Horror Films, Data Analytics, and Customer Journey Mapping

A Forrester analysis of customer journey-mapping tools, released this month, analyzes 13 different vendors, including Adobe, NICE, SuiteCX, Teradata, Kitewheel and others, examining each vendor in terms of:

• Data fusion, that is, how well a vendor can connect data across different journey steps to create a unified view of each individual customer’s experience;
• Design and planning, which includes creating a repository of journey-maps to analyze current and future behaviors and track interactions;
• Testing and optimization, or the ability to test various hypotheses, measure results, and optimize future journeys; and
• Automation and orchestration, which requires predictive and prescriptive analytics to enable automation of future interactions. 

You can download Forrester’s report from NICE’s web site and read it for yourself, but something I found very interesting about their analysis was the way they categorized the vendors into two groups, based on how they create their journey maps – either Bottom-Up or Top-Down.

Bottom-Up Journey Mapping

Ten of the vendors were identified as bottom-up journey mappers, because their journey-mapping capabilities spring from the streams of transactional and other data generated by customers. With enough contact center, web site, VOC and CRM data you can map out different kinds of journeys for different types of customers and situations. But the data-driven approach is based on inductive reasoning, which basically means cataloging lots of details first, and then trying to hypothesize the general principles from those details.

This kind of “brute force analytics” is well suited for data-rich environments, but it has a very big drawback. Let me illustrate with a question:

Do you like horror films?

According to OK Cupid’s matchmaking database, when it comes to forecasting relationship longevity, “Do you like horror films?” is the single most predictive question! But why is that? No one knows why. It just is. This illustrates the nature of the problem we face when we use inductive, data-informed reasoning all by itself – a problem that is becoming increasingly clear in the era of Big Data.

And it’s a problem that a bottom-up, inductive approach to journey mapping also has. A bottom-up, data-only approach might be able to give you a detailed picture of what your customers are doing, but it simply can’t tell you why they’re doing it, even when the reason may not be at all obvious.

Top-Down Journey Mapping

The other three vendors (including SuiteCX, the journey-mapping company I advise) were classified as top-down journey mappers, meaning that their mapping process is based on deductive reasoning, rather than inductive. Deductive reasoning involves a kind of “meta” thinking process – that is, it starts with informed thinking about the general principles that apply, and then deduces specific facts and details from these general principles. Think of deductive reasoning as a kind of story-telling, disciplined by data and judgment.

But this top-down approach has its own pitfalls, beginning with the requirement that to be useful it must be informed by the reasoner’s “judgment” – which is one of those words that applies only to the human intellect, and not to an algorithm or a database. People have varying degrees of judgment about things, and the quality of your judgment about any subject will be based on your general level of expertise, as well as your familiarity and prior experience making decisions in this field.

What this means is that while top-down journey mapping begins with story-telling about different types of customers, the quality of these stories, as well as the types of customers and situations chosen, makes a big difference. If your full-time job is unrelated to how customers experience the brand – say, your job function is in payables, or maybe logistics, or HR – then the quality of your judgment when it comes to customer journey mapping probably won’t be as high as the quality of judgment shown by, say, a digital marketing analyst. Not that you won’t have interesting and useful ideas, but top-down journey mapping requires taking an informed and disciplined approach if it is to be effective.

Both inductive and deductive reasoning are important if you really want to understand your customers with any confidence, but either form of analysis, when used by itself, will have flaws.

And the Forrester report says as much, cautioning that journey-mapping is not a tool, but a practice. To be done professionally, according to Forrester, it should be thought of as both science and art.