The age of adaptive journeys

Somebody lands on your site from a search result, reads for four minutes, looks at two case studies and leaves. You know almost nothing about them. They have not logged in, they will not fill in a form, and there is a good chance they are on their phone between meetings with no intention of telling you anything at all. That visitor is not an edge case you handle after the important work is done. On most sites that visitor is the norm, and everything we build for the people who do identify themselves is built for the minority.

The industry has a well-rehearsed answer to this, which is to make the stranger less strange. Identify them, enrich the record, match the company, drop them into a nurture flow. I understand the pull of it completely - marketing teams are measured on pipeline rather than on how pleasant the reading was, and a named lead is a number you can put in a report on Friday, while a good experience is not. The instinct is not stupid. It is just aimed at the wrong problem, because it treats anonymity as the obstacle when anonymity is simply the condition, and it will still be the condition after you have spent a year and a budget trying to solve it.

The cost of getting it wrong is not zero

There is a comfortable assumption that personalisation is upside-only, that a slightly-off recommendation is merely a wasted slot, and BCG's research is a useful corrective. Across more than 23,000 consumers, two-thirds said they had recently had at least one personalised experience with a brand that was inaccurate or invasive. Those are two different failures with the same root: inaccurate is being confidently wrong about someone, and invasive is being uncomfortably right in a way they did not sign up for. Both come from a system trying to work out who you are rather than paying attention to what you are doing.

Inaccurate is being confidently wrong about someone. Invasive is being uncomfortably right in a way they did not ask for.

Behaviour tells you more than identity does

Here is what I find genuinely strange about the identification instinct: the anonymous visitor is already telling you a great deal, and almost none of it requires knowing their name. Someone who reads two pieces about workflow automation and then opens a retail case study has described their problem more precisely than most contact forms manage. Someone who arrives on a technical deep-dive from a search for a very specific integration question has told you what they are worried about. Someone who has been through four pages in ninety seconds is scanning rather than reading, and would probably like the summary rather than the essay.

None of that is personal information. It is context, it lives entirely in the current visit, and it is a far better guide to what to show next than any inferred demographic, because it describes the job somebody is trying to do right now rather than the category a data provider has filed them under. The uncomfortable truth for a lot of martech is that the most useful signal was always sitting in plain sight, and it got ignored because it could not be exported into a CRM.

What adaptive actually means

The word gets used loosely, so it is worth being precise about the three things a journey has to do before it earns the description. It has to respond, which means the next screen reflects what just happened rather than what a campaign planner decided six weeks ago. It has to learn, in the modest sense of accumulating context across a visit so that page six is better informed than page one. And it has to adapt, which is the hardest of the three, because it means being willing to change what is on the page and not merely which banner sits at the top of it.

There is a fourth requirement that matters more than the other three, and it is the one most systems skip. An adaptive journey has to be honest about what it actually did. If a page tells a visitor it is showing work from their sector, it had better be showing work from their sector, and if there is not enough of it then the right move is to say something smaller and true rather than something larger and hollow. That sounds like a small piece of copywriting discipline. It is not. It is the whole difference between a system that feels attentive and one that feels like it is guessing at you, and a visitor can tell the two apart immediately even if they could not say why.

Asking is allowed, as long as you earn it

None of this is an argument against ever asking someone a question. It is an argument about the order of things. Ask before you have been useful and you are running a toll gate, which is why gated content so often collects a fake email address and a small amount of resentment. Ask after you have been useful, about something that visibly changes what happens next, and it stops being data collection and becomes a preference - closer to choosing a seat than to filling in a form. The test I like is simple: if the answer would not change a single thing on the page, you have no business asking for it.

What we are building on our own site

We are putting this into practice on touchfoundry.co.za under the name Adaptive Journeys, and I would rather describe it plainly than oversell it. A visitor will be able to tell us as much or as little as they like about their sector, their role and what they are trying to do, and the site will reorganise itself around that - the work we lead with, the writing we surface, the order of the things we offer. Everything stays available; nothing gets hidden behind an answer. If somebody tells us nothing, they get our best general view, which is the version that exists today.

The part we have been strictest about is the part nobody sees. It runs in the visitor's own browser for the length of their visit, it is written to be legible rather than clever, and the summary it shows will only claim what the page beneath it can actually back up. We are building it this way because it is the version we would want to be on the receiving end of, and because a studio that argues for restraint in public should be able to point at its own site as the worked example rather than the exception.

The ambition is smaller than it sounds

Adaptive journeys get talked about as though the goal were a system that knows you, which is both a bigger claim than anyone needs to make and a slightly unnerving one. The goal is much smaller and much more useful than that. It is to be genuinely helpful to somebody who has told you nothing, on the strength of what they are doing in front of you, and to keep being helpful whether or not they ever decide to say who they are. Get that right and the identification problem mostly solves itself, because people tend to introduce themselves to things that have already been good to them.