How AI can (and can’t) help with customer journey mapping

Where AI genuinely helps

Synthesising large volumes of data

If you’ve run a research programme and you’re sitting on transcripts from a series of in-depth interviews, AI can help you identify patterns, themes, and recurring language much faster than a manual read-through. It won’t replace the human interpretation, but it accelerates the process significantly, and gives you another pair of “eyes” through which to sense make.

It is also very useful to help you combine and coalesce a variety of information sources you likely already have in house — sales data, CX data, reviews and other social media monitoring.

Where you shouldn’t skimp — getting real data from real people.

Building the assumptive map framework

AI tools can help develop a first-draft assumptive journey map by drawing on publicly available information about how customers in a given category typically behave.

However, they are generic and lacking in helpful nuance.

We recently did this as an experiment on a client project. The difference between the generic AI generated customer journey map and what the team, with their deep, in-person experiences of the business developed over a day is chalk and cheese.

However, these broad, generic AI maps are useful as a starting point for the internal assumptive mapping workshops. They can provide the broad framework for the journey — building possible key stages. This saves time in the early stages of a workshop, freeing up space to get into the weeds (which is where the interesting nuggets lie) with the team.

What’s more, doing an AI generated CJM also helps external facilitators like us familiarise ourselves with the category, so we are more knowledgeable from the get-go. It’s another source of information, rather than the final output.

Competitor and landscape research

AI-powered research tools can rapidly survey competitor digital experiences, review platforms, and industry commentary. This kind of landscape research used to take days; now you can get a good overview in a few hours.

Again, it should be assessed as another source of information, rather than taken as truth. But an easily accessible one that will contain worthwhile information.

Structuring and formatting outputs

This is perhaps one of the most exciting uses of AI with the CJM process.

Take photos of the walls of post-it notes generated in the workshop, and a well-instructed AI agent can transcribe them and pull them into a digestible format far easier and quicker than humans can.

And, with its capacity to “think”, AI gets you 60% of the way to your final product, and gives you easily accessible outputs you can then interpret, add to and finalise based on what you’ve learned about the business.

Once the strategic thinking is done, AI can help create well-structured documents, presentations, and visual briefs faster than traditional drafting.

We probably spend as much time here as before — but we are spending it crafting and building rather than transcribing and doing that initial structure. As a result, the assumptive maps are deeper and richer than they might have been without AI’s practical help.

Where AI doesn’t replace human judgement

Real customer conversations

No AI tool can replicate a well-run in-depth interview with a real customer. The nuance, the body language, the things people say when they feel heard, the follow-up question that opens up an unexpected insight come from human skills and lived experience.

Real customer conversations hold the most valuable material for building customer-centric, winning organisations. And they require a skilled human researcher to get this value.

This is not the place to skimp or substitute with AI or synthetic customers or moderators. This is the gold from which insight emerges. Talk to real people. Use skilled moderators to untangle what people say from what they do, and go deeper where it matters. You can use AI to help you make sense of this data later, but gather the data the old-fashioned way. Not only do you get richer insights, but you are also building your intuition and keeping customer understanding within the business.

Interpreting and applying outputs to the business

AI is pretty good at pattern recognition. At least, in terms of the data set in front of it.

However, knowing which patterns matter for your organisation, your customers, and your competitive context is another matter.

Humans see better patterns — ones that lead to ideas and opportunities for your business because we have an embodied experience of the world. The patterns we see emerge because of our experiences living in the world, and our organisations. It comes from lived experience, commercial judgement, and an understanding of the business that AI doesn’t yet have. Even ones that are “part of the team”.

Making the strategic recommendations

AI can tell you what the data says. It cannot tell you what to do about it in a way that accounts for your organisation’s culture, capability, budget, and risk tolerance. It will, however, tell you it can, with 100% confidence. Do not be fooled.

The bottom line

AI makes good journey mapping faster and more efficient.

It does not make bad journey mapping good.

The quality of the output still depends on the quality of the thinking, the rigour of the research, and the experience of the people doing the work.

Want to understand how AI fits into a research and strategy engagement? Let’s have a conversation.