News & Insights

The reality of personalisation: why most programmes underperform 

Written by Stephen Gillespie | Dec 24, 2025 9:14:59 AM

Personalisation isn’t a feature. It’s a discipline.  

Enterprise platforms offer to deliver “highly personalised experiences at scale.” 

It sounds persuasive. But leaders who’ve tried to make it work will know: without foundations in the organisation, personalisation won't deliver the business advantage you're after. 

This is in no way because the technology isn’t capable, but because the organisation isn’t ready to deliver.

The uncomfortable fact is this: when personalisation fails, it's not at the point of execution. It fails because it doesn't deliver relevance, in the moment, based on need. 

The three things you can’t skip 

1. Structured, reusable content 

You can’t deliver the right experience without the right content. 

That means: 

  • Relevant content variants, not just a single master asset.
  • Intended journeys that map to intent, alongside Metadata and titles that match.
  • A publishing model that supports agility. 

Without this, all the segmentation in the world won’t move the needle. 

 

2. Segments built on behaviour 

“First-time visitor” or “UK audience” is personalisation, but unless it's tied to a business goal or journey, it’s really just labelling. 

Relevance comes from recognising intent: 

  • What are they trying to do right now, and why? 
  • How urgent is it? 
  • What behaviour signals can we attribute? 

The shift from static demographics to dynamic signals is where personalisation can become more powerful if you have the traffic volumes to support this model. 

3. An experimentation culture

Personalisation without testing can leave you working purely with guesswork.

Ideally, you'd construct a hypothesis and test against it. Then, if it fails, in our ideal scenario you'd have a culture that accepts valuable learning and refines the hypothesis.

That's the hardest part - not the tooling, but creating a mindset where the organisation wants to learn fast, rather than prove itself right. 

Why most brands start too soon

Does this sound familiar? 

  • A platform is introduced.
  • Someone says “let’s try some personalisation.” 
  • A few rules are created. A banner gets swapped. 
  • A box gets ticked. 

Six months later, the whole thing hasn't moved the needle and there is an absence of fresh ideas. 

Why? Because the content plan wasn't configured around intent. There was no meaningful segmentation that aligned intention to those segments.

And fundamentally, there's no culture of testing.

Having a feature enabled isn’t the same as delivering relevance. 

Where Optimizely helps

When the foundations are there, Optimizely is one of the strongest platforms we’ve worked with. 

Because it doesn’t bolt testing on as an afterthought - it’s embedded. And because it treats audience logic and structured content as core, not add-ons. 

That combination gives teams the ability to start small, learn quickly, and scale confidence alongside capability and achieve results. 

Relevance is earned 

Done well, personalisation can transform your digital performance.

Done badly, it’s unfortunately little more than a gimmick. 

It’s not about “surprising and delighting” or 'being creepy'... It’s about building relevance, in the moment, based on real need. 

If your content isn’t structured, your signals aren’t meaningful, and your teams aren’t ready to learn - you're not really personalising.

Pause. Build the groundwork. Then personalise.