<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Dr Andy Lewis on Medium]]></title>
        <description><![CDATA[Stories by Dr Andy Lewis on Medium]]></description>
        <link>https://medium.com/@Decisionista?source=rss-2d07e0a8d645------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*f7-Z69uK_shcVUkcbQPTGQ.jpeg</url>
            <title>Stories by Dr Andy Lewis on Medium</title>
            <link>https://medium.com/@Decisionista?source=rss-2d07e0a8d645------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 00:48:37 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@Decisionista/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Three ways the value of data is destroyed within your CDP]]></title>
            <link>https://medium.com/@Decisionista/three-ways-the-value-of-data-is-destroyed-within-your-cdp-b896a10eec68?source=rss-2d07e0a8d645------2</link>
            <guid isPermaLink="false">https://medium.com/p/b896a10eec68</guid>
            <category><![CDATA[digital-marketing]]></category>
            <category><![CDATA[real-time-personalization]]></category>
            <category><![CDATA[real-time-marketing]]></category>
            <category><![CDATA[customer-experience]]></category>
            <category><![CDATA[real-time-analytics]]></category>
            <dc:creator><![CDATA[Dr Andy Lewis]]></dc:creator>
            <pubDate>Thu, 01 Sep 2022 11:14:44 GMT</pubDate>
            <atom:updated>2022-09-01T11:14:44.619Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jJrraZQRUniLU4bdOOAf8A.png" /></figure><p>There are more definitions of what a CDP is than there are people who will read this blog post. And no two vendors will agree on what the most important functionality is for a CDP.</p><p>At heart though, a CDP is the latest attempt to reach the nirvana of a Single Customer View where All Business Problems Will Be Solved. What is genuinely good about the latest focus on this task is the inclusion of several new types of functionality that help create a dynamic real-time customer profile, including good identity matching, real-time and streaming data analytics, and sophisticated machine learning.</p><blockquote><em>A CDP creates the potential then to listen to many real-time sources of data about your customers, apply sophisticated processing and then make that valuable insight available to many business systems and customer channels.</em></blockquote><p>However, the value of the data in your CDP is only as good as the way you make use of that data. Getting the most value from customer data requires ditching old-fashioned ideas about data processing that stubbornly linger in organisations.</p><p>So, what are common ways data value is destroyed?</p><h3>#1 Segmenting your Customers</h3><p>Segmentation of customers into groups with similar important characteristics is an artefact of a time when we had little data on our customers and rarely had the computing power to treat customers as individuals. With modern big data stores and cloud environments, that is not true. There is no fundamental reason why you cannot assess your customers one-by-one as individuals with a rich, vast and dynamic data profile around them, and construct messages and actions for that customer that are unique to that individual.</p><p>Indeed, that is why you have invested in a CDP, is it not? You want to make the best possible decisions about each customer?</p><p>So, why do you take your customers and put them in big buckets — segments — and treat all the members of that segment as if they are the same? Your optimised strategy for that segment is only optimal for the average customer in that segment. You will be missing opportunities and value with all the customers that are not like the average one, as you are not resolving that customer with the resolution that can yield significant differential insight.</p><p>Segmentation smooths out all the insight you might glean from your customer profile. If you put a customer in more than one segment then you will have difficulty in understanding for that customer which is the most important segment or journey they are on, and which is the best action to take for that customer. You are prioritising segments, and the associated campaigns, rather than individual and personalised actions for customers.</p><p>Using your customer data to focus on one customer at a time, and applying Next Best Action approaches, will ensure you maximise the value of your data to pick the optimum set of actions to take with that customer.</p><blockquote><em>Maximise data value by always resolving your customer as an individual and calculate your personalised next best action for that individual.</em></blockquote><h3>#2 Batching your analysis</h3><p>Real-time CDPs now exist that allow you to apply data analytics to inflight real-time data. Your data is at its most valuable the moment you capture it. Every second that then elapses will result in the decay of value in that data. The rate at which the value decays will depend on the type of data.</p><p>The value of capturing a year of birth will decay very little over time. However, the value of understanding what page on your web site a customer is looking at may well decay in seconds or minutes. Maybe your customer is stuck on filling in a form, or is looking at the contract cancelation terms. Every second you delay means you have a smaller window of opportunity to act. Delay long enough and the window shuts.</p><p>It is the fast decaying data that can be the rocket fuel for decision making. A date of birth gives little insight into a customer’s intent. What transaction they last made, or web page they visited, probably does.</p><p>Delay by seconds or minutes and you can no longer influence the current web session to nudge the customer towards a valuable goal. Delay by hours and your customer will be researching competitors. Delay by days and the customer will have bought elsewhere or cancelled their contract.</p><p>Understanding your current customer context (where they are, what transactions they have just made, what channel they have been on, what service issue they have etc) will always lead to insight into their intent and will allow you to predict what the best action to take is. This will require the execution of many rules and predictive models to optimise your actions. If you wait for a batch run to compute much of this, the value of the data will decay.</p><p>Any latency in your CDP and decision making will make your actions less relevant to the current customer context. Your rule evaluations may be no longer valid. Your predictive model may be using out-of-date predictor values.</p><p>Data value can be maximised by ensuring the analysis you do with your data is done the moment you capture a new significant piece of data or see a customer state change. Being able to act on that data change in the moment ensures you can act within all the windows of opportunity you have available to you.</p><blockquote><em>Maximise data value by removing all latency in data analysis, insight generation, and activation of actions. This maximises contextual relevancy for your actions.</em></blockquote><h3>#3 Distributing decision making</h3><p>A core capability of a CDP is to make available its data and insight to applications and channels that can make use of it. A weakness here is that each channel and application may use that data in inconsistent and incoherent ways. For example, a web channel may use the insight that a customer is near the end of contract to put up a banner for a 10% discount renewal offer. In contrast, the email campaign tool may decide a free gift offer is the best approach for this customer.</p><p>The inconsistency comes from different decision making strategies existing in each activated channel. Rules may be different. Definitions of offers and campaigns may differ. Treatment selection strategies may be different. Despite important insight being gained on the CDP, how that insight is used may vary wildly and incoherently.</p><p>A centralised decision authority mitigates this problem. This may be an integral part of the CDP in that complex decision strategies may be run at the time of data capture to decide on the best action to take, what time to take the action, what channel(s) to use/activate, what calls to action to make, what visual templates and content to use.</p><p>If you make the decision centrally, and then use channel applications for delivery only and not decision making, you can ensure the customer experience is always coherent and coordinated. You can orchestrate channels seamlessly in guiding customers along their own personalised journeys. You can ensure your contact history is consistent and being used to nudge customers towards business goals maximally. You can decide what channels are most likely to succeed and only go for the expense of paid channels when that adds sufficient incremental lift and a positive business case for doing so.</p><blockquote><em>A single decision authority ensures your data is being used consistently. Value is not destroyed by clashes and incoherence.</em></blockquote><h3>Finally, focus on the customer’s experience and the decisions you make about them</h3><p>It may be strange to have to say this, but using a CDP be should always be about the customer, their experience and the decisions you want to make about them with their data to optimise their experience. The use cases you want to enable with a CDP will drive the types of decisions you need to make. And the decisions you need to make will drive what data is valuable in that decision. This insight allows us to see that not all customer data is equally valuable and that the CDP, if it is to add value, should be ensuring valuable data is being exploited and not just creating a single customer view per se. Focussing just on data will ensure you collect a lot of data that has little value. Focussing on decisions puts the spotlight on how that data drives business value.</p><p>And that business value is maximised when we make decisions about one customer at a time, understanding their current real-time context, intents and needs, and can ensure that this complex decision is consistent across any point of use and channel.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b896a10eec68" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Make 2019 “The Year of Predictions”]]></title>
            <link>https://medium.com/@Decisionista/make-2019-the-year-of-predictions-c0054fc70e3d?source=rss-2d07e0a8d645------2</link>
            <guid isPermaLink="false">https://medium.com/p/c0054fc70e3d</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[customer-experience]]></category>
            <category><![CDATA[predictive-modeling]]></category>
            <dc:creator><![CDATA[Dr Andy Lewis]]></dc:creator>
            <pubDate>Fri, 08 Mar 2019 14:19:43 GMT</pubDate>
            <atom:updated>2019-03-08T14:19:43.927Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*JC4rP2j_2BAfl66CMlCrDQ.jpeg" /></figure><p><em>This article was first published on </em><a href="https://www.pega.com/insights/articles/to-market-better-make-2019-the-year-of-predictions?utm_source=tw&amp;utm_medium=oso&amp;utm_content=100000471967531&amp;utm_campaign=sm&amp;linkId=100000005252560"><em>Pega Blogs</em></a><em>.</em></p><p>My prediction for 2019 is that, for many companies, this year will be “The Year of Predictions.” Companies will discover that there is a more cost-efficient method to building “mountains of models” and making huge volumes of executable predictions, and they’ll use this streamlined predictive analytics capability to their advantage over their competitors.</p><h3>Making predictions is a valuable activity.</h3><p>Predictions allow you to anticipate your customers’ intents and needs. For example, predictions allow companies to focus investment in valuable customers that may be about to cancel a contract, or make certain a web page is populated with content that will most likely create engagement. Predictions ensure you only initiate outbound communications with customers who would actually welcome that contact at that moment in time.</p><h3>In short, predictions allow you to have the most relevant conversations at the right time and in the right channel — preventing time-wasting, annoying, irrelevant, and value-destroying interactions.</h3><p>If you can predict lots of aspects of what your customers want and how they will behave, you can create experiences and interactions that increase the value of the customer, rather than detract from a customer’s expectations and destroy that value. Predictions that are tightly embedded into executable interaction strategies inform business decisions.</p><h3>Driving down the cost of making predictions is important, if you are to be a prediction-based company.</h3><p>There is no point making a prediction if you do not act on it. Historically, cost stops companies from making and using predictions to improve engagement. Imagine a travel company that has built out predictive models allowing them to rank different destinations for a family’s next summer holiday. Would the family prefer to visit Italy or Florida? Predicting Italy as their most likely destination is not enough to create an engaging experience. The travel company might want to predict what is more appealing: food, history, beaches, or villas? They probably ought to know when someone is likely to book, over what channels, and if the family also wants to hire a car (and what class of car). Regardless of the scenario, the number of predictions required to tailor a highly personal and engaging experience can quickly escalate.</p><p>Instead of making predictive models to get the Italy experience right, the travel company might consider lots of crude, but cheaper, substitutes, like A/B testing various images of Italy, or some rough segmentation of customers. But to create a truly personalized, one-to-one experience, you need predictive models for every element of that experience. The cost of doing so often stops this from happening.</p><h3>Why has it cost so much to create predictions?</h3><p>There are many reasons. First, lots data is needed. Huge volumes of data may be available, but often it is in the wrong place, is old, and may be missing much of the required context. Worse, data about your new products and services may simply have not yet been generated.</p><p>Second, building models takes time and requires lots of skills. These skills are in high demand and are often best deployed to business-critical predictive models, such as fraud and risk. It can be hard to justify using such skills to decide whether to place a Facebook ad with a picture of Rome or a picture of Venice. Companies are used to creating more value from predictive models by concentrating on creating better models with advanced techniques, not creating huge numbers of “good enough” models. In the travel company example, most models you’d want to deploy need only be good enough to rank competing options on how to present that Italian holiday. Building best-in-class models here is unlikely to change that ranking. You do not need to be “correct,” you need to be effective.</p><p>Third, companies need to demonstrate that the models they create are compliant with corporate model standards. They need to be transparent as to how they used data to make predictions, and to be shown as such.</p><p>Fourth, these models need to be operationally deployed to inform decisions, strategies, and real-time engagements. An un-deployed model generates zero value. The effort demands significant IT resources to code, test, and integrate models across channels, applications, and marketing tools. Plus, models need to be refreshed often as they quickly become stale after having been deployed.</p><p>Fifth, such models need continuous reporting, monitoring, and governance to be set up in the enterprise system architecture around them.</p><h3>New digital tools are changing the economics of predictive modeling.</h3><p>Modern customer decisioning platforms are able to “close the loop” of making <a href="https://www.pega.com/predictive-analytics">real-time prediction-based decisions</a>, capturing behavioral responses, and training predictive models based on these responses. This allows business users to control the automated creation, deployment, and refreshing of potentially thousands of predictive models fully embedded within a real-time decision execution engine.</p><p>Companies are already putting this technology to use. For example, the Commonwealth Bank of Australia uses Pega’s <a href="https://www.pega.com/products/pega-platform/decision-hub">Customer Decision Hub™</a> to solve many of these cost issues and deliver “mountains of models” to scale up the number of predictions they make by orders of magnitude when engaging with customers. In the first week, CBA was able to automatically build and deploy 250 predictive models. This <a href="https://www.pega.com/customers/cba">increased lead volumes 10 times and trebled conversion rates.</a>According to their Chief Analytics Officer, Andrew McMullan, the effort would have normally taken 25 of their data scientists more than three years to complete using traditional methods.</p><p>CBA was able to drastically reduce the cost of predictions by employing the full-cycle automation capabilities of the Customer Decision Hub. Predictive models are built, deployed, and executed all within the hub, solving many of the cost issues discussed above. Because the models are built in context, the training data sets can be taken directly from contextual data available at the time of decision making and immediately tied to the outcomes of those decisions — like a click, a purchase, or an accept. Expensive data collection and transformation is not required.</p><p>The models are built using a templated approach that can be validated and tested before deployment. Efficient modelling techniques that produce “good-enough” predictive models can be scaled to produce potentially thousands of models simultaneously. The Customer Decision Hub takes care of periodically refreshing models into the real-time decision strategy engine and auditing this process. Various levels of reporting and visualization allow data scientists to monitor the models, understand how data is being used, and work with business owners to improve messages and propositions.</p><h3>This approach effectively creates a high-performance production line for predictions and predictive models.</h3><p>This approach will help collapse costs, making 2019 the Year of Predictions, as enterprises move from handcrafted models built by skilled practitioners to fully automating the process of creating training data sets, continuous and autonomous model building, seamless deployment, and closed-loop feedback.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c0054fc70e3d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[5 REAL-WORLD CHALLENGES OF DIGITAL PERSONALIZATION]]></title>
            <link>https://medium.com/@Decisionista/5-real-world-challenges-of-digital-personalization-ed6af0db2f0a?source=rss-2d07e0a8d645------2</link>
            <guid isPermaLink="false">https://medium.com/p/ed6af0db2f0a</guid>
            <category><![CDATA[marketing]]></category>
            <dc:creator><![CDATA[Dr Andy Lewis]]></dc:creator>
            <pubDate>Fri, 29 Jun 2018 17:16:01 GMT</pubDate>
            <atom:updated>2018-06-29T17:16:01.093Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*m_k9aO0iJY_Bz8Uy6s8PBw.png" /></figure><p><a href="https://www.pega.com/insights/articles/5-real-world-challenges-digital-personalization"><em>originally posted in pega.com</em></a></p><p>Not long ago, your bank manager knew your name. The butcher knew the cuts of meat you liked, and the High Street travel agent knew where your family enjoyed a holiday. Today’s businesses are trying to recreate that intimacy and empathy in their digital channels so they can better connect, serve, and sell to their customers. The problem is, the techniques and tools businesses are using to improve digital personalization have originated in the wrong place.</p><h3>The old way — optimizing content with segments and batches</h3><p>Most tools used have a heritage in content optimization. That is, they are designed to test and deploy variants of the web experience with the goal of optimizing specific metrics like bounce rates, click-through, or conversion. Different page layouts are tried. Alternative images and text are tested against each other. Through an experiment-and-learn approach, poor design and unappealing content can be eradicated to create a more compelling experience.</p><p>One big step in personalization was to develop tools that could apply different content based on the type of visitor. For example, with behavioral segmentation you could present individually-optimized experiences based on the visitor’s persona — which significantly improved the digital experience.</p><p>However, marketers that rely on behavioral segmentation will eventually hit a roadblock — because it’s difficult to further personalize an experience with segments when most of the essential customer context for the current situation lies outside of the digital channel.</p><p>To overcome this roadblock, marketers’ attempts to inject customer insight into digital channels have tended to rely on co-opting batch campaign engines to produce “leads.” But this approach doesn’t create better personalization either. Batch applications were designed for bulk outbound “spamming” and are poorly suited for the digital world. For example, they will be slow to pick up on the fact that the customer was in a retail shop this morning with a complaint.</p><h3>The challenges marketers need to address</h3><p>Content optimizations, even when targeted at customer segments and personas, can never be true personalization. Trying to make this work results in five personalization challenges that marketers need to overcome:</p><p><strong>1. Relevancy:</strong> There is no point working out the best message variant and presentation for a customer if that message is not relevant to them.</p><p>The tools that are very good at targeting the right content at a customer to get the highest click-rate do not necessarily generate content that is right for your visitor. By focusing on content variations, you ignore the bigger issue — relevancy.</p><p><em>It is far more important to get the most relevant message in front of your customer, than it is to find the “perfect” variation of an irrelevant message.</em></p><p><strong>2. Timeliness:</strong> There is no point delivering relevant messages if the timing is wrong.</p><p>You may be running lots of offline analytics to understand your customers’ next likely need or purchase — but this analysis is delayed, and there’s a big lag processing and loading the insights that you need to engage into your channels. Customers are now moving faster than the speed at which you can understand what they might need — their journeys happen across channels in real time, and are often over before you even can blink.</p><p><em>Customer relevancy is not confined to a channel and can change quickly.</em></p><p><strong>3. Consistency:</strong> You will undo your good work of being timely and relevant if you can’t be consistent in your next actions.</p><p>You may have placed the right message in front of your customer, and they have responded — but you’ll undo that good work if you can’t maintain that conversation consistently across your other channels. Nudging the customer along their journey requires a response that is consistent everywhere you engage customers, and the intelligence gained needs to be shared with your other channels and paid media — fast.</p><p><em>Outbound and paid channels often live in their own world, largely oblivious to the customer’s current state, or what’s happening on inbound — just as digital channels are oblivious to the outbound messages being sent out the door.</em></p><p><strong>4. Segmentation:</strong> You may optimize your messages for each segment, but your customer is not a segment.</p><p>Much of digital targeting relies on segmenting visitors. Customers are assigned personas, which are then associated with segments that are shaped by your views on what you think your customers <em>ought</em> to be doing. For every segment there is a message you want to deliver. However, your organization’s personalization strategy may often exceeds its capability to define and build all the segments required to produce the fine-grained targeting you desire.</p><p>Worse, customers rarely fit neatly into those segments — their behavior may deviate markedly from the “average” defined by the segment.</p><p><em>Your customers are not neatly defined by segments. Their intents, contexts, and needs are individual.</em></p><p><strong>5. Optimization:</strong> You may achieve better click-throughs and checkouts with optimization, but are you actually increasing customer value? Revenue? Profitability?</p><p>The actions you’re trying to influence may happen days, maybe weeks, in the future. For example, what you really want is to retain a customer at end of contract, or for them to draw down a loan, or activate their SIM card, or purchase a flight upgrade. But optimizing in-channel KPI may not be the best way to make that happen.</p><p><em>You need to be capturing true business successes and relating them to the decisions you make in the digital channel. Personalization should then be driven by predictions as to what content will result in the best business outcomes.</em></p><p>These challenges result in much of your valuable digital real estate being populated with content that is irrelevant, untimely, and incoherent — ultimately driving the wrong behaviors. You end up focused on pushing product, dominated by segmentation-based thinking, and driven by convenient (but unhelpful) metrics.</p><p>The relentless pursuit of increasing click-throughs and checkouts will mask deeper problems in customer experience. If you are not consciously driving and measuring experiences to directly manage both, you may be reducing customer value and harming retention rates, instead of setting yourself up for <a href="https://www.pega.com/challenges/customer-retention">longer-term, more sustainable success</a>.</p><h3>Steps to take to improve digital personalization</h3><p>Problems will persist while the culture of the organization sees digital personalization as an issue for the digital channels, and where metrics are only focussed on the performance of these channels. To provide <a href="https://www.pega.com/topics/customer-engagement">true digital personalization</a>, a new approach is needed. To conquer these personalization challenges, marketers should take the following steps:</p><ul><li><strong>Do not personalize “in channel.”</strong> Relevancy is not always visible there. Decisions about how to engage with a customer should be delegated to a centralized decision authority that has a full view of the customer across all channels, inbound and outbound, and all customer journeys. Channel consistency is only guaranteed by it being informed by your central “customer brain.”</li><li><strong>Make decisions in real time.</strong> Relying on batch processes and feeds cannot deliver timeliness. Current contextual data adds rocket fuel to your knowledge of the customer.</li><li><strong>Abandon segments and use automated machine learning at scale.</strong> <a href="https://www.pega.com/challenges/customer-acquisition">Use your interactions with customers</a> to automatically build predictive models as to how they behave. Personalize on a one-to-one basis using these models, taking into account the specific real-time situation, context, and history of each individual customer.</li><li><strong>Select the best experiences based on individual predictions of the best expected business outcome.</strong>Abandon the product push mindset and forget clicks and checkouts. Focus on the predicted value of the interaction and how this digital session helps reach overall goals for the customer and the organization.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ed6af0db2f0a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Does Your Chatbot Dream of Electric Customers?]]></title>
            <link>https://medium.com/@Decisionista/does-your-chatbot-dream-of-electric-customers-9ba3d00a7524?source=rss-2d07e0a8d645------2</link>
            <guid isPermaLink="false">https://medium.com/p/9ba3d00a7524</guid>
            <dc:creator><![CDATA[Dr Andy Lewis]]></dc:creator>
            <pubDate>Tue, 10 Apr 2018 10:36:01 GMT</pubDate>
            <atom:updated>2018-04-10T10:36:01.099Z</atom:updated>
            <content:encoded><![CDATA[<p><em>This post by me was first published on </em><a href="https://www.pega.com/insights/articles/chatbots-understanding-text-understanding-customers"><em>pega.com</em></a><em>.</em></p><p>Over the past few weeks, a Twitter joke was doing the rounds about Natural Language Processing (NLP):</p><p>WHAT DO WE WANT?</p><p>Natural Language Processing!</p><p>WHEN DO WE WANT IT?</p><p>Sorry, when do we want what?</p><p>I am sure we all recognise the frustration we often find when engaging with AIs as they try to understand us. In the past few years, speech response systems have become ubiquitous — they are in our phones, kitchens, and cars, helping us to navigate, find restaurants, and play our favorite music.</p><p>And, yes, we do marvel at them, but also get extremely frustrated as they fail to do the sort of things humans find simple. When I ask, “Do I need an umbrella in Boston right now?” my Amazon Echo remarkably tells me I do not need one, but fails to note that I am not in Massachusetts where the weather is fine, but in Lincolnshire, England where the rain is streaming down. Very few of my human associates would make such a mistake.</p><p>Your customers may have already interacted with a chatbot you have deployed to attempt to get answers to queries or even with help buying something. But you can bet they know they are engaging with a bot and not a human. Chatbots can appear clunky, limited, and be quick to give up. Most importantly, they show no understanding of the customer as an individual. They treat all customers the same, responding in the same way to all, lacking any contextual awareness of what the customer is trying to do. Can you be sure that your chatbot is improving your customer’s experience? Or is it just one more step in the way of the customer getting what they really want or need?</p><p>If your customer can easily spot they are talking to a chatbot because it cannot contextualise what is said to them, then your chatbot lacks customer empathy and should be retired.</p><h3>NLP lacks contextual understanding</h3><p>Typically, in Natural Language Processing (NLP), words are treated as the sole data input in each individual processing request. NLP tries to extract the topics, intents, entities (names, account numbers, etc.), and sentiments from this text. But often, there is not enough information to fully express the complete meaning of the sentence. This is because meaning is wrapped up in the context of the conversation. By just looking at words alone, the meaning is often lost, ambiguities persist, and errors are made. To be truly intelligent, a chatbot needs to do more than just process words, it needs to maintain stateful information and inject external contextual understanding and insight. In my Boston weather example above, the essential external contextual data concerning my current location was absent, resulting in entirely the wrong response.</p><p>Which brings us to a huge problem when we want to deploy NLP technologies to improve customer experience. NLP has customer experience applications in chatbots, email handling, social media monitoring, IVRs and call center assistance. In each of these areas, we want to automate as much customer interaction as possible, ask the relevant questions, and initiate the right actions, while maintaining great customer experiences. Poorly understood requests and inappropriate and unpersonalized responses degrade that experience.</p><p>NLP technologies only process the words presented, but organizations want and need more. They want to understand the customer’s holistic context, full intent, and best action to take. They want to engage in a way that is responsive, consistent, timely, and accurate. To do that, you need to personalize the response depending on the customer’s segment, life-time value, business issue, and stage in their journey. None of this vital context is in the text. So how do we blend customer insight and data with the presented text to extract deep meaning and insight, and transform this engagement into valuable action? The solution is combining context with intelligence.