product Managers to become users in Cx MOMENTS - Last but not least, you can automate this feedback process by deploying a solution like ProductBoard, which can automatically share tagged Zendesk tickets with the Product team. All you need to do is use Cx MOMENTS to automatically tag tickets in Zendesk when they talk about a specific product feedback. Want to get started and share customer feedback and issues with your Product team? Why not try Cx MOMENTS now?
Developed in 2003, CSAT measures customer satisfaction based on a 1 - 5 (very unsatisfied - very satisfied) scale. This metric shows how happy customers were with the overall support process. It also helps identify pain points to uncover areas of the process that could be improved. Any worthwhile customer support team will use CSAT, even if only as one factor, as a benchmark for customer satisfaction. Your satisfaction rating is the average customer satisfaction rating given during a reporting period. It helps to give you an overall view of how well your support team is meeting your customers’ needs and is an easy metric to benchmark yourself with against your peers. Zendesk have done a great benchmark analysis available based on their thousands of customers, split into industries and company size for you to see how well you compare to the rest. Your satisfaction rating is hugely important to your business, because if your customer aren’t happy, they won’t be coming back and they won’t be telling their friends about all the great things you do. But an average across all your tickets very often hides important information. Cx MOMENTS can show customer satisfaction in much more detail, for instance drilling down to see customer satisfaction by the reason to contact your support. Now we can see that the bulk of customers are satisfied, but customers contacting us about high priced upgrades, who make up a small but important group, are not. We now have a specific insight into what might be causing a decrease in customer satisfaction and we can fix it at its route cause. A big strength of customer satisfaction score lies in its implicitly: it’s an easy way to close the loop on a customer interaction and determine whether or not it was effective in producing customer happiness. If for some reason the experience wasn't satisfactory, its easy to pin point that moment and take actions to remedy the experience. With the extra tier of data you receive from Cx MOMENTS this allows you to spot specific customer issues or pain points which can easily be resolved to quickly improve your customer’s experience. Not only that, but you can track customer satisfaction across the customer lifecycle very simply using customer satisfaction scores. Since its such a quick survey you can ask it across multiple experiences during the customers journey and get a big picture overview of how your customer feels at each touchpoint during the process. This makes it easier to find potential bottlenecks that could arise in your customer experience allowing you to work on these. With Cx MOMENTS, we use AI to analyse your customer support tickets and provide you with a detailed and highly accurate view of why your customers are calling your support. We automatically tag all of the tickets in your Helpdesk based on the customers reason to contact providing you with a detailed analysis, for the first time of what issues your customers are coming in contact with. This data gives a new depth and insight to the one dimensional metric of customer satisfaction scores. Before, customer satisfaction scores could only provide a general overview of how your support was performing and very often important information can hide underneath broad metrics like this. Our AI analytics provides you with a breakdown of customer satisfaction based on the customer issue being discussed. You may have a high overall satisfaction score and be struggling to find areas to improve, but with an AI plugin in your Helpdesk, you can now see which specific customer issues may be causing lower satisfaction scores than others. This information is invaluable to improving your support and moving the coveted CSAT ever closer to the mystic 100. Issues with low satisfaction scores can be investigated further to find the underlying root cause by looking into the correlations between issues, products, agents or partners there may be. Customer issues related to shipping by be seen to have a lower CSAT score compared to other issues your team deals with. With one of our online retail customers we were able to find some interesting insights when looking further into the data using Cx MOMENTS. We were able to see that users who mentioned shipping issues also mentioned UPS at a disproportionate level. The support team was able to bring up this issue with management and the logistics team with detailed data from thousands of tickets related to this issue. Management was then able to follow up with the shipping partner and remedy the situation. Customer satisfaction scores can also be broken down by agent and issue. This means that you can see which agents may be struggling to answer specific customer issues. Certain agents may struggle when dealing with more technical customer issues related to the product. With Cx MOMENTS you can quickly spot these patterns in lower CSAT scores for specific issues and either move this agent from the agent group dealing with these issues or create personalised training for your agent to quickly bring them back up to speed. The possibilities are endless once you have the data to see what issues are causing lower than average customer satisfaction scores. With this information you can make improvements in your support processes, bring all agents up to speed and escalate recurring issues to other departments for permanent fix. Cx MOMENTS helps you to uncover these hidden issues beneath the broad metric of CSAT helping you to provide ever better customer support.
