Personalisation in Digital Marketing: Where It Works and Where It Wastes Budget

Personalisation in digital marketing means delivering content, offers, and experiences tailored to individual users based on data about their behaviour, preferences, or context. When it works, it increases relevance and drives measurable commercial outcomes. When it is poorly executed, it burns budget, erodes trust, and produces a version of marketing that feels intrusive rather than useful.

The gap between those two outcomes is almost always a strategic one, not a technical one.

Why Most Personalisation Programmes Underdeliver

I have sat in enough agency briefings and client strategy sessions to recognise the pattern. A brand invests in a customer data platform, integrates it with their email and paid channels, and then spends six months configuring rules. The first results are promising. Open rates improve slightly. Click-through rates tick up. Someone in the room declares it a success and moves on to the next project.

What rarely gets examined is whether the personalisation is actually driving incremental revenue or whether it is just making existing intent look more responsive. That distinction matters enormously when you are trying to justify continued investment.

The problem is structural. Most personalisation is built around what data is available rather than what decision it needs to support. You have email open history, so you optimise send times. You have browse data, so you retarget on product pages. You have purchase history, so you recommend adjacent products. None of this is wrong, but it is reactive. It optimises the channel rather than the customer relationship.

Personalisation that genuinely moves commercial metrics starts with a different question: what does this customer need to believe or feel in order to take the next step? Everything else flows from that, including which data is actually useful and which is just noise dressed up as insight.

If you are thinking about where personalisation fits within a broader commercial marketing framework, the Go-To-Market and Growth Strategy hub covers the strategic context in more depth, including how to sequence investment across channels and customer lifecycle stages.

The Data Problem Nobody Wants to Talk About

Early in my career, I was working on a campaign for a travel brand and we had what felt like a rich data set: destination browse history, previous booking behaviour, email engagement. We built out what we thought was a sophisticated segmentation model and launched a personalised email programme. The results were mediocre. Not terrible, but not the step change we had promised.

When we dug into why, the answer was uncomfortable. A significant portion of the browse data was stale. People who had looked at beach holidays in January were still being targeted with beach content in April, by which point many of them had already booked elsewhere. The data was technically accurate but commercially useless. It was describing a past version of the customer, not their current intent.

Data freshness is one of the most underappreciated variables in personalisation strategy. It is not enough to have data. You need data that reflects where the customer is right now, not where they were three weeks ago. The shelf life of behavioural signals varies enormously by category. In travel, intent signals can decay within days. In financial services, they can persist for months. If you are not building decay logic into your segmentation, you are personalising against a ghost.

There is also the question of data completeness. Most brands are working with partial pictures. They can see what happens on their own properties but have limited visibility into the broader context of a customer’s life. Someone who stops engaging with your emails has not necessarily lost interest in your category. They may have bought from a competitor, changed jobs, or simply shifted to a different channel. Treating silence as disengagement and ramping up frequency in response is one of the most common and most damaging personalisation mistakes I see.

Where Personalisation Actually Creates Commercial Value

The clearest commercial case for personalisation is in reducing friction at high-intent moments. When a customer is close to a decision and the experience helps them complete it faster or with more confidence, personalisation earns its budget. This is not a sophisticated insight. It is just good marketing applied with precision.

When I was running paid search campaigns at lastminute.com, we were working with a relatively simple form of personalisation by today’s standards: matching ad copy to search query, landing page to ad message, offer to browsing history. But the commercial impact was immediate and measurable. A campaign for a music festival drove six figures of revenue within roughly a day, not because the targeting was complex but because the message was precisely matched to what people were looking for at the moment they were looking for it. Relevance, delivered at the right time, is still the most powerful lever in digital marketing.

The categories where personalisation consistently delivers strong returns share a few characteristics. They tend to have high consideration cycles, where customers research before they buy. They tend to have meaningful product variation, where the right recommendation genuinely helps. And they tend to have data environments rich enough to support inference about intent, not just identity.

E-commerce, financial services, travel, and subscription businesses all fit this profile. In these categories, well-executed personalisation can meaningfully improve conversion rates, reduce cost per acquisition, and increase average order value. The BCG analysis on financial services go-to-market strategy makes a related point about how understanding evolving customer needs at a granular level changes the economics of customer acquisition.

In categories with shorter consideration cycles or more commodity-like products, the returns are less clear. Personalising a grocery shop has some value in convenience but limited impact on brand preference or margin. The effort-to-return ratio shifts, and the investment case becomes harder to sustain.

The Segmentation Trap

One of the most common failure modes I have seen across agencies and in-house teams is what I would call over-segmentation. The logic sounds sensible: the more granular your segments, the more relevant your messaging. In practice, it often produces the opposite effect.

When I was growing an agency from around 20 people to over 100, we were working with clients across 30 industries, and the segmentation question came up constantly. The clients who built too many segments invariably ran into the same problems. Creative production could not keep pace with the number of variants required. Statistical significance became impossible to achieve in testing because audiences were too small. And the operational overhead of maintaining dozens of audience definitions meant the strategy was always six months behind the data.

