October 17, 2026 · Martina Pérez
What if your product knowledge updated itself?
You are in a call with product managers presenting a design and they start challenging every piece of it. You are in a call with engineers, and they don't understand why you are even working on something. You are working with another designer, and they keep challenging the reasoning behind your designs.
Let's assume you're doing the process right. The problem is clear. You've done the discovery work. You've explored different solutions. The design is solid.
So what's happening?
I think sometimes the problem isn't the design. It's that everyone is looking at the problem from a different place.
You have the research in one place. The product metrics somewhere else. Customer feedback in another. Previous decisions are buried in documents or Slack. And everyone arrives at the conversation carrying a slightly different version of what they think is true.
You have the problem clear, and then you get ten different opinions about it. Is it actually affecting the journey? How much? Is it happening to many users or just a few? Is it getting worse? Is there evidence elsewhere? Does another product solve it differently? Everyone has a different hypothesis, and there is no common ground.
A lot of product disagreement isn't disagreement about the solution. It's disagreement about the problem, the evidence, and the context behind it.
The problem isn't that product teams don't have enough information.
Our understanding of a product changes every week. Our product artefacts usually don't.
A new research session reveals a pain point. Analytics show a drop in a funnel. Someone on the team notices a competitor doing something differently. A new pattern appears somewhere else.
But the journey map gets updated every few months, and it slowly drifts apart from everything else.
That's what made me start using AI differently: not to generate more research, but to keep what we've already learned connected.
So I started wondering: what if the user journey could become a living artefact?
What if we could scan, every week, the information that's already there for us and map it against our current user journey? Research findings. Funnel data. Customer feedback. Internal observations. Competitive patterns.
The idea isn't entirely new. Tools like TheyDo, Smaply, UXPressia and Dovetail already help teams connect parts of this picture. But most teams I know still have the pieces scattered across a research repository, an analytics tool, a feedback inbox and internal documents.
So I wanted to see how far you can get by connecting what's already there, without adding another product.
I've already built a few automations in n8n to explore this. I work at n8n, so it's the tool I reach for, but the idea doesn't depend on it. Any tool that can connect your research repository, your analytics and your feedback inbox would do.
A research session mentions that someone never got the verification email. It's tagged to Sign up, and appears next to three earlier sessions describing the same moment. One complaint is an anecdote. Four is a pattern.
The funnel shows that almost half the people who discover the product never sign up. That drop now sits in the same column as those four sessions, so the number and the quotes describe the same moment.
A similar product starts letting people try a template before creating an account. That pattern is mapped to Sign up as a possible opportunity. Another product's new pricing page isn't, because it doesn't touch this journey.
And there's another one I'm working on: investigating the hypotheses behind the problems we see.
Together, these don't create another dashboard. They create something closer to a living product map.
But there is one important distinction.
That journey doesn't represent your opinion. It represents what the different qualitative and quantitative data is showing. The team's opinions are in there too, as one input among the others, not as the frame.
Once you have this, you can start forming your strong opinions about the best solution. But now you have a shared understanding. You can actually communicate and speak the same language.
And you have informed opinions, rather than opinionated ones.
That, to me, is the interesting part.
AI shouldn't replace product thinking. It should reduce the cost of keeping product knowledge alive.
This approach has an obvious risk: it can look objective while being full of choices.
How are research sessions summarised, and by which model? What goes into a summary, and what gets left out? Who decided this funnel matters more than another? What feeds an opportunity, and who decides how much it weighs?
Every one of those questions needs a person behind it. Depending on the company, that might be a product manager, a product designer, a data analyst or a researcher, or several people for different parts of the journey.
It matters less who does what than that someone does. Each part of the journey needs a named human owner who checks that the automation works and that what it brings in is reliable. That isn't optional. Each update also links back to its source, so anyone can check where a claim came from and challenge it.
AI is the collector, not the editor. It gathers what the team already knows. The people who know each area decide what it means.
Instead of asking "What do we know about this part of the journey?", the team can start with an always-current picture.
And a simple rule:
If you've spent more than three minutes debating what the problem is, what users are struggling with, or how it affects the metrics, stop and look at the journey.
If the answer is there, the conversation moves on to what to do about it. If it isn't, you've found a gap in what the team knows, and that becomes the next thing to learn.
Teams stop spending their time agreeing on what's true, and start spending it on deciding what matters.
I'm going to share each automation individually: what worked, what didn't, and what I'd change.