1. Track and trend your customer support queries. Properly. This is where it starts. Why are your customers calling support? How often do they complain about a specific software issue? How frequently do they ask about their order delivery? Which products generate the most support calls? As a Cx Manager, if you don’t know this, you are blind. You can’t improve self-service. You can’t train your agents. You can’t escalate problems to other teams for permanent fix. You can’t do much at all to improve customer experience. But how do I do this …“properly”? might you ask. Traditionally, you rely on support agents manually applying tags or custom fields to each support interaction. And unfortunately this does not work well. Manual tagging is simply not reliable, and we see it every time we run a report to trend a specific customer issue. It never shows what we expect, because manual tagging is not reliable. At Cx MOMENTS, our own industry stats show that only 15% of the reasons to contact are properly identified by manual tagging. But Cx MOMENTS will do this automatically for you. It will detect and trend the reasons why your customers contact your support teams. And when you discover a new issue or reason to contact, Cx MOMENTS will classify your entire ticket history against it. So you’ll always know if a new problem is really …new. 2. Share new product feedback with … your Product team. When launching a new product or a new feature, Product Managers are really eager to hear what comes back from the field. And quick. Early customer complaints or queries will help detect problems or design weakness before they spread to a wider customer base. There again, agent tagging is not reliable enough to filter the right tickets or chat transcripts. And Cx Managers are left with the task of manually reviewing tickets to spot the ones they should share with product teams. Cx MOMENTS will do this automatically for you. Even with few or no customer queries at start, as you would expect with a new product, you can configure Cx MOMENTS to track your new products as soon as they launch. 3. Stop running after recurring support issues. Escalate them to the team that should fix them for good. There is a phenomenal amount of literature about how to better listen to customers when they contact your support team, how to measure their satisfaction, how to improve the efficiency of your support process in terms of call handling time, time to close and so on. But when it comes to fixing things so that customers don’t have to call in the first place, the literature becomes thinner. This burden is left on the shoulders of the Cx Manager. When a product or process issue generates a large amount of recurring support queries, they have to figure out how to best escalate it to the team in charge of fixing it for good. That team is generally also very busy, and will not bulge unless proper metrics and evidence is there. Because agent tagging is not reliable, Cx Managers end up spending weekends reading tickets to make up these metrics, and find “some” evidence, or examples should we say. Cx MOMENTS will do this automatically for you. In a few click, it will show the trend of a specific customer support query and filter the associated tickets, so you can share this with the other team, who will look very differently at a trend and list of tickets you can generate in 2 clicks. 4. Keep the difficult customer queries for experienced agents, and assign the easy ones to new agents. Every successful company faces the same challenge. When sales grow, so does the support activity. And you never get new support agents hired and trained on time to support this. So why not be pragmatic about it? You know your business, you known which customer queries need experienced agent to be resolved, perhaps because they are complex to troubleshoot through your different business processes, or because they require a strong personal relationship in place with other stakeholders to be sorted out quickly. Put simply, make sure the easy queries get routed to your new, inexperienced agents. And keep the difficult ones for the experienced staff. Cx MOMENTS will do this automatically for you. Our AI classification can be used by your Helpdesk solution to generate workflow automation, including assigning tickets about specific issues to specific agent groups. So why not try Cx MOMENTS today? We integrate on one click with most Helpdesk and Chat solutions , and offer a 2-week free trial.
Everyone starts somewhere and in customer support a baptism of fire and being thrown in the deep end can often end in tears. 33% of Americans say they’ll consider switching companies after just a single instance of poor service and U.S. companies lose more than $62 billion annually due to poor customer service. Some tickets are far more complex than others to solve and if it’s your first few days as an agent it would be much easier for you and the customers if you could be fed easier tickets first and slowly work your way up to tickets which require more experience and expertise. Today your agents either cherry pick the tickets they want to answer first or get assigned the next tickets waiting in line. You may request customers to self-diagnose their issue or the agent group they should be put in contact with before they submit a ticket, but very often customers don't know what help they need or just randomly click in order to get through to a person as quickly as possible. Following this first contact the ticket may then need to be routed to the right agent for the customer issue. If you have newly started agent, you may not want this same system to apply for them. Training agents up on easier tickets can help to bring them up to speed quickly without sacrificing customer satisfaction. And it's not just about training, it's also about maintaining excellence in Customer Service at all time. Most B2C companies need to quickly hire a large number of customer service agents during each seasonal peak period. These temporary agents are not all familiar with your products or business processes, and you might find it helpful to assign them the "easy tickets", while keeping the "difficult" ones for your full time experienced staff. The challenge, as always, is how to assign these without reading each ticket. Read Tickets Without Opening Them Using Natural Language Processing and Machine learning, our AI is able to understand unstructured text and tag your support tickets with your customers reason to contact your support along with any other keywords with in the ticket. This happens as tickets are coming into your Helpdesk and allows you to already have tickets categorized, before you begin solving the customers issue. This is a huge development for support, as up until now you have been in the dark with regards to what your ticket consists of until you actually read it. Turning your qualitative customer conversations into quantitative data is the first step in taking advantage of your support data and using it to improve your support productivity and your customer experience. Creating Rules in Your Helpdesk to Route Tickets Cx MOMENTS can also apply these topics as tags into Freshdesk or Zendesk Support tickets. This enables you to configure observer rules or triggers in the Helpdesk solution, to automate (for example) the assignment of specific tickets to specific agent groups, or automatically prioritize certain types of tickets in a backlog queue. This feature is now available within the Cx MOMENTS dashboard, allowing you to create custom tags based on our automatically detected topics in your support data. By automatically routing tickets based on the customers issue, high priority tickets can be handled by your best agents, and easier tickets like customers looking for a delivery update or to change their password can be passed on to your new agents to get them into the swing of things and help them find their feet. This can be done by automatically assigning particular customer issues to certain agent groups. As agents become more comfortable in answering customer problems you can move them into other agent groups where they can deal with progressively harder customer issues. This is a great way to maintain high customer satisfaction scores and prevent new agents disrupting the customers experience. Monitoring your New Agents Performance But how do you know when an agent has mastered a customer issue and is ready to move on to more complex or specialised customer problems or queries? Cx MOMENTS provides you with detailed analytics on your customer support tickets. We provide new depth to traditional metrics you may be using in your Helpdesk today like CSAT and agent first response time. We provide you with a detailed and highly accurate view of your customer issues and how they are trending over time, and we then splice this data with your traditional metrics within Cx MOMENTS. This allows you to see important customer support metrics like customer satisfaction or first response time broken down by each agent and each customer issue. Cx MOMENTS provides a customer satisfaction score for each of your agents based on each customer issue they handle. This means you can view their performance on each new customer issue. If they are struggling to maintain a high customer satisfaction score when dealing with customer issues like lost deliveries, you may want to either route these customer issues away from this particular agent or create individualised training for this agent on this customer issue to bring them back up to par. Final Thoughts Knowing what a ticket consists of before you even open it provides you with a world of possibilities to start improving your support. Training up new agents is a necessary part of a support agent's job, but it can leave you open to the risk of possibly providing below par support. And this can be more harmful to your brand then just one poor experience. Americans tell an average of 15 people about a poor service experience, versus the 11 people they’ll tell about a good experience. AI can help you to train up your staff on real customer issues while mitigating this risk by choosing the tickets they handle. As agents mature in their ability to handle issues, and in their knowledge of your business, you can start routing more complex tickets to them and closely monitor their performance on these topics from within your Cx MOMENTS dashboard.
As technology advances we are met with more and more possibilities to improve support productivity and better help our customers. Artificial intelligence has become one of these tools which is now at the disposal of customer support teams. There is a host of different challenges and processes which you can use AI for to augment your support capabilities. The first approach is to use reputable Bot Platforms such as Google DialogFlow, IBM Watson Assistant, Amazon Lex, or Microsoft Bot Framework to automate responses to recurring, simple queries. These solutions are great when it comes to serving your KB articles, but they face 2 challenges. Customers sometimes ask for more than just a simple pre-canned answer, and we will see further down in this blog post how to extend these front-end automations to real back-end Business Process Automations. Bots learn from their mistakes, but this takes time. If only you could train them with much more data at start, wouldn't it be great? Well, this is exactly what we are going to explain next. The big thing that Artificial Intelligence and Machine Learning need is data. Lucky for you, support teams have a lot of data. They receive thousands of support tickets every week. Turning your qualitative customer conversations into quantitative data is the first step in taking advantage of your support data and using it to improve your support productivity and build out entire automated processes to help your customers even faster. Customers come to you with questions and Artificial Intelligence can help provide the right answers as quickly as possible. To assist you in the entire automation process, Cx MOMENTS have broken this process into 4 simple steps explaining how you can build out support process automations. 1. Select the Customer Issues or Reasons to Contact Knowing your customers top reasons to contact and how these are trending over time is vital to select which processes and customer issues to try and automate. Having data on the number of monthly tickets, customer satisfaction and time spent by your team on the customer issue is also important in getting buy-in from management to undergo such a project. With Cx MOMENTS the history of your tickets are analysed with AI, providing you with a detailed and highly accurate view of why your customers are calling. From here you can find an issue which may be taking up a significant portion of your teams time, which may be a good candidate for process automation. 2. Find and Use Real Customer Queries to Train the Bot's AI In order to build an automated process or chatbot, such as Google DialogFlow, IBM Watson Assistant, Amazon Lex, or Microsoft Bot Framework, you need to implement what is known as supervised machine learning. Supervised learning is so named because the data scientist (or customer support team!) acts as a guide to teach the algorithm what conclusions it should come up with. It’s similar to the way a child might learn to count from a teacher. Supervised learning requires that the algorithm’s possible outputs are already known and that the data used to train the algorithm is already labeled with correct answers. Supervised machine learning algorithms uncover insights, patterns, and relationships from a labeled training dataset – that is, a dataset that already contains a known value for the target variable for each record. Because you provide the machine learning algorithm with the correct answers for a problem during training, it is able to “learn” how the rest of the features relate to the target, enabling you to uncover insights and make predictions about future outcomes based on historical data. For example, a classification algorithm will learn to identify animals after being trained on a dataset of images that are properly labeled with the species of the animal and some identifying characteristics This can be an arduous task, often left to teams of data scientists, but with Cx MOMENTS we have taken much of the work of building a labelled dataset out. Our AI provides you with your customers reasons to contact and every ticket related to this intent. This means we have already generated the training dataset for your automation bot, whether it is Google DialogFlow, IBM Watson Assistant, Amazon Lex, or Microsoft Bot Framework. Cx MOMENTS ML Training User Interface enables agents to rapidly tag customer conversations to train the Bot AI. It's actually as simple as this: And as Cx MOMENTS Analytics has already pre-built a long list of support tickets that best match the label(s) you are training on, users simply confirm if each conversation matches the label or not, and the next one comes up instantly. In the space of 5 to 10 minutes you can have a verified dataset of 200 tickets to train up your automation bot. This process can simplify and speed up the process by 2-5 times and saves you reading through countless tickets to build your training dataset. Cx MOMENTS then makes this training dataset directly available to the top Bot platform of your choice, such as Google DialogFlow, IBM Watson Assistant, Amazon Lex, or Microsoft Bot Framework. You can use these to build your conversational workflows once they are trained on recognizing the customers' intents. 