Conversion · Updated · 12 min
Website conversion rate: why more visitors are rarely the answer
Conversion rate is the most honest number a website has, because it cannot be talked up. It says one thing only: out of a hundred people who were here, this many did what they were supposed to do. Everything after that is interpretation.
Alexander HertiFounder of Business Growth Solutions
In short
A website conversion rate is the share of visitors who complete the desired goal: conversions divided by visitors, times 100. 30 enquiries from 2,000 visitors make 1.5 percent. For B2B service firms the number that matters is the booked meeting or the enquiry, measured separately by traffic source, and the page gets fixed before you buy more reach.
Definition and formula: how to calculate the conversion rate
The conversion rate is the share of a page's visitors who complete a defined goal action. The formula could hardly be simpler: conversion rate = conversions ÷ visitors × 100.
An example: your service page had 2,000 visitors last month, and 30 of them sent an enquiry. 30 divided by 2,000 is 0.015, times 100 makes 1.5 percent. If 12 of those 30 booked a slot in the calendar straight away, the conversion rate for booked meetings is 0.6 percent. Both numbers are correct, they just answer different questions.
What sits in the denominator matters. You can count visitors (users) or sessions. Someone who comes back three times and enquires on the third visit is one user but three sessions, so the session-based rate is almost always lower. Pick one and stick with it, or you will be comparing two different metrics month after month.
Rule of thumb: never calculate the conversion rate across the whole site, only per page type and per traffic source. An average across blog, home page and landing page tells you nothing.
Decide first what actually counts: macro and micro conversions
Before improvement can be discussed, it has to be clear what is being counted. On a B2B site that is almost never a purchase, it is a booked meeting, an enquiry or a call. Those are the macro conversions, the actions the page exists for.
Next to them are micro conversions: smaller steps that show someone is heading in the right direction. They are useful for diagnosis, but no substitute for the actual goal.
- Macro: a meeting booked in the calendar, a submitted enquiry form, a tap on the phone number followed by a call, a quote request.
- Micro: calendar opened but not booked, form started, pricing page or case study viewed, whitepaper downloaded, newsletter subscribed.
- Not a conversion: time on page, page views and scroll depth. That is behavioural data that helps understanding, not something to optimise for.
Newsletter sign-ups and downloads belong on their own line, because they are a different promise. Adding everything together produces a number that looks good and steers nothing. Five percent sounds respectable until you notice that four of those points are brochure downloads and one point is real enquiries.
A second point that often gets missed: count separately by source. Visitors from an ad, from search and from an email behave so differently that a combined average makes both groups invisible.
Measuring conversions in GA4: setting up key events properly
In Google Analytics 4, conversions are now called key events. The principle is the same: you capture an event and mark it as especially important. Standard properties allow up to 30 key events, which is plenty for any B2B site if you stay disciplined.
- Decide which two or three actions are macro conversions, and write it down before you set anything up.
- Fire a dedicated event for each of them, for example on reaching the thank-you page after the form or on the confirmation step of the embedded booking tool. Loading the form page is not a conversion.
- Mark those events as key events under Admin, Events. Micro conversions stay ordinary events so they do not dilute reports and bidding strategies.
- Test every key event yourself once with a dummy enquiry, and after four weeks compare the count in GA4 with the count in your inbox or CRM.
The last step is the most important one. The gap between analytics and inbox shows you how much your measurement is missing, and therefore how much weight any further number in this article can bear for your site.
Consent, Consent Mode and server-side tracking: the measurement gap in the EU
In the EU, analytics and advertising cookies may only be set after consent. Anyone who declines the cookie banner does not show up at all in a classic setup, neither as a visitor nor as a conversion. The share who agree is your consent rate, and it varies considerably by banner, audience and device.
This skews the conversion rate in two ways. First, absolute numbers go missing: GA4 shows fewer enquiries than arrive in your inbox. Second, the rate itself is not necessarily right, because people who accept can behave differently from people who decline. Anyone who redesigns the banner and then sees a rising conversion rate may only have changed the measurement, not the page.
Google's Consent Mode softens this. In basic mode, Google tags only load after consent. In advanced mode they load straight away with denied defaults and send cookieless signals from which Google models missing conversions. Modelled numbers are estimates and need enough data volume, they do not replace a count. Whether advanced mode is legally right for your company is something to settle with your data protection adviser.
Server-side tagging is the next step: data goes to your own server first and only from there to analytics or ad platforms. That can save load time because fewer scripts run in the browser, and it gives you more control over which data is passed on. It does not replace consent. We set this up as part of building websites.
The most reliable number for macro conversions on a B2B site is not analytics but your CRM or inbox. Use analytics for ratios and trends, your own count for the absolute truth.
What website conversion rate is normal in B2B
The question comes up in every first call, and the honest answer is: there is no universal number. For B2B services that need explaining, enquiries and bookings on most sites we see land in the low single-digit percentage range. With cold ad traffic it sits below that, with visitors who searched for your name well above.
The table below is therefore explicitly not a study and not an industry benchmark, but rough orientation from our project experience. The ranges are deliberately wide, because offer, price and audience matter more than any average.
| Traffic source and page type | Enquiry (experience-based) | Booked meeting (experience-based) |
|---|---|---|
| Brand search to home or service page | 3-10% | 1-5% |
| Organic search to service page | 1-4% | 0.5-2% |
| Organic search to blog article | 0.1-1% | below 0.5% |
| Search ads to landing page | 1-5% | 0.5-3% |
| Cold social traffic to landing page | 0.3-2% | 0.2-1% |
| Email or referral to landing page | 2-10% | 1-5% |
That is exactly why comparing yourself to an industry figure from a blog post is worthless, this table included. The only comparison worth anything is your own number from last quarter, split by source.
And the rate alone decides nothing. A site at one percent with five-figure deals beats a site at five percent with small ones. What a meeting may be worth we work through here: what may a B2B meeting cost.
Where it actually snags: four places
There are four places where visitors get lost, and they need completely different answers. First, before the page: the ad or the search result promises something other than what the page delivers. Then the page is not at fault, and no button will save it.
Second, at load: anyone who leaves over load time never sees your offer. Third, at comprehension: the page loads, gets read, and after thirty seconds it is still unclear who this is for. Fourth, at the action: everything is understood, but the form wants nine fields or the calendar shows no slot this week.
Telling those four apart is the actual work. The usual mistake is to tinker at place four when the problem is one or three, changing the button colour while the copy is what fails.
Load time and Core Web Vitals as a conversion factor
The second place, load time, is the only one that can be measured objectively with ease. Google bundles it into the Core Web Vitals and names three thresholds for a good user experience, measured at the 75th percentile of page loads:
- Largest Contentful Paint (LCP): the largest visible content is there within 2.5 seconds.
- Interaction to Next Paint (INP): the page responds to clicks and input within 200 milliseconds.
- Cumulative Layout Shift (CLS): the layout barely shifts, with a score of 0.1 or less.
That speed moves money is well documented. The Google-commissioned study "Milliseconds Make Millions" by 55 and Deloitte analysed 37 brand websites with over 30 million mobile sessions. For lead generation sites, the largest effect observed from a page that was 0.1 seconds faster was a 21.6 percent increase in users who progressed to the form page. These were not B2B service firms and it is a correlation, but the direction is clear.
The most common brakes on B2B sites are uncompressed hero images and videos, too many tracking and chat scripts, heavy site-builder themes and a booking calendar that blocks the whole page. Check your page with PageSpeed Insights on mobile, not only on desktop. How we think about speed and design together is in modern UX/UI design.
How to see this without expensive tools
Three sources are enough to start. First, time on page relative to the length of the text: if people stay fifteen seconds on average, nothing was read, it was clicked away. Second, scroll depth: if hardly anyone reaches the offer, the top of the page is the problem.
Third, and this is the underrated source: the enquiries themselves. If every second enquiry contains the same question, your page is not answering it. That is the cheapest conversion analysis there is, and it is already sitting in your inbox.
What you do not need for this: a heatmap subscription, a testing tool and an agency for multivariate tests. Serious A/B testing needs sample sizes most B2B sites do not reach in a month. How many exactly is further down.
The changes that most often do something
First: delete fields. Every mandatory field costs completions. Name and email are almost always enough, ask the rest in the conversation. A phone number as a required field is the single most expensive item we see regularly.
Second: name the next step concretely. Not get in touch, but what happens and how long it takes. Twenty minutes, free, no sales pressure tells a reader more than any adjective about your company.
Third: state the price range, at least as a span or a floor. It costs you enquiries from people who would never have bought anyway and clearly raises the share of useful conversations. Fourth: put the evidence where the decision happens, so reviews and cases right next to the button rather than on a separate page.
Fifth: embed the calendar directly. Anyone who has to wait for a reply after enquiring cools off. An embedded calendar with open slots in the next few days removes that wait. It does have to load fast and show real, near-term slots, or it becomes an obstacle itself.
Checklist: conversion optimisation for B2B websites
We go through these points on every existing site, roughly in this order. The first three concern measurement, because without it every further change is flying blind.
- Macro and micro conversions defined in writing and reported separately.
- Key events set up in GA4, tested and reconciled with inbox or CRM.
- Consent rate known, Consent Mode set up correctly, measurement gap quantified.
- Conversion rate per traffic source and page type rather than one overall average.
- Ad, search result and page headline promise the same thing.
- The first screen says who the offer is for and which problem it solves.
- Core Web Vitals in the green on mobile.
- One main path per page, with the next step repeated after each major section.
- The button describes what happens, for example "Book a 20-minute call" instead of "Submit".
- Form with three to four fields at most, phone number optional.
- Calendar embedded, with open slots in the next few days.
- Evidence such as testimonials, logos and case studies right next to the button.
- Price range or starting price stated.
- Common objections from real enquiries answered on the page.
- Thank-you page with a clear next step, so that the enquiry turns into a meeting.
When A/B tests make sense and when they do not
An A/B test shows two versions of a page, randomly split, and compares the conversion rate. That is clean, but only with enough data. How many visitors you need depends on your current rate and on how large a difference you want to detect. The table gives approximate numbers per variant, calculated with the standard formula for comparing two proportions.
| Current conversion rate | +20% relative | +30% relative | +50% relative |
|---|---|---|---|
| 1% | approx. 42,700 | approