Your traffic is fine. Maybe it's even good. But the number of people actually turning into leads, demos, quote requests or sales hasn't moved in line with it.
You've probably had this conversation internally more than once: "We're getting the visitors, why aren't we getting the enquiries?"
So someone says the words "we need to do some CRO (conversion rate optimisation." And the moment you start asking what that actually costs, you get answers ranging from a few hundred quid for a landing page tweak to five-figure quotes for a "full conversion programme" - for what sounds, on the surface, like the same job.
Let's cut through it.
Conversion rate optimisation for a B2B website typically costs somewhere between £3,000 and £15,000 for a one-off project, or £2,000-£6,000+ a month if you're running it as an ongoing programme - depending on how much research, testing and content sits behind the work.
For most mid-sized B2B businesses running a structured CRO project (audit, a handful of new landing pages, testing infrastructure, a quarter of iteration), you're realistically looking at £8,000-£20,000 for the initial phase, followed by an ongoing spend if you want the gains to compound rather than plateau.
That's the headline. Now let's talk about why the range is so wide, what you're actually paying for, and why treating CRO as a one-off "fix the homepage" job is one of the most common ways businesses waste the traffic they've already paid to acquire.
This is where a lot of budgeting goes sideways, because "CRO" gets used to describe a few genuinely different pieces of work.
You take a defined set of pages - typically your homepage, a handful of service or product pages, and your highest-traffic blog content - and improve how they convert. New layouts, sharper copy, better calls-to-action, maybe some A/B testing on specific elements. It has a start and an end date.
You treat conversion as something you continuously test, measure and improve, the same way you'd continuously optimise paid ad spend. New hypotheses each month, new experiments, a backlog of pages to work through, and reporting that shows what's actually moving the needle.
This is less "change the button colour" and more "rebuild the buyer journey" - persona-led content, nurture sequences, and landing pages built around what your ideal customer actually needs to see before they'll fill in a form. It sits closer to inbound strategy than to a technical testing programme.
This is something we complete with every website project - after all, that's how we ensure that a website starts its life as a part of your sales team and converts, as opposed to a brochure of your services.
Almost every CRO conversation we have starts as Option 1 and, if it works, turns into Option 2 or 3 - because a single round of landing page tweaks rarely explains why conversion was low in the first place.
We hear a version of this constantly from prospects before they come to us:
"We're getting good traffic and posting some great content, but the leads just aren't converting as much as we'd like."
That's the exact starting point of most CRO engagements. The traffic isn't the problem. The journey once someone arrives is.
Here's how the money typically splits across a structured CRO project, using a mid-market B2B engagement as the example:
Cost Area |
What it Covers |
Typical % of Budget |
| Research & Audit | Analytics review, heatmaps/session recordings, funnel analysis, competitor benchmarking | 10-20% |
| Persona & Journey Work | Buyer journey mapping, messaging by funnel stage, qualification criteria | 10-15% |
| Copywriting | Headlines, page copy, CTAs, forms, email nurture where relevant | 10-15% |
| Design | Landing page templates, wireframes, visual treatment of key journeys | 15-20% |
| Build | Page build, form logic, personalisation modules, tracking setup | 20-30% |
| Testing | A/B/multivariate testing setup, heatmap and session recording tools | 5-10% |
| Reporting & Iteration | Results analysis, next-round hypotheses, ongoing recommendations | 10-15% |
Notice that "build" - the bit most people picture when they think of CRO - is rarely more than a third of the actual cost.
The research and journey work at the front end is what determines whether the changes you make are the right ones, and it's also the part that gets skipped most often when budgets are tight.
Engagement |
What's Included |
Typical Cost |
| Single Landing Page Optimisation | One page, copy + design refresh, basic A/B test | £1,500-£4,000 |
| Sitewide CRO Audit & Roadmap | Full-site audit, prioritised recommendations, no build | £3,000-£8,000 |
| Structured CRO Retainer | Audit, persona work, 3-6 new/optimised landing pages, testing setup | £8,000-£20,000 |
| Ongoing CRM Retainer | Monthly testing, iteration, reporting, new asset creation | £2,000-£6,000/month |
If your number sits well outside these ranges, ask why.
Too cheap, and the research and testing phases are probably being skipped - which means you're paying for a redesign with a CRO label on it, not an actual optimisation process.
Too expensive, and you're either being quoted for a full site rebuild dressed up as CRO, or the agency's testing infrastructure is more elaborate than your traffic volume actually justifies.
This is the one people underestimate most. A/B testing needs volume to reach statistical significance.
If a page gets 200 visitors a month, running a formal split test on it could take the best part of a year to produce a reliable result - which means the spend is better directed at qualitative research (heatmaps, session recordings, user feedback) and considered, evidence-led changes rather than a live testing programme.
The busier your funnel, the faster and more defensible formal testing becomes - and the more a testing-heavy retainer earns its cost.
Optimising your homepage and one landing page is a very different job to optimising the six or seven pages that sit across your entire buyer journey, from first blog visit through to demo request.
Every additional funnel stage is another set of hypotheses, another round of copy and design, and another thing to measure.
Formal A/B testing needs infrastructure.
If you're on HubSpot's Marketing Hub Professional or Enterprise tiers, some of this is built in. If not, tools like heatmap and session recording software, or dedicated testing platforms, typically run anywhere from a couple of hundred to a couple of thousand pounds a month, on top of the agency or freelance fee for actually using them properly.
"Show a different headline to visitors from manufacturing companies versus visitors from membership organisations" sounds like one line in a brief.
In practice it's dynamic content logic, list segmentation, and additional copy variants - all of which add build time and, usually, additional software cost.
If your existing pages are structurally fine but the copy doesn't speak to what your buyer actually needs to hear, the cost sits more in copywriting and messaging strategy than in design or development.
If the pages themselves are the problem - slow, cluttered, badly structured - the cost shifts towards design and build.
There's a meaningful cost difference between "we ran a formal A/B test and know which version wins" and "we made the change we were fairly confident about and moved on."
Genuine testing takes longer, costs more, and needs enough traffic to be worthwhile - but it removes the guesswork. Plenty of good CRO work happens without formal testing, particularly on lower-traffic pages, but it's worth knowing which one you're paying for, because the two are priced differently for good reason.
We've written before about the total cost of ownership question when it comes to websites generally, and CRO has its own version of the same problem: the headline project fee is rarely the whole number.
Heatmap tools, session recording software, and dedicated A/B testing platforms are usually billed separately from the agency fee, and they don't stop once the initial project ends if you want testing to continue.
Every test variant needs to be built, not just designed. A "simple" three-variant headline test still means three sets of copy reviewed, three builds, and someone checking none of them break on mobile.
You can't optimise what you can't measure properly. Getting goal tracking, form submissions and attribution set up correctly - so that "conversion rate" actually means something reliable - is often underestimated, particularly on sites where tracking has been patched together over several years by whoever happened to be available at the time.
Someone on your side needs to sign off copy, approve designs, and interpret results alongside the agency.
That's not a line on an invoice, but it's real time, and it's the reason CRO projects often take longer than expected when the internal reviewer is also running three other things.
This is the one that's easiest to ignore because it never shows up as a bill.
If your site gets a reasonable amount of traffic and converts at, say, half the rate it could, you're not saving money by skipping CRO - you're paying for traffic (through SEO effort, ad spend, or content marketing) that never gets the chance to convert.
The "cost" of CRO has to be weighed against the value of the leads it's currently failing to capture, not against zero.
There are three broad routes, and the right one depends on how much testing infrastructure and internal expertise you genuinely have - not how much you'd like to have.
Platforms like HubSpot include A/B testing and reporting on their Professional and Enterprise tiers, which can be a reasonable starting point if you've got someone internally who understands both the tool and how to interpret the results. It's a good fit for smaller-scale, page-level tests.
Technically possible with a heatmap tool and a testing platform bolted on. Realistically, this is where a lot of "we ran a test but couldn't tell what actually happened" conversations start, because DIY CRO tends to skip the research phase that tells you *why* a page underperforms before you start guessing at fixes.
For anything beyond a single-page tweak, a partner earns their fee by doing the unglamorous parts properly: the research, the persona work, the hypothesis prioritisation, and the discipline to keep testing after the first "quick win" rather than declaring victory and moving on.
Here's the high-level process we run CRO projects through with clients:
Analytics review, heatmaps, session recordings and stakeholder input, so recommendations are based on evidence rather than opinion.
Understanding who's actually landing on each page, what they need to know, and what's stopping them from converting.
New or optimised pages, built with testing and tracking in mind from the outset, not bolted on afterwards.
Results reviewed, hypotheses refined, and the next round of tests prioritised - because CRO earns its keep through cumulative improvement, not a single redesign.
Dale Office Interiors, a commercial office design and fit-out company, came to us with a familiar problem: decent traffic, a strong reputation, but a conversion rate that wasn't moving in line with visitor numbers. They'd been relying heavily on repeat business and word of mouth, and needed their website to start doing more of the work.
We focused the engagement squarely on conversion - persona and buyer journey work, new landing pages built around specific offers, and sitewide conversion opportunities on the content that was already getting traffic.
In one quarter, that work produced a 556% increase in contacts generated, with conversion rates of 30-60% across six new asset landing pages, and a key blog's call-to-action conversion rising from 0% to 9.23%.
Their own words on it:
[Axon Garside] understood exactly where we wanted to go with things, correspondence was always sharp and to the point, and quick, and beyond that it was recommendations from peers - stating how good AG were and how supportive they've been to their business."
It's not the only example.
Huthwaite International used a similar approach - repositioning existing content around clearer conversion paths - to grow online revenue by 433% in a year.
Winnow saw an 18x increase in organic leads and a 45% boost in landing page conversion rate as part of a broader inbound and CRO effort.
None of these came from a single round of button-colour testing. They came from research-led changes to the pages and journeys that mattered most.
If you want a rough rule of thumb before you get quotes in:
A single high-priority page, minimal testing infrastructure: budget towards the lower end, £1,500-£4,000.
A structured project across your key funnel pages, with proper research and a handful of new assets: £8,000-£20,000 is realistic.
An ongoing programme, where testing and iteration continue month over month rather than stopping after one round: £2,000-£6,000+ a month, on top of any testing tool subscriptions.
Whatever number you land on, make sure it includes research and measurement, not just design and build. Those are the parts that don't show up in a mock-up but absolutely show up in your conversion rate three months after launch if they're skipped.
Broadly, yes - not because the work is priced differently by sector, but because different sectors are optimising for different things.
Engineering and manufacturing businesses are usually less concerned with raw conversion volume and more concerned with lead quality - fewer, better-qualified enquiries from buyers who are actually in-market, often against long, considered sales cycles and detailed product ranges.
CRO here tends to lean on qualification logic within forms, and on content that filters tyre-kickers out before they reach the sales team, rather than simply maximising the number of form fills.
Membership bodies, charities and training organisations are usually optimising a different set of actions entirely - event bookings, course sign-ups, donations, membership renewals - each with its own journey and its own drop-off points.
CRO for this kind of organisation frequently means mapping several distinct conversion paths rather than one, which adds research time upfront but tends to uncover more quick wins, simply because these journeys are optimised less often than a standard "contact us" form.
Neither of these is a reason to expect a dramatically different quote. It's a reason to make sure whoever you're working with understands what "a good conversion" actually looks like for your specific buyer, not just for a generic B2B lead form.
Cost isn't the only variable, but it's rarely independent of who you choose either.
A cheap quote from someone who skips the research phase will often cost more in the long run than a properly scoped project, because you end up testing changes with no clear hypothesis behind them - and no reliable way of knowing whether an uplift is real or just noise.
A few questions worth asking anyone quoting you for CRO work:
Do they start with research and analysis, or go straight to redesigning pages based on opinion?
Do they have a documented approach to prioritising which tests or changes to run first?
Is reporting built around genuine conversion metrics, or vanity numbers like time-on-page?
Will testing and iteration continue after the first round, or does the engagement stop the moment the new pages go live?
Do they understand your buyer journey specifically, or are they applying a generic B2B playbook regardless of sector?
We're not neutral on this - we'd obviously like that partner to be us.
But even if it isn't, asking these questions of anyone quoting you for CRO will steer you away from the most common and most expensive mistake in this space: paying for a redesign, calling it optimisation, and wondering six months later why the conversion rate looks exactly the same as it did before.
It depends on your traffic and average deal value, but most structured projects show measurable movement within the first quarter - a handful of tests live, early signal on what's working.
The bigger, compounding gains tend to show up over two to three quarters, once you've had enough rounds of testing to stop guessing and start knowing what your specific audience responds to.
Formal split testing needs volume, so if you're working with a smaller audience, the value shifts towards qualitative research - heatmaps, session recordings, direct feedback - and considered, evidence-based changes rather than live A/B tests.
It's still worth doing. It just looks different, and it should cost less, because you're not paying for testing infrastructure your traffic can't yet support.
Treating it as a one-off spend rather than a compounding one.
A single round of changes might lift conversion by a noticeable amount, but the businesses that see the biggest long-term gains are the ones who keep testing after the first win, rather than treating "we redesigned the landing page" as the finish line.
Usually both, in different proportions depending on the page.
A page with strong content but poor structure needs design and UX work. A well-designed page that isn't converting often has a messaging problem - it's not speaking to what the visitor actually needs to hear at that point in their journey. The audit phase is what tells you which one you're dealing with, which is exactly why skipping it tends to produce disappointing results.
We build websites exclusively on HubSpot CMS, which means conversion tracking, testing and reporting are connected to the same system as your CRM data from day one - so you're not just measuring form fills, you're measuring which pages actually produce leads that turn into revenue.
CRO, for us, isn't a bolt-on service that happens once after launch. It's built into how we approach B2B website design generally, and it's usually one of the highest-return pieces of work we do with clients, because it doesn't require spending more on traffic - it requires spending better on the journey that traffic already takes.
If you're weighing up a CRO project and want to see what a realistic, tailored number looks like for your specific site and sector, take a look at our pricing guide - it breaks down what's involved, what to expect, and how we price CRO alongside broader website and inbound work.