Personalisation has become one of those words that gets used so freely it's stopped meaning very much. Ask most ecommerce teams what their personalisation strategy looks like and you'll hear about first-name tokens in subject lines, a welcome flow, maybe a win-back campaign for anyone who hasn't bought in 90 days. That's not personalisation. That's scheduling.
The distinction matters because those two things produce very different results, and once you've seen what connected, behaviour-led personalisation actually does to conversion rates and customer lifetime value, the gap is difficult to ignore.
Why static segments stop working
A segment is a description of who a customer was at a specific point in time based on a set of criteria. "Purchased once." "High spender." "Lapsed." These labels were accurate when you applied them. They may not be accurate now.
The problem is that many ecommerce CRM strategies are built around these fixed categories. Customers get placed into a bucket at acquisition and the experience they receive is determined by that bucket, sometimes indefinitely. A customer who bought a single product eighteen months ago and has since visited the site four times, read three emails, and added something to their basket twice is being treated identically to someone who genuinely hasn't engaged since that first purchase.
This is a structural problem. The way personalisation has been set up doesn't have a mechanism to respond to them.
The same issue plays out on-site and in paid media. A returning customer sees the same product grid as someone visiting for the first time. A paid retargeting ad shows a product someone bought last week. An email campaign goes out to the entire database with identical messaging because the segmentation logic hasn't been updated since it was first built.
When acquisition, conversion, and retention each operate with their own definition of who a customer is, the experience fragments. And fragmented experiences don't convert as well as coherent ones.
What changes when behaviour drives the experience
The shift we're describing isn't about which tools you use. It's about what data you use, when you use it, and whether that data is shared across the customer journey or locked inside a single channel.
Take the post-purchase email sequence. A standard flow sends email one, then email two, then email three – regardless of what the customer did after receiving email one. A behaviour-led sequence asks different questions. Did they open it? Did they click through? Did they return to the site? Did they buy something else? Each of those outcomes points toward a different next step, and the customer experience should reflect that.
This is exactly the kind of logic we applied when we worked with REESON Beauty ahead of their Black Friday period in 2025. Rather than deploying a single campaign to the whole database, we built a strategy around lifecycle stage and engagement signals – the 60-day engaged segment received different messaging to recent sign-ups, and highly engaged non-purchasers received content focused on product value rather than promotion. Flows were rebuilt ahead of the season specifically so new subscribers entering during the peak period entered the Black Friday narrative immediately rather than receiving a generic welcome sequence.
CRM-attributed revenue increased by +166% year-on-year. Campaign revenue was up +283%. Those numbers came from segmentation and timing, not from a bigger budget or a new tool.
The channel gap nobody talks about enough
Most conversations about personalisation stay within a single channel. Email personalisation. On-site recommendations. Paid retargeting. Each one is treated as its own problem to solve.
The more useful question is what happens when the signals from each channel inform the others.
Your CRM data tells you who your best customers are: their purchase frequency, the categories they return to, the price points they're comfortable with. That information should be shaping your paid media audiences, not sitting unused in Klaviyo while your Meta campaigns run on demographic targeting alone. Your on-site browse behaviour tells you what someone is considering right now. That should be feeding into what they see in their next email, not getting lost because the two systems don't talk.
For Lily Charmed, connecting audience signals more tightly across Google and Meta (and aligning campaign structure to actual purchase intent rather than broad demographic reach), produced a +105% increase in ROAS on Google and a +36.94% increase in checkouts initiated on Meta. The spend didn't change materially. What changed was the relevance of who was being reached and when.
This is the argument for treating personalisation as a cross-channel strategy rather than a channel-specific tactic. When acquisition, conversion, and retention share the same customer data, each stage gets sharper.
On-site: where personalisation often has the most immediate impact
Email and paid media get most of the attention in personalisation conversations. The on-site experience often doesn't, which is worth examining, because it's where the customer is closest to making a purchase decision.
A returning customer who has bought from a specific category twice doesn't need to be shown the same homepage as a first-time visitor. A customer who has had a product in their basket for three days doesn't need a generic promotional banner. A customer who consistently buys at a certain price point is being poorly served by recommendations that sit significantly above or below it.
None of this requires a major rebuild. When we worked with Daily Skin on their CRO strategy, the improvements were targeted and incremental: reorganising site architecture, adding a free shipping progress bar in the cart, introducing a Quick Buy option on product listing pages, integrating verified reviews through Judge.Me. The conversion rate went up 91%. Add-to-cart rate increased by +206%. These weren't dramatic changes to the site. They were changes to how the site responded to what customers were actually doing.
Where to start
The most practical question to ask isn't which personalisation technology to invest in. It's where in your customer journey the experience is most static relative to what your customers are actually doing.
For most ecommerce brands, that's one of three places: the CRM flows that haven't been updated since they were first built; the on-site experience that treats every visit the same regardless of history; or the gap between paid acquisition and what happens when someone arrives on the site.
Pick the one where the disconnect is sharpest. Build a hypothesis about what a more responsive experience would look like. Test it, measure it, and let the data tell you what to do next.
This is a way more useful starting point than asking which AI features your email platform offers. If you'd like a second pair of eyes on where that disconnect sits in your business, we're happy to talk it through.
