What is conversion rate optimisation?

Conversion rate optimisation, usually shortened to CRO, is the practice of increasing the percentage of visitors who complete a desired action on your site — most often a purchase.

It improves the return on traffic you already have instead of buying more of it. That is the whole idea, and it is why conversion is upstream of nearly everything else: change it and you change what every other channel can afford to pay.

How to calculate conversion rate

Divide conversions by total visitors, multiply by 100.

200 orders from 10,000 visitors is a 2% conversion rate.

Measure it per channel as well as overall. Paid, organic, email and direct traffic convert at very different rates, and a single blended number hides which one is failing. An overall figure of 2% could be email converting at 8% and paid converting at 0.4%, which is a completely different problem from a site that converts evenly.

What counts as a good conversion rate

Most e-commerce sites land between 1% and 3%.

That range is close to useless as a target, though, because it spans categories that are nothing alike. A site selling $2,000 furniture and a site selling $15 socks face entirely different decisions. Price point, category, traffic source and whether the buyer has heard of you all move the number more than any design decision.

The benchmark worth tracking is your own, over time. Industry averages mostly tell you how much company you have.

Why it matters more than it sounds like it should

Run the arithmetic once and it becomes hard to unsee.

Say your site converts at 1%. You pay $50 for a visitor’s click, your average order is $80, and your gross margin is 40%.

A hundred visitors costs $5,000. One of them buys. You collect $80 and keep $32.

You turned $5,000 into $32.

Now run it at 2%. Same traffic, same spend, same product. Two people buy. You keep $64.

Nothing changed except what happened after the click. The auction did not move, you did not find a better audience, you did not spend more. And you will keep collecting that difference on every dollar you spend from now on.

This is why buying more traffic so often makes a business worse rather than better. Scaling a funnel that leaks scales the leak. You end up larger, with a thinner margin, wondering why the ad account stopped working when it never really started.

What the work actually involves

A decade of tool marketing has convinced people CRO means A/B testing button colours. On the sites we work on, that is not where the gains are.

A/B tests measure small changes and need serious traffic to reach reliable results. On most e-commerce sites a button-colour test takes weeks to reach significance and then fails to replicate.

The wins that move a business are structural — what a page leads with, the order it answers questions in, what the offer looks like at the moment of decision. Those are not tests. They are rebuilds, informed by watching where real people stall.

Nocs Provisions sold high-end binoculars to people who go outdoors. The site led with atmospheric lifestyle photography — beautiful, on brand, and wrong. Optics buyers want specifications, comparisons, and confidence they are not wasting $300. The people most likely to buy were bouncing first, because the page answered a question they had not asked and skipped the ones they had.

Restructuring the decision path took them from 0.8% to a 1.5% baseline. An 87% lift, no A/B test involved.

The four questions, in order

Most underperforming pages answer these out of sequence, or skip one:

1. What is this? Within a couple of seconds, in plain category terms. Clever headlines routinely fail here, because cleverness needs context the visitor does not have yet.
2. Is it for someone like me? People need to recognise themselves. This is what specificity and social proof actually do.
3. Why should I trust you? Reviews, guarantees, real photography, named humans, an address.
4. What happens if I am wrong? Returns, warranty, sizing help, how to reach a person. Skip this and you lose everyone risk-averse, which in most categories is the majority.

The other half nobody looks at

Conversion rate is one half of the equation. Average order value is the other, and it is usually easier to move.

Boardies sells swimwear. People rarely buy one item — they buy for a holiday, a family, a season. The site was built for single-item purchases, so every order had to recover its full acquisition cost alone.

Restructuring merchandising around how people actually shop the category lifted average order value 52% on an unchanged catalogue. No new products, no new photography.

A 50% AOV lift does the same thing to your unit economics as a 50% conversion lift. It is frequently less contested, and almost nobody looks there first.

How much traffic you need

Less than people assume, provided you are not running formal tests.

Statistical testing needs volume — often thousands of conversions to detect small differences reliably. Diagnosing why a decision path fails and rebuilding it does not. MOR Collections lifted conversions 200% in seven days on a partial implementation.

If you have enough traffic to feel the problem, you have enough to fix it.

Where it sits in the sequence

The order that works, nearly every time:

1. Fix conversion. It changes what every click is worth.
2. Raise order value. It changes it again.
3. Then scale acquisition. Now the same ads that lost money make money.

Nocs is the cleanest illustration of the cost of getting this backwards. Conversion went 0.8% to 1.5% first. Only after that did the ad account move from 0.5X to 4X — on spend that had been unprofitable for months. The ads did not start working. The destination did.

Run steps two and three before step one and you are scaling a leak. It looks like growth for a quarter. Then you are a bigger company with a worse margin.