Conversion Rate Optimisation: The Last Ten Metres

Most marketing budgets are spent getting people to the site. Almost none is spent on what happens when they arrive.

That's the strange economics of the average growth plan: enormous effort on the first thousand metres of the journey, near-total neglect of the last ten, where the buyer either acts or leaves. Conversion rate optimisation is the discipline of fixing the last ten metres.

What it is, in one breath

Your conversion rate is the percentage of visitors who do the thing you built the page for: enquire, book, buy, subscribe, apply. Conversion rate optimisation (CRO) is the practice of raising that percentage through evidence: finding where buyers hesitate, changing what's in their way, and measuring whether the change actually worked.

That last clause is the whole discipline. Without measurement it isn't optimisation, it's redecoration.

Why it beats buying more traffic

The maths of CRO is the most persuasive in marketing. Every improvement to conversion applies to all the traffic you already have and everything you buy afterwards, which means it compounds while ad spend just repeats.

Raise the conversion rate and every channel you run gets cheaper on the same day: the ads you pay for return more per click, the SEO traffic works harder, the email list converts better, all without a dollar more at the top.

And the alternative keeps getting worse, from two directions at once. Ad prices rise nearly every year, so a business with a leaking experience pays more each quarter to pour water into the same holed bucket.

And AI answers now settle a growing share of searches before anyone clicks through, which means the visits that still arrive are scarcer than they were, and worth more. When traffic stops being abundant, converting the traffic you get stops being optional.

More traffic won't fix a broken experience. Increasingly, more traffic isn't even on offer.

Start from the number, not the page

Before touching a single page, do the arithmetic that most CRO work skips. Take the revenue the business needs, divide it by what an average customer is worth, and you have the number of customers required. Divide that by your close rate and you have the leads required.

Now compare it with the leads the site currently produces. That gap is the brief. It tells you whether you need a conversion lift, a traffic lift or both, and it turns "improve the website" into a number someone can be accountable for.

Optimisation without a target is a hobby with a dashboard.

What it isn't

CRO has a reputation problem, and it earned it. A decade of "change the button colour" content reduced the discipline to superstition with a spreadsheet. So, to be clear about what serious conversion work is not: it isn't cosmetic tweaks, it isn't dark patterns that trick people into clicking, and it isn't a debate your team settles in a meeting.

The loudest voice in the room shouldn't decide what goes on the page. The best page should, and the only way to know which page is best is to let real buyers vote with their behaviour.

It's also bigger than the website. Buyers judge everything: the site, the form, the speed of the follow-up, the first weeks after they say yes.

A conversion is not a click, it's a decision, and decisions are shaped by the whole experience. Optimising the page while the follow-up takes three days is polishing one link of a chain that snaps somewhere else.

Where conversions actually die

After enough audits, the same culprits keep appearing. Forms that ask for too much, too soon, so buyers get most of the way there and vanish. Pricing pages that hide the one fact every serious buyer came for. Landing pages that describe the business instead of answering the buyer's question.

Sites that look fine and say the right things but never ask anyone to do anything. Speed, because every second of loading bleeds intent. Missing trust signals at exactly the moment of commitment, when the buyer is looking for a reason to feel safe and finds a stock photo instead. And mobile experiences that were checked once, on the designer's phone, in perfect light.

None of these are exotic. That's the point. Most conversion problems are ordinary hesitations that nobody has walked through the buyer's eyes to find. And they're worth finding in order of leverage: the pages where money decides, pricing, key landing pages, forms and the calls to action between them, before the pages where nobody was deciding anything anyway.

The process, honestly

The serious version of CRO is a loop, not a project. Walk the buyer's full path and record every hesitation, dead end and drop-off. Rank what you find by revenue impact, not by how easy it is to fix, because effort spent on a page nobody visits is effort donated to nobody.

Form a hypothesis about why buyers stall, grounded in what you can observe rather than what the team assumes. Ship the change as an experiment with a defined measure of success. Keep what wins. Remove what loses. Go again.

Two words of caution about that loop. First, it never finishes, and that's a feature: buyer behaviour drifts, competitors move, and last year's winning page quietly becomes this year's control to beat.

Second, the loop runs on evidence, which means the tools that show you real behaviour, session recordings, heatmaps, funnel and form analytics, matter more than the tools that show you opinions. Watch what people do, not what they say.

The part the testing vendors won't tell you

Here's the honest bit that's missing from most CRO content, usually because the people writing it sell testing software. Proper A/B testing needs volume. Below a certain amount of traffic and conversions, a split test takes months to reach a trustworthy answer, and most tests quietly get called early on noise.

Plenty of Australian businesses simply don't have the numbers for it, and nobody selling a testing platform is rushing to say so.

That doesn't put CRO out of reach. It changes the toolkit. With modest traffic, you lean on behaviour evidence instead of split statistics: recordings and heatmaps to find the friction, form analytics to find the field where buyers give up, disciplined before-and-after measurement with a defined window, and bigger, bolder changes rather than micro-tweaks, because small differences need big samples to detect and big differences don't.

The principle survives at every traffic level: evidence over opinions. Only the instruments change.

The 2026 layer

Two things have changed recently that most CRO guides haven't caught up with. The first is personalisation that's actually practical now: a first-time visitor, a returning buyer and an existing customer shouldn't be handed the same page, and on a modern stack they don't have to be.

The second is stranger: some of your visitors aren't human anymore. AI agents now arrive on behalf of buyers to extract facts, compare options and shortlist suppliers, and they convert or discard you based on structure, not persuasion.

Optimising the experience now means optimising it for both kinds of visitor, which is a discipline of its own. We've built a whole Program around that second one.

Measure what matters

One last discipline: a single site-wide conversion rate is a vanity number. It mixes buyers with browsers, mobile with desktop, first visits with fifth ones, and it moves for reasons that have nothing to do with your changes. Measure conversion where decisions happen: by page, by traffic source, by intent.

And track the small commitments that lead to the big one, the pricing page visit, the started form, the return visit, because the path to a conversion is a series of smaller yeses, and each one can be optimised on its own.

The quiet truth underneath all of this: your next customer is already on your site. They arrived this week, hesitated somewhere specific, and left without telling you why.

CRO is the work of finding that spot and fixing it, which is exactly what our Conversion & Experience Program exists to do. The traffic is paid for. Make the last ten metres worth it.