4. Automate Support Processes beyond Helpdesk Now that your Bot AI has been trained to understand the customer queries, another challenge arises: very often, customers do not only ask for pre-canned answers taken from your KB (Knowledge Base). They sometimes need "things to be done", or "checked" well beyond your Helpdesk's internal information repositories. For example, checking an order status could get 2 answers: - the easy one is to send back a KB article explaining how to find it yourself. We've seen with many of our deployments that actually most of these queries come from customers that have already tried this. - the proper answer is... get this answer! Verify the order, get the tacking order, select the shipper, request delivery status and serve the answer back to the customer, or escalate to an agent if an order or delivery exception arises. For this, Cx MOMENTS also integrates with major eCommerce platforms, such as Shopfiy, Salesforce, Magento or Opencart, helping you to self-help your customer even more. For example, Cx MOMENTS can assist you in automating the common customer query of order status. Customers contact your support team asking where their order is and today, your team needs to answer this as quickly as possible. They need to obtain the customers order number, if it has not yet been provided, and then need to check the order status with the ecommerce platform you may be using. Finally they relay this information back to the customers. It's a simple but repetitive task which takes up agents time that could be better spent solving harder customer issues. Cx MOMENTS can automate this entire process. Cx MOMENTS will integrate with your Helpdesk to read customer tickets as they come into your support team. We then integrate with Google Dialog flow which handles conversational flows based on the intent identified, and grabs the customers order number. If the order number is not present, an automated response can be fired asking the customer for this information. Order status is then automatically checked with Shopify or any other integrated ecommerce platform and the information is sent straight back to the user. Your customer gets the answer they are looking for almost instantly and your team didn't have to lift a finger. This process was set up with very little effort and can save your team countless hours fielding commonly asked and simple tasks. There are countless recurring issues and commonly asked questions that you can start leaving to machines. AI is perfectly suited to repetitive, menial tasks, leaving your support team more time to focus on the meaty stuff which will always exist. Aspect’s 2017 survey on the agent perception of chatbots found that 79% of agents feel that handling more complex customer issues improves their skills and offers more opportunities for career growth. Conclusion Cx MOMENTS assists in the integration, decision tree, training and analytics for your chatbot implementation making it even easier for you to answer your customers problems automatically. Using cx MOMENTS ML training UI (user interface) you can quickly build a training dataset for your chatbot, whether it is Google DialogFlow, IBM Watson Assistant, Amazon Lex, or Microsoft Bot Framework, train it on your support data and deploy it all by yourself. Using Cx MOMENTS multi Helpdesk, Bot Framework and eCommerce platform integrations, you can rapidly deploy automations using your existing platforms.
The holiday season with all its joys can be a nightmare for a support team. There is an influx of orders coming up to Christmas and customers can be very keen to make sure their purchases arrive on time. This means your support team is left with the difficult job of keeping this spike in customer inquiries smooth and comfortable for every customer, despite the ever-increasing inflow of tickets on top of tickets. But there’s only so many of you to field all of these issues and questions, and there’s only so much time in the day. What can your team do to handle this spike in activity? How do you keep the ship sailing on calm waters without turning into the storm and ruining the experience for your customers just before the most important holiday of the year? Customer expectations are higher than normal at this time of year and AI (Artificial Intelligence) can help you to exceed them despite the spike. Using AI to Make Your Life Easier Customer support teams can use AI to improve the service they provide today and make life just a little bit easier for everyone along the way. Using Natural Language Processing and Machine Learning techniques, it is now possible to easily automate many of the processes you currently have to do manually. Here’s how AI plugins like Cx MOMENTS for your customer support Helpdesk gives your team the secret weapon to manage the spike in support tickets your team is fielding. Automated Replies Let’s face it. Many of the questions and issues your customers bring up with your support team at this time of year are common ones that are often asked. You may have even created some FAQ content on them but still, customers get stuck and need a helping hand. And though you may have canned responses to these easy issues like estimated delivery times or your returns policy, it still takes time to read the ticket and decide on the appropriate response. With an AI solution plugged into your Helpdesk, it doesn’t have to be this way. Using AI algorithms, tickets are tagged based on what the customer means rather than the exact words they use. AI is now able to mathematically understand the context of a sentence and see what alternative words are frequently used as well. That means someone “requesting a refund” and someone demanding “their hard-earned cash back” will both get tagged as refund requests. These tags on all tickets can then be used in smart ways to help you automate all of those commonly asked questions. By using simple triggers or rules in your Helpdesk which can be configured based on the tags that have been applied, you’ll never again have to explain how many days delivery usually takes or how long customers have to return goods. It will all be done automatically for you and your team, leaving more time in the day to focus on the tough stuff. Ticket Routing We are all better at some things compared to others. It’s the way of life and it’s the same in a support team. Some