Conversion rate optimisation teams talk a lot about friction. But most conversations focus on the wrong kind.

Slow load times are friction. Complicated forms are friction. Buried CTAs are friction. These are real problems, and they’re worth fixing. But they’re also the easy problems — the ones you can measure, A/B test, and iterate on with data.

The harder friction is harder to see.

The friction you can’t see in your analytics

When someone lands on your page, reads the headline, and leaves — your analytics record a bounce. They don’t record why.

Was it the wrong audience? Probably not — they clicked your ad or found you through search. Was it a slow page? Maybe. But more often than not, it was something subtler:

  • The value proposition was clear to you, but not to them
  • The CTA answered the wrong objection
  • The proof points addressed the wrong concern
  • The page felt designed for a different kind of buyer
Interpretive friction The page works technically — it loads fast, the form is simple, the CTA is visible. But it fails to land with the specific person reading it. The gap isn't in the UI. It's in the meaning.

This kind of friction doesn’t show up in your analytics. It shows up in your conversion rate.

Why buyer role matters more than demographic

Most teams segment their audience by job title, company size, or industry. These are useful proxies. But they miss something important.

The same person — same title, same company, same stage in the buying journey — will read your page completely differently depending on their role in the decision.

One page. Four completely different experiences. And the copy written for one of them is almost always the copy that loses the other three.

Buyer roleWhat they’re askingWhat they need from your page
Economic buyerIs this worth the risk?ROI framing, credibility signals, urgency
Technical evaluatorDoes this actually work?Integration details, security posture, implementation clarity
ScepticWhat’s the catch?Specific proof, honest limitations, third-party validation
High-intent prospectWhat do I do next?Clear CTA, fast path to action, no friction before the next step

Each role reads the same copy through a different lens. A headline that reassures the economic buyer can read as vague marketing to the sceptic. A proof section that satisfies the technical evaluator can feel irrelevant to the high-intent prospect who already knows it works.

What low-friction copy actually looks like

Low-friction copy isn’t jargon-free. It isn’t short. It isn’t written in plain English.

Low-friction copy meets each reader where they are. It answers the question each buyer type is actually asking. It reduces the gap between what they understand about your category and what you’re offering.

That’s harder to write than most teams think. And it’s harder to test than most analytics tools allow.

Pre-launch audience testing — with agents shaped around your actual buyer types — is one of the few ways to close that gap before campaign spend is on the line.

Frequently asked questions

What is conversion friction?

Conversion friction is the gap between what your page says and what a specific reader needs to hear before they will act. It is rarely about page speed or form length; more often it is an unanswered question, a claim without evidence, or a price the reader cannot find.

Why doesn’t my analytics show conversion friction?

Analytics records what happened, not why. It can show you that people left a page, but not that they left because the proof on it came from a role they do not recognise. That reasoning is the part you have to surface some other way.

What are the main types of conversion friction?

We publish a named taxonomy of twelve: category ambiguity, value opacity, unproven superlative, proof mismatch, missing metric, price opacity, price-anchoring failure, gated tier, implementation opacity, verification gap, CTA hesitation and commitment cliff. Each is defined with its fix in the Friction Index.

How do you find friction before you have any traffic?

You cannot measure it, so you have to simulate or borrow judgement. Pre-launch conversion testing covers the four realistic options, including their limits.