Conversion rate optimization is the practice of increasing the share of visitors who take a meaningful action. Its appeal is economic: acquisition costs recur, whereas a conversion improvement applies to every visitor who arrives afterwards. That is why CRO usually returns more per pound than an equivalent increase in media spend.
Start with evidence, not opinion
The most common failure in CRO is redesigning on instinct. A team dislikes a page, rebuilds it, and the number moves for reasons nobody can identify. Six months later there is no learning to carry forward.
- Analytics: where do sessions end, and on which step of which flow?
- Form analytics: which field causes abandonment, and how long do people hesitate on it?
- Session behaviour: where do people scroll back, hover without clicking, or rage-click?
- Enquiry quality: are you converting more people who cannot buy?
- Sales conversations: what objection recurs, and does the page address it before the call?
The last two matter more than most CRO advice admits. A change that increases form fills while lowering close rate has made the business worse and the dashboard better.
The friction inventory
Most conversion problems fall into a small number of categories, and naming them makes the fix obvious.
- Unclear value: the visitor cannot tell within seconds what this is and whether it is for them.
- Unclear next step: several competing calls to action, or none that is obviously primary.
- Unearned commitment: asking for a phone number before demonstrating any value.
- Missing proof: a claim the visitor has no reason to believe at the moment they must decide.
- Unanswered objection: the question that stops the purchase is answered nowhere on the page.
- Mechanical friction: slow loading, an awkward form on mobile, a broken validation message.
Testing honestly
A test needs enough traffic to distinguish a real effect from noise. Many pages do not have it, and pretending otherwise produces confident conclusions from random variation. If a page receives a few hundred visits a month, an A/B test will not resolve a modest difference in any reasonable timeframe.
- Concentrate tests where volume makes a result readable — usually a handful of pages carry most conversions.
- Decide the sample size and duration before starting, and do not stop early because the numbers look good.
- Change one meaningful thing per test, so a win is attributable.
- On low-traffic pages, apply judgement and best practice, and record it as judgement rather than evidence.
What to fix first
Sequence by traffic multiplied by drop-off. The highest-volume step with the steepest loss is where a fix pays back fastest. In practice that is usually the primary landing page, the form itself, or the mobile experience of both.
Mobile deserves separate attention rather than a responsive assumption. A form that is merely tolerable on a laptop is often genuinely difficult on a phone, and for many businesses the majority of first sessions arrive there.
Measuring what matters
Conversion rate alone is a weak metric because it is trivially gamed by lowering the bar. Track it alongside lead quality and downstream close rate so an improvement in one is visible against any deterioration in the others. Attribution helps here, with the caveat that multi-touch models apportion credit using assumptions rather than measuring truth.