Personalisation – is it experimentation with purpose?

Personalisation gets talked about like it’s a feature — something you plug in and switch on. In reality, it’s a way of working. And more than anything, it’s a form of experimentation.

If you're starting to build a personalisation practice, or trying to improve one that’s underwhelming — it helps to reframe it not as a “thing to deliver,” but as a discipline of learning: trying something specific, based on data, observing what happens, and adapting accordingly.

This shift in mindset can make a big difference. It removes the pressure to get it “right” first time and opens up space to ask better questions.

Start with questions, not answers

Who are we personalising for? What decisions or behaviours are we trying to influence? What hypotheses do we have about what might help?

These questions give you something to test, not just something to implement. And they force a bit of strategic discipline. Personalisation should be tied to outcomes — not just layered on top of existing journeys for the sake of it.

Your data is your starting point

Good experiments rely on good data. That doesn’t mean you need a full customer data platform or real-time targeting engine on day one. But you do need some sense of how your users behave, and how different signals (like visit history, behaviour, or preferences) might inform better decisions.

Often the first job is just to connect enough dots to make an educated guess — and to be able to tell if your guess was any good.

Build small, test fast, learn quickly

You don’t need a grand personalisation strategy to get going. What you need is a loop: identify an opportunity, make a small change, measure what happened, and decide what to do next.

For example:

If you think new users from organic search want different content than repeat visitors, change the homepage modules they see — and watch how behaviour shifts.

If you think a certain segment responds better to a different tone in onboarding, run an A/B test in your emails.

If you think product recommendations are helping conversion, isolate and measure the impact over time.

Each of these is a form of personalisation — but more importantly, they’re structured experiments that help you learn what works.

Teams, not tech, make this work

It’s worth saying: no personalisation programme succeeds without collaboration. You need marketers, product people, analysts, and engineers working with shared intent. That means aligned goals, open conversations, and mutual respect for the different levers each team controls.

It’s not about building a perfect tech stack first. It’s about building the right habits — asking better questions, measuring impact, and iterating with purpose.

In summary

If personalisation feels hard, it might be because it’s being treated like an end state — not the experimental, iterative practice it really is.

The teams that make it work aren’t just better at targeting. They’re better at learning. They try things, measure carefully, and use what they learn to make the next step more valuable.

And when you think of it that way, personalisation stops being about complexity — and starts becoming a way to build smarter, more responsive digital experiences over time.

If this topic is of interest to you, why not join us for our upcoming webinar Painless Personalisation in 180 days - where we'll show you practical ways to start small, prove value early, and build momentum with Optimizely. Register here.