The clients who performed best were the ones who identified three to five genuinely distinct audience states, each with a clear commercial logic, and built their personalisation around those. Not demographic segments. Not persona archetypes. Behavioural states: people who are aware but not engaged, people who are engaged but not converting, people who have converted once but not returned, and so on. Each state has a different job to do and a different message that serves it.

This approach scales. It is testable. And it keeps the focus on commercial outcomes rather than on the sophistication of the segmentation architecture itself.

Personalisation and the Privacy Shift

The deprecation of third-party cookies and tightening privacy regulation have changed the data environment significantly. Brands that built their personalisation programmes on third-party audience data are now reckoning with the consequences. First-party data strategies, which should have been the foundation all along, are now a commercial necessity rather than a nice-to-have.

This shift is not a crisis for personalisation. It is a correction. The most durable personalisation has always been built on the relationship between a brand and its own customers, not on data rented from intermediaries. Brands with strong first-party data assets, built through genuine value exchange rather than data harvesting, are better positioned now than they were three years ago.

The value exchange point is important. Customers will share data when they can see a clear benefit. They will not share it to be tracked. A loyalty programme that delivers genuinely personalised offers based on purchase history is a value exchange. A cookie wall that demands consent to continue browsing is not. The distinction is obvious to users even when it is not obvious to the marketers designing the experience.

Tools like Hotjar and CrazyEgg have long been used to understand on-site behaviour in aggregate, and they remain useful inputs into personalisation strategy, particularly for identifying friction points in the conversion experience. But they are inputs, not outputs. The insight they generate needs to be connected to a commercial hypothesis before it can drive a personalisation decision.

Dynamic Creative and the Relevance Illusion

Dynamic creative optimisation has become one of the most widely adopted personalisation tools in digital advertising. The premise is straightforward: assemble ads in real time from a library of creative components, matching the combination to audience signals. It sounds compelling. In practice, it is frequently misapplied.

The issue is that dynamic creative can produce ads that are technically personalised but strategically incoherent. If the underlying creative components are not built around a consistent brand narrative and a clear commercial objective, the algorithm will optimise for click-through rate rather than for the outcome that actually matters. You end up with highly personalised ads that drive traffic from audiences unlikely to convert, or that optimise for engagement signals that have no relationship to revenue.

I have judged the Effie Awards, which recognise marketing effectiveness rather than creative awards, and the pattern is consistent. The campaigns that demonstrate genuine commercial impact through personalisation are the ones where the creative strategy and the data strategy were designed together from the start. Not where the data team built an audience model and then handed it to the creative team to execute against.

Personalisation is a strategic discipline, not a production workflow. Treating it as the latter is why so many programmes produce activity without outcomes.

Scaling Personalisation Without Losing Control

At scale, personalisation requires governance structures that most marketing teams are not set up to provide. When you are running personalised experiences across email, paid social, display, and on-site simultaneously, the interactions between those channels become complex quickly. A customer might receive a discount offer via email on the same day they are being retargeted with a full-price ad on social. The left hand does not know what the right hand is doing, and the customer experience reflects that.

The BCG framework on scaling agile operations is useful context here, not because personalisation is an agile problem specifically, but because it highlights the governance structures that allow complex, cross-functional programmes to maintain coherence as they grow. The same principles apply: clear ownership, shared data standards, and a decision-making process that can keep pace with the volume of decisions being made.

Frequency capping across channels is one of the most basic governance requirements and one of the most commonly neglected. Without it, customers in high-intent segments get hammered with messaging from every channel simultaneously, which produces diminishing returns at best and active brand damage at worst. The Forrester research on scaling makes a related point about the organisational structures required to sustain complex programmes, and the same logic applies to personalisation at scale.

Personalisation governance also means having clear rules about what you will not do. Which data signals are off-limits for targeting? At what point does relevance become intrusion? These are not just ethical questions. They are brand questions with commercial consequences. A brand that feels like it is watching you too closely loses the trust that makes personalisation effective in the first place.

The Measurement Problem

Measuring personalisation is harder than it looks. The instinct is to compare the performance of personalised content against non-personalised content and declare a winner. But this comparison is rarely clean. Personalised audiences are almost always self-selecting in ways that inflate apparent performance. If you are personalising based on high-intent signals, the personalised group was always going to convert at a higher rate. The personalisation may have had nothing to do with it.

Proper incrementality testing, where you hold out a statistically valid control group from the personalised experience, is the only way to measure the true contribution of personalisation to commercial outcomes. It is also the approach most commonly skipped, because it requires holding back potential revenue in the short term to generate credible measurement in the medium term. Most organisations do not have the patience for it.

The Vidyard research on pipeline and revenue potential for go-to-market teams touches on a related point: the gap between activity metrics and revenue metrics in digital marketing is significant, and personalisation is one of the areas where that gap is most pronounced. Clicks and opens are not outcomes. Revenue is an outcome. Build your measurement framework around the latter.

If you want to think more broadly about how personalisation fits into a commercial growth strategy, including how to sequence investment across the customer lifecycle and how to connect channel tactics to business outcomes, the Go-To-Market and Growth Strategy hub is the right place to start.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions