What Is Conversion Rate Optimisation?
Conversion rate optimisation is the deliberate, ongoing work of improving your website so that more visitors take the action you want, most often completing a purchase, without needing a single extra visit. It is a different lever from marketing that brings people to your site in the first place, which our guide to ecommerce marketing covers in full. Where ecommerce marketing gets people to the door, CRO decides how many of them buy once they are inside.
Picture a small candle shop selling online, getting 3,000 sessions a month at a 1.8 percent conversion rate, so 54 orders. Push that rate to 2.4 percent through better product photography, a clearer returns policy and a simplified checkout, and the same 3,000 sessions now produce 72 orders, a third more revenue with no extra spend on adverts. That is the entire case for CRO in one sentence. You are not trying to find new customers. You are trying to stop losing the ones who already found you.
How Do You Calculate Conversion Rate?
The basic formula is orders divided by sessions, multiplied by 100. If your shop had 5,000 sessions last month and 100 of them ended in a completed order, your conversion rate is 2 percent. The detail that trips owners up is deciding which action counts as a conversion. Some shops track completed purchases only, others also track adding an item to the basket or starting checkout, which is useful for spotting where people drop off but should never replace the purchase number when you are judging whether the shop is working.
It also helps to know that conversion rates are not a single, universal number you should measure yourself against blindly. They vary widely by sector, by device and by average order value, which is why industry benchmark trackers such as Smart Insights publish figures broken down by category rather than one headline average. A shop selling a £15 impulse purchase will convert differently to one selling £400 furniture, and comparing the two tells you very little. The more useful comparison is your own shop against its own past performance, month on month.
Why Do Online Shoppers Abandon Their Carts?
Here is the sobering context for anyone running an online shop. The Baymard Institute, which aggregates dozens of separate studies on the subject, puts the average documented online shopping cart abandonment rate at 70.19 percent. Roughly seven in ten shoppers who put something in a basket leave without paying for it. That is not a sign your shop is broken. It is the normal, well-documented state of online retail, and it is exactly the gap CRO exists to close.
The same Baymard research names some of the most commonly cited reasons shoppers give up: unexpected extra costs appearing at the final step, being forced to create an account before they can pay, and a checkout process that simply feels longer and more effortful than it needs to be, especially on a small mobile screen where every extra field is more annoying than it would be on a laptop. None of these are exotic problems. A shop that shows delivery costs early, offers a guest checkout, and keeps the payment form to as few fields as it can get away with is already closing a meaningful part of that gap before it runs a single test.
How CRO Fits the Way Shoppers Decide
CRO works best when it respects two things that are true about how people buy, and both point in the same direction. The first is that visitors arrive at your product pages at very different stages of readiness. Someone who found your shop through a broad search is often still working out whether they even have the problem your product solves, while someone who typed your brand name directly into Google already knows what they want and is one click from paying. A product page that tries to sell to both groups with the same message usually persuades neither. The fix is to match the page to the visitor, giving the curious browser reassurance, detail and social proof, and giving the ready buyer a fast, unfussy route straight to checkout.
The second is that friction is the enemy of a sale in exactly the same way that a hard-to-reach shelf is the enemy of a sale in a physical shop. Every extra click, every unexpected form field, every moment of doubt about whether a size will fit or a delivery will arrive on time, is a small tax on the sale. A shop that removes that friction, clear sizing guides, an obvious returns policy, delivery dates stated plainly, is not doing anything clever. It is simply making itself easier to buy from, which is worth more than almost any clever bit of persuasive copy. A homeware shop that added a simple "fits a standard UK doorway" note to its furniture listings, after noticing size questions kept arriving by email, saw fewer abandoned checkouts within weeks, because it had removed a doubt rather than argued a customer out of it.
Where Should a Small Online Shop Start with CRO?
With limited time, the order you tackle things in matters. Start with checkout, because it is the last hurdle and the one Baymard's research shows most people fall at. Strip out unnecessary account-creation steps, show the full cost including delivery before the final page, and make sure a guest can pay without registering.
Next, look at the mobile experience specifically, since most shopping journeys now start and often finish on a phone, and a checkout that feels merely inconvenient on a laptop can feel genuinely unusable on a small screen. Then turn to your product pages themselves: photographs from more than one angle, honest descriptions of size and material, and reviews from real customers, since a shopper deciding between you and a stranger's shop is looking for exactly this kind of reassurance.
Online retail is not a fringe channel to treat as an afterthought either. The Office for National Statistics tracks internet sales as a share of total UK retail spending every month, a reminder of how central online shopping now is to UK retail as a whole, and how much is riding on getting the buying experience right rather than treating a website as a static brochure. A clear marketing strategy is what tells you which of these fixes matters most for your own shop, rather than guessing at all of them at once.
A Practical Way to Run a Conversion Rate Optimisation Test
You do not need expensive software to start, and you should not try to fix everything simultaneously. A workable approach looks like this.
- Pick one page and one problem. Use your analytics to find the page where the most visitors drop off, often the checkout or a specific product page, rather than guessing.
- Form a clear hypothesis. Write down what you think is causing people to leave and what change you expect to fix it, so you can judge the result afterwards.
- Change one thing at a time. Test a single variable, a shorter form, an earlier delivery-cost display, a clearer size guide, so you know what caused any change in the number.
- Give it enough time and enough visitors. A test judged after two days and forty visitors tells you almost nothing. Let it run long enough to cover a normal week of trading before drawing a conclusion.
- Keep the winner and move on. If the change helps, make it permanent and move to the next problem. If it does not, revert it and try a different fix, since not every idea works.
This kind of testing works best as a habit rather than a one-off project, which is exactly the sort of recurring activity a marketing plan should schedule in, a standing slot each month to review one page rather than a single frantic week before it gets forgotten.
What Mistakes Undermine Conversion Rate Optimisation?
A handful of habits waste the effort. The first is testing several changes on a page at once, so that when the number improves, nobody knows which change caused it and the lesson is lost. The second is stopping a test too early because early numbers looked encouraging, when a handful of visitors is not a fair sample and the real trend often looks different a week later. The third is fixating on desktop performance while most shoppers are on a phone, since device-level benchmark trackers consistently show mobile shoppers behaving differently to desktop ones, often converting at a noticeably lower rate purely because the experience is more fiddly on a small screen. The fourth is treating CRO as a single project to finish rather than a habit to keep, when the shops that improve fastest are the ones running a small test every month rather than a big overhaul once a year.
How Do You Measure Whether CRO Is Working?
Track your conversion rate monthly, alongside two companion numbers that give it context: average order value, since a higher conversion rate paired with a falling order value can cancel itself out, and revenue per visitor, which combines both into one figure that tracks the true health of the shop. The wider UK retail picture published in the ONS's monthly retail sales bulletins is worth keeping half an eye on too, since it shows how much of any change in your own numbers is the market moving and how much is genuinely down to your own improvements.
None of this needs to be complicated. A simple spreadsheet with one row per month, tracking sessions, orders, conversion rate and average order value, will show you the trend clearly enough to know whether the changes you are making are working, which is the entire point of doing this in the first place.
