What's the difference between CRO and A/B testing?

A/B testing is one method within CRO, not a synonym for it. CRO is the broader discipline: identifying where your funnel leaks, forming hypotheses about why, and running experiments to fix them. A/B testing (splitting traffic between two variants to see which performs better) is the most common experiment type, but CRO also includes user research, session recording analysis, heatmaps, copy audits, and technical performance work. You can do CRO without A/B testing, particularly when traffic volumes are too low to reach statistical significance quickly.

How much traffic do you need before CRO is worth doing?

There is no universal threshold, but a common practical benchmark for running a reliable A/B test is around 1,000 sessions per variant per week. Below that, tests take so long to reach significance that the business context may change before you get a clear answer. That said, CRO is not only A/B testing. User interviews, session recordings, and copy audits are all high-value at any traffic level. For founder-led businesses with modest traffic, qualitative CRO often delivers the most insight per hour spent.

Is conversion rate optimisation only for e-commerce?

No. CRO applies anywhere you have a defined desired action and can measure whether visitors take it. That includes B2B lead generation (demo requests, contact form fills, content downloads), SaaS free-trial sign-ups, professional services enquiries, and booking flows for hospitality or events. The mechanics are the same: set a baseline rate, identify friction, test changes, measure the outcome. E-commerce is the loudest part of the CRO conversation because purchase data is immediate and the volume of transactions makes testing fast, but the principles carry across every vertical.

How long does CRO take to show results?

Quick wins, such as fixing a broken form, clarifying a call-to-action label, or removing a page element that creates confusion, can show results within days. Structured A/B tests on higher-traffic pages typically need two to four weeks to reach statistical significance. Deeper programme work (rebuilding a landing page, restructuring a checkout flow, improving offer clarity) shows meaningful results over one to three months. CRO is iterative: most programmes run a continuous cycle of hypothesis, test, measure, and implement rather than a one-off project.

Can you do CRO without paid traffic?

Yes. CRO works on organic traffic just as well as paid. The source of your visitors does not change the fundamentals: you measure how many complete the target action, identify what is stopping the others, and test improvements. Paid traffic is useful for CRO because you can scale it quickly to hit the volume you need for faster testing cycles, but it is not a requirement. Many effective CRO programmes run entirely on organic search, direct, and referral traffic, particularly in B2B where sales cycles are longer and paid channels are expensive per click.

What is conversion rate optimisation?

The short answer, and then the longer one.

Conversion rate optimisation means improving the percentage of people who arrive at a page or funnel and do what you want them to do. That action can be anything with a measurable outcome: requesting a demo, signing up for a trial, making a purchase, downloading a resource, or sending an enquiry. If you can count the people who arrive and count the people who act, you can optimise the ratio between those two numbers.

The longer answer is that CRO is a discipline, not a single tactic. It combines data analysis, user research, copywriting, design, and iterative experimentation. A CRO programme starts by identifying where in a funnel people drop off, then asks why, forms a hypothesis about how to fix it, tests a change, and measures whether the change moved the number. Repeat. That cycle, applied consistently, is how conversion rates improve over time.

What CRO is not: it is not guessing, and it is not cosmetic redesign. Changing a button colour because someone read that blue outperforms green is not CRO. Running a structured experiment because your session recordings show that a particular form field causes people to abandon is CRO. The difference is evidence and intent.

Why it matters more than most founders think.

Most businesses, when they want more leads or sales, buy more traffic. More ad spend, more content, more outreach. That is one route to growth, and sometimes it is the right one. But it leaves a significant question unanswered: what happens to the traffic you already have?

According to Unbounce's 2023 Conversion Benchmark Report, the median landing page conversion rate across industries is 4.6%. The top quartile of pages convert at 11.45% or higher. That is more than double the median, from the same traffic, to the same audience, often with the same offer. The difference is almost always in how well the page communicates the offer's value and how clearly it directs the visitor to act.

For a founder-led business, this has a direct financial implication. If your current lead page converts at 4% and you get 500 visitors a month, you generate 20 enquiries. If structured CRO work moves that to 8%, you get 40 enquiries from the same traffic budget. You have doubled your pipeline without spending an extra pound on acquisition. That is the case for conversion and funnel optimisation in a sentence.

How conversion rate is actually measured.

The formula is straightforward:

Conversion rate = (conversions ÷ visitors) × 100

If 1,200 people visit a page and 60 complete the target action, the conversion rate is (60 ÷ 1,200) × 100 = 5%.

The one thing worth noting is the denominator. Different tools measure it differently. Google Analytics 4 defaults to events per session, which means a single person visiting a page twice counts as two sessions. Some tools use unique users instead, so the same person counts once regardless of how many times they visit. Neither is wrong, but mixing them across reports creates confusion. Pick one denominator, define it clearly, and stick to it across your tracking.

For most B2B pages, unique users is the more useful denominator because the purchase decision is not made in a single session. A prospect might visit your demo page three times before booking, and you do not want to count that as two failures and one conversion.

You also need to be clear on what counts as a conversion. A form submission and a confirmed meeting are both plausible conversion events, but they measure different things. Form submissions include tyre-kickers; confirmed meetings measure genuine intent. If you are running a CRO programme on a demo request page, track both and watch the ratio between them as well as the absolute conversion rate.

A practical example: the B2B software demo page.

Say you run a B2B software company. Your product homepage sends qualified traffic to a demo request page. That page gets 800 sessions a month. Thirty people fill in the form: a conversion rate of 3.75%.

A CRO audit finds three issues. First, the headline describes what the software does but not what it does for the buyer, so visitors are unclear whether it applies to their situation. Second, the form has seven fields, including company size and annual revenue, which creates friction before the prospect has any reason to trust you with that information. Third, the social proof on the page is from a sector your target buyers do not recognise themselves in.

You run three changes: a headline rewrite that leads with the outcome the buyer wants, a form reduced to three fields (name, work email, and a brief description of their situation), and updated testimonials from companies in the right sectors. After four weeks at sufficient traffic, the conversion rate moves to 6.2%. From 30 enquiries a month to roughly 50, from the same ad spend and the same organic traffic.

None of those changes required new technology or a larger marketing budget. They required someone to look carefully at what the page was actually communicating to the person reading it, and then test whether clearer communication produced a better result. That is CRO.

Where CRO fits in the growth stack.

CRO is most effective when it is applied to a funnel that already has the right foundations. There are two things to check before running experiments on individual page elements.

The first is offer clarity. If your product or service is confusing, or if the positioning is vague, CRO will only improve things at the margin. The page is communicating something; if what it is communicating is not compelling, changing button placement will not fix that. Start with the offer itself: who is it for, what does it do for them, and why now? If you can answer those three questions clearly, you have something worth optimising.

The second is targeting. If the traffic arriving at your page is not well-matched to the offer, conversion rates will be low for the wrong reason. CRO cannot compensate for sending people to a page that was never relevant to them. Check your traffic sources and the intent behind the keywords or audiences driving visits before assuming the page is the problem.

Once offer and targeting are solid, CRO works on the gap between intent and action: the friction, the confusion, the missing reassurance, the unclear next step that stops a willing prospect from converting. That is where structured conversion rate optimisation services have the most to add.

Most businesses have more upside in their existing funnel than they realise. The traffic is often there. The intent is often there. What is missing is a page that does the job of converting it.

What is conversion rate optimisation? A plain-English answer.