of your agents are just better at certain questions than others due to experience, extra training or maybe they’re just damn good at dealing with a customer’s missing delivery issues. How much time would it save your team if you could route all tickets to the right agent or group based on skills or experience? “A lot” is the answer to that question. With the help of AI that automatically tags all your customer conversations and Helpdesk triggers or rules, this can be easily achieved. AI can now understand what a customer means based on the sentence used and how semantically similar it is to other words used in the same context (confusing? Like cash, money, dough, and cheese). Automatic triaging to the right team can now be extremely accurate no matter what words your customer uses to complain. That means customers complaining about their new shirt being “damaged”, “ripped up”, “creased”, “tearing” or “torn” can all be quickly routed to the team that handles damaged goods. Backlog Prioritization In an ideal world, your support team would have a very small backlog. The number of new tickets opened would roughly be equal to the number of existing tickets you close, with the majority of tickets getting resolved within the normal response time. But the reality is ticket volume can spike or grow too fast especially around this time of year. Backlog can sometimes grow into an unclimbable mountain of new and opened tickets that we’ve all faced. With this ever-increasing backlog of tickets and a shortage of time on your support team’s hands, it becomes extremely difficult to know which customer issues require prioritization. Unless you open and read every ticket, there is no way of knowing what the ticket is about and which ones to deal with first and reading those tickets can take up a serious amount of your teams time. AI can once again come to the rescue by automatically finding and tagging your Helpdesk tickets. Cx MOMENTS provides you with a dashboard showing you your customers top reasons to contact. From here you can make decisions on what issues require priority and assign these to your agents first. Issues like lost deliveries or faulty products may need to be resolved much quicker than customers looking to reset their password. By applying a few smart settings and features within your Helpdesk account, you can now fight the biggest fires first before they get out of hand. Conclusion The customer support team is vital to providing an amazing customer experience and can make a customer loyal when an interaction goes well. But during peak seasons like Christmas, teams can quickly find themselves drowning in support tickets and fighting just to try and open them all. An AI solution and some smart Helpdesk features and settings can help make the life of a support agent so much easier. AI helps to automate those easy-to-answer responses, route tickets to the best-suited team and prioritize the most important queries when the ticket count starts piling up.
Proactive support and engagement can help build customer trust and long-term relationships, but what exactly do they mean? Proactive support means getting in front of a customer issue before it escalates or even happens. Proactive engagement means identifying ways in which a customer experience can be enhanced without the customer asking for it or even knowing that it’s possible. For example, a text letting a customer know that her flight has been delayed, an email informing a customer their issue or requested feature is now available, or a phone call checking to ensure that the shipping address on file is correct. Proactive support and engagement go above and beyond standard reactive support. Reactive support means waiting for a customer to reach out with an issue—but by that time, a customer is likely already frustrated with your brand. Waiting for a customer to reach out with a problem is like waiting for your houseplants to start wilting before you water them. How AI can Help Today, artificial intelligence can assist your team in providing proactive support automatically. Cx MOMENTS uses natural language processing and Machine Learning to automatically tag all of the tickets in your Helpdesk based on why a customer is contacting your support. From connecting your Helpdesk to Cx MOMENTS in a few clicks, our AI analyses and tags all of the tickets in the history of your Helpdesk based on a range of keywords and reasons to contact. Providing you with a detailed and highly accurate support dataset. Turning your qualitative customer conversations into quantitative data is the first step in taking advantage of your support data and using it to improve your support productivity and your customer experience. From here you can start using tags in your business rules (automations, macros, and triggers) to segment your customers, build custom workflows and proactively help your customers. The Benefits of Proactive Engagement Increased customer loyalty
When you engage customers, you have a better chance of retaining them. In a brick-and-mortar store, a salesperson makes a sale by connecting with customers individually—they use everything they know, including information about the customer, the types of products they're interested in, and their own sales experience. In the age of e-commerce, proactive engagement lets salespeople build a personal connection with customers, creating loyalty online. Increased CSAT
92% of chats receive a positive CSAT score. We like attention—it’s human nature. That’s why reaching out to customers, for example via live-chat, is generally well-received. And when customers are happy, they’re more likely to make a purchase with your brand. Increased sales
