Roughly 750 retail marketing executives and industry experts will descend on Chicago June 5 for the Customer Relationship Management Conference (CRMC), one of the premier retail marketing and CRM events of the year. The theme for this year’s event is “Join the Expedition”; the event’s homepage depicts a circuitous route to a mountaintop destination, informing attendees that the dozens of speakers – “specialists who know the route” – will guide them to the peak of retail marketing: creating an incredible, real-time customer experience.
A mountaineering expedition is akin to today’s non-linear, omnichannel customer journey. A continuously connected consumer with multiple devices and an endless number of interaction touchpoints is on a multi-dimensional journey, appearing to a brand at any time and on any channel. Marketers know they must create a personalized customer experience across the entire journey, and recognize the customer as an individual throughout the highly customized path-to-purchase.
According to a recent Harris Poll survey commissioned by RedPoint, neither consumers nor marketers give brands high marks for their ability to deliver an exceptional customer experience. Just 18 percent of consumers rated brands’ ability to deliver an exceptional experience as “excellent”. Marketers had a better view of themselves, with 34 percent rating themselves as “excellent”. Clearly, marketers have their work cut out for them.
To engage with the connected consumer, marketers must forego the traditional channel-specific “batch and blast” approach that sufficed when customer choice was limited and the buying journey followed a mostly straight-line path. To deliver highly personalized experiences that are relevant for each customer in the context and cadence of an individual journey requires advanced analytical tools. AI and machine learning help marketers create real-time, personalized experiences at scale with models that intelligently orchestrate the complete customer journey.
My colleague Steve Zisk, RedPoint’s senior product marketing manager, and I will present a discussion at CRMC about the hype vs. the reality of AI and machine learning for marketing organizations tasked with creating real-time customer engagements. In our experience, marketers recognize that AI and machine learning are invaluable tools to help engage with the connected consumer, but they struggle with how to reconcile a specific business problem with advanced technologies. Our session will focus extensively on best practices, common use cases, and speed-to-value for marketing organizations who want to drive value with next-level customer engagement, personalization, and path-to-purchase optimization.
Supercharging Engagement with Machine Learning
Marketing’s response to shopping cart abandonment offers a typical example for how AI and machine learning supercharges engagement strategy. Sending the owner of an abandoned cart an email reminder is a basic response that can be automated without the use of a machine learning algorithm. The issue, though, is that an email that treats every customer the same regardless of the items abandoned, the reasons for the abandonment, or the context of a specific journey will likely introduce friction into the customer journey. An email reminder to a customer who already bought the abandoned items in-store, for example, creates frustration for the customer.
Machine learning models help eliminate friction with algorithms that tailor an abandonment response to a wide range of factors using first-party, second-party, and third-party customer data. A model that explores a customer’s abandonment history, for example, uses what it learns about each abandoned cart to inform a personalized next-best action or recommendation. Beyond basics such as quantity and cost of abandoned items, detailed analysis that examines everything the customer later bought in-store, whether an item had been abandoned before, or even the influence of sentiment analysis and other unstructured data provides marketers with a window into intent.
AI and machine learning models can also provide marketers with an ability to sequence the entire customer journey, or path to purchase. Leveraging all customer data and applying models such as propensity, regression, or finite state machine models, retailers can discover patterns in omnichannel activity that recommend the next 4-5 steps in the process, whether it’s an email, a website offer, a mobile app push, or direct mail piece. AI and machine learning can ensure that marketers focus on the strategy and content, and let the machine do the heavily tactical lifting of who gets what, when, where, and how.
No More Guesswork: Data Trumps Human Intuition
A real-time, intelligent cart abandonment strategy and path to purchase use cases require a lot of behind-the-scenes instrumentation to ensure the validity of results. Marketers need to understand what data to collect, they need to verify that it is accurate and representative, and they need to map the business use case with the technology they’re using to ensure that the model they’re creating will answer the right questions.
The access and use of metadata helps marketers devise a more data-driven strategy by providing greater context to interactions, offers, transactions, and touchpoints. Infusing metadata from the campaign back into an intelligent model provides marketers with greater confidence in results and allows for more intelligent orchestration.
Metadata also is “the tip of the spear” in improving accountability in modeling. Models embody a point of view on customer engagement that can appear biased or neutral, hip or stodgy. Younger consumers will embrace or shun brands based on perceptions of transparency, inclusiveness, and honesty. Marketers must use metadata to measure, test, and adapt their models to ensure their customer engagements are aligned with consumer values. Beyond data questions, marketers must also know how to activate a model once it has been built, they need to know how to make it work in a specific campaign, how to produce results, how to feed those results back into the model, and how to verify that the model is working as intended. Our session will address these and other concerns, as well as delve into the concerns marketers have about the impact machine learning will have on their day-to-day operational responsibilities.
Machine Learning: A Partner to Marketers
Some marketers are reluctant to embrace AI and machine learning because they’re fearful that by feeding data into a model they’re actively ceding creative control to the technology and that they will spend their days in servitude to the intelligent machine.
The concern is one of a few common misconceptions about AI and machine learning that are easily dispelled. In reality, machine learning capabilities give marketers the tools to become more creative. An analysis of a shopping cart abandonment, for example, will reveal new patterns that raise new questions for marketers about an individual customer’s behavior. This might influence the creation and testing of a new model, and put marketing in the driver’s seat for creating new hyper-personalized engagements based on new findings. The more that marketers embrace machine learning as an ally, the more they’ll distance themselves from a one-size-fits-all email as they move closer to hyper-personalized engagement.
Another common misconception is the idea that marketing organizations that have yet to begin using AI and machine learning are far behind the curve and are doomed to be perpetually behind. Similarly, some marketers feat that they must incorporate AI and machine learning into every campaign or process all at once or risk remaining stuck in the “batch and blast” dark ages.
The truth is, the real value AI and machine learning deliver to marketing is in many ways just an extension of what marketing software has delivered for years; the ability to try new things, to amplify creative strengths, to reveal new discoveries, and to enable collaboration across teams and channels. When marketers get bogged down in the hype surrounding AI and machine learning, they often worry about what they have to do. By asking instead what they want to do, they may be surprised at how easy it is to embrace AI and machine learning as guides to help them reach the retail marketing summit of real-time, hyper-personalized experiences across an omnichannel customer journey.
You can find more information about CRMC and our session, “Leveraging AI and Machine Learning for Real-Time Customer Engagement: Hype vs. Reality” by visiting the CRMC detailed agenda. To learn more about how to simplify analytics with AI and machine learning, the RedPoint AI Studio solution brief details how to align business objectives with advanced analytic models and put the power of AI into the hands of marketers.