</p><h3>The customer is the context</h3><p>Context demands we bring into the mix customer memory, insight, predictions, rules, and a learning capability. Real-time Next Best Action (NBA) technologies are designed to process customer data and context to recommend the next action that will produce the best business value within that interaction. Combining NBA with NLP can produce an intelligence that contextualizes the words said, and predicts the most relevant and timely responses — like the next best question to ask; or the next best case information to offer.</p><p>Context allows the chatbot to hand off the customer to the most appropriate agent to manage a request. If a customer says, “I want to close my account,” then the action the chatbot takes will depend on if the customer has recently made a complaint and has a high churn risk, or is merely closing an old, unused facility.</p><h3>To evolve, chatbots need to learn from contextual data</h3><p>In order to gain deep empathy with customers, the chatbot needs to learn from each customer interaction. Given any NLP classification of a request, an NBA engine may have many possible responses to that request. Predictive models can be used to rank competing responses according to their likelihood of a positive business outcome. Using automated machine learning within the chatbot allows responses to improve over time as the chatbot becomes more experienced.</p><p>When customer context is taken into account, predictive models can apply outputs of textual analysis, such as conversation categories, intents and sentiments — plus combine these with customer data, product and transactional data, and real-time contextual data, like weather, location, and recent activity. This is the path to true personalization and accurate, relevant responses.</p><h3>Does Your Chatbot Dream of Electric Customers?</h3><p>So if we have many competing responses to a request, our empathic chatbot needs to understand if a chosen response was successful or not, and then spend some time dreaming of our customers — that is, analyzing responses, processing to understand why responses worked or did not, and then evolving to a better and more human chatbot.</p><p>However, given that we may be managing many hundreds of potential responses to customer requests, relying on manual human intervention to help chatbots learn will be cost-prohibitive. The chatbot needs to rely on automated machine learning, managing hundreds, if not thousands of predictions, and should be continuously building new predictive models in response to each customer interaction.</p><p>Our chatbot can learn how words are used in the context of your business and the specific intent for each individual customer within the context of their interaction. But to truly create a seamless experience for a customer, this intelligence needs to extend across channels. At some point the chatbot will want to conclude the conversation and, maybe, hand off to another channel or system. Unless that channel is also driven by the same Next Best Action contextual awareness, the customer will experience a disjoin and your good work may be undone.</p><p>The reverse is true, too: The chatbot should be learning from interactions in other channels. A common “brain” is essential for this. It provides insight across all of your customers’ experiences. For example, if a telco chatbot is asked what phone upgrade offer is best, the chatbot needs to draw on real-time learnings from other channels about what handsets and tariffs are best suited for this customer.</p><h3>Human-like empathy is essential</h3><p>Getting this right means chatbots will be able to offer customer experiences that are nearly indistinguishable from those presented by your best human agents now. Processing words is not enough. Electric empathy is the key. To be successful we need to fully focus on the customer, their context, their personal history, and not just the words they are saying.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9ba3d00a7524" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[THE HIGHLANDER PRINCIPLE OF MARKETING: ONE CUSTOMER BRAIN]]></title>
            <link>https://medium.com/@Decisionista/the-highlander-principle-of-marketing-one-customer-brain-cf38b3ac7798?source=rss-2d07e0a8d645------2</link>
            <guid isPermaLink="false">https://medium.com/p/cf38b3ac7798</guid>
            <category><![CDATA[marketing]]></category>
            <dc:creator><![CDATA[Dr Andy Lewis]]></dc:creator>
            <pubDate>Thu, 25 Jan 2018 19:59:33 GMT</pubDate>
            <atom:updated>2018-01-25T19:59:33.877Z</atom:updated>
            <content:encoded><![CDATA[<p><em>Originally published on </em><a href="https://www.pega.com/insights/articles/highlander-principle-marketing-one-customer-brain"><em>pega.com</em></a></p><p>Most of the customers I speak to now are looking for ways to make their marketing efforts more effective and their customers’ experiences more relevant, rewarding, and valuable. It is now almost universally accepted that the old methods of engaging with customers through mass outbound communications are failing in this regard. Response rates are vanishingly small because most bulk communications are utterly irrelevant to the majority of customers. Marketing departments are often targeted on the volume of communications they can make to customers, regardless of the relevance or success. In short, companies have become very good at training their customers to ignore them.</p><p>For a few years now, efforts have been made to improve relevancy and engagement through adding intelligence into each channel interaction. Segmentation, AI, and personalization strategies are used to better target messages at customers. Emails need not be identical, but can contain personalized elements. Webpages can contain targeted banners, text, and calls to action. Call center agents can be prompted with specific offers from CRM systems.</p><p>There are many vendors who specialize in creating intelligent, targeted content through channel-specific applications. And when implemented, valuable uplift occurs in offer acceptance, click-throughs, and engagement. There are good reasons for this. Whilst CRM systems may provide good knowledge of a customer to help personalize messages and offers, very often the best source of intelligence comes from gathering the context of the customer in the channel. If we know what the customer is doing right now, what they are asking or looking for, or can infer it from their digital behavior, we can better personalize their experience. Real-time context matters.</p><h3>Contextual data drives relevancy.</h3><p>Contextual data is then the rocket fuel of personalization. But businesses are now realizing it’s not enough to be just “of the moment” in the current channel.</p><p>Customers do not just interact on one channel when making a purchase or chasing a service request. They go on their own journeys, which may be difficult to predict, across their devices and channels at their own speed. If we have all these channels working independently, trying to nudge customers in different directions, then confusion and incoherence results. We do not want to offer a credit card on the web when the customer was in a branch yesterday discussing mortgages. We should not be sending emails with the latest offers when their last product is being repaired or replaced. In short, each channel needs to be aware of every other channel and work together to create consistent experiences. Digital channels need to be joined up with stores and branches. Messages in mobile apps need to be congruous with outbound email. Paid and social media needs to respond in real time to what customers are doing in an organization’s owned channels.</p><p>However, businesses are now realizing their marketing architecture makes this hard. Joining up channels is expensive with all these competing channel intelligences. It is just too difficult to shunt the data around across channel silos in a timely way, share and maintain rules and predictive models, and understand customer history and events. Customers move between channels in real-time. Their data and history does not.</p><h3>“There can be only one.”</h3><p>This is where the <em>Highlander Principle</em> needs to come in. The <a href="https://en.wikipedia.org/wiki/Highlander_(film)">1986 Christopher Lambert film</a> about Scottish warrior immortals told us “there can be only one” immortal to rule the world. While there are many, they will compete and fight with each other. And so it is with “customer brains.” With many intelligences fighting to get the attention of the customer, your world will be one of chaos and confusion. There can be only one customer brain. Your customer intelligence cannot reside in multiple brains in each channel, but needs to be unified and coordinate journeys across all channels.</p><p>With a single brain, we have one customer authority. There is one central and shared intelligence that decides on the next best action for that individual customer. The single brain is not just directing each channel, but is also aware of what has been happening in each channel in real time. It listens to the customer no matter where they might be, remembers what the customer has done, knows the customer’s details intimately, understands their intent, predicts their future, and arbitrates over all possible actions to decide which action makes the most business sense right now. The result is a coherent set of messages across customer journeys, channels, and interactions that guides and nudges the customer to an optimal outcome.</p><p>Pega customers who have implemented the <a href="https://www.pega.com/products/pega-7-platform/decision-hub">Pega® Customer Decision Hub</a> across multiple channels to create joined-up journeys are now seeing remarkable results.</p><ul><li><a href="https://www.pega.com/customers/transavia">Transavia</a>, the “Netherlands’ friendly airline,” uses the Customer Decision Hub to provide each passenger with a contextual, consistent experience. Pega provides an intelligent interaction platform to break down organizational silos and sufficiently equip employees at all customer touch points to make smart decisions based on relevant data.</li><li><a href="https://www.pega.com/customers/rbs">The Royal Bank of Scotland</a> has joined up 18 channels for its 17 million customers, and supports over a hundred customer journeys with the Customer Decision Hub. They have seen a 20 percent increase in mortgage retention, a web-to-branch transition increase from zero percent to 40 percent, and a 10–1 ROI from this approach in 2015.</li><li><a href="https://www.pega.com/customers/sprint">Sprint</a>, a U.S. tier one carrier serving 60 million customers, saw a 10 percent reduction in customer churn and a 40 percent increase in transactional NPS using next-best-action retention offers from the Customer Decision Hub.</li></ul><p>Pega customers who are following the Highlander Principle — putting the customer at the heart of their business and coordinating their journeys with a unified Customer Decision Hub — are setting the standard for customer engagement.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cf38b3ac7798" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>