With research indicating that 55% of online shoppers will abandon a purchase if they can’t quickly find an answer to a question, proactive support and engagement have become necessary. If a customer is hesitating on your checkout page, a simple check-in could solve a problem they might have, or a question they might need answered. According to Forrester, reaching out in real-time and offering proactive support can increase sales by up to 29%. You can also use your tagged data to segment customers calling into your support. Create micro segments based on highly specific issues customers may have had like users complaining about high salt content in their food, or a faulty light on their new drone. These customer issues will be automatically tagged as they come into your Helpdesk. Using these tags you can then build proactive outreach campaigns targeted specifically at these users to notify them of any improvements or offer them personalised offers to stimulate winback and prevent churning. This is an interesting way of using your tagged data to build automated outreach campaigns to add a personalised touch to your support and keep upset customers from churning. Freshly Builds Proactive campaigns with Cx MOMENTS Freshly’s team uses the data provided by Cx MOMENTS to advance their customer retention and winback outreach efforts by treating each customer as an individual with individual concerns. The granular analysis of all customer conversations enables Freshly to initiate well-targeted and proactive outbound engagements via Zendesk’s product “Connect.” The tight integration between Cx MOMENTS and Zendesk creates target customer segments that can then be used by Zendesk Connect outreach campaigns to follow up with specific users based on their specific concerns. Cx MOMENTS helps the team solve future problems as well. With the detailed insights into target customers, the web or marketing teams can gather feedback from customers who have had issues or comments in their area in the past and work on improvements and new products. Freshly is now on the road to providing some of the most proactive and individualised customer service around. In a world of increasing customer service expectations from consumers, this will surely act as a strong differentiator and help to retain and win over customers as they grow—and Cx MOMENTS is happy to be a partner in that journey.
One key reason for a Customer Service Manager to look at individual agents performance is to see what can be done to improve overall customer satisfaction and keep customers happy. Agent interaction is an important part of customer’s perception of your company and its products, so a good place to start is to naturally look at and assess agents performance, and how they score in terms of customer satisfaction. But how? For each support ticket closed, a survey is generally sent out asking the customer to rate their interaction with this specific agent. When enough surveys come back from customers, then a satisfaction rating can be applied to each agent. Bad agents need training, good agents need rewards. Et voila. Sounds easy? Well, not so fast… There is certainly some truth to be found in this data, but what if some agents had to handle more difficult issues than others did, such as order not received, customer overcharged, refund never paid, and so on? Obviously these agent are bound to get much worse satisfaction ratings and performance than other agents that just had to confirm the opening hours of a specific store, or whether a product is back in stock. Now: To be fair, and real, a Customer service Manager has to look at the diversity of issues that are handled across its agents group, and try and compare his agents performance against the same issues. Again, this sounds easy. The Customer Service Manager simply decides to run a report across agents for one specific ticket category, generally identified by a tag or a custom field. But then a new challenge appears: unfortunately, even if the reports looks good, sampling a few tickets quickly shows that the ticket tags or fields are not correctly applied by your customer service agents. In short the tickets are not correctly categorised, and the report is worthless. And this will probably always be the case , whatever new categorisation or tagging system you put in place. The next step is then for the Service Manager to read tickets manually over the week-end hoping to come up with enough similar tickets for each agent to compare them and detect which ones need coaching. That might happen one week-end, but that’s it. It won’t happen again. But do not despair….there is a simpler and much more effective way to do this: use AI to read and classify your tickets. Cx MOMENTS does this automatically, for new and past tickets. We detect what topics, issues, queries, problems were raised by your customers in each of their support tickets. We then map these topics against tickets volumes, trends and… customer satisfaction. Once you have that, the next steps are sooooo easy! First, focus on an important and meaningful topic. Depending on your specific business challenges, this could be one of most recurring issues (e.g. product not being delivered), one of the fastest trending ones (e.g. Poor performing new app release), or one with poor satisfaction rating (e.g. refund not processed). Whatever your criteria is, select the topic you are interest in by clicking on it. This is in effect a one-click filter to compare how your agents perform against this specific topic: And if it’s time for a one-2-one coaching discussion, another click will give you all the tickets this specific agent has processed for this specific topic. With Cx MOMENTS, performance assessment, QA and coaching of customer service agents can focus on the precise topics and issues that specific agents need help on. And it does this in just a few clicks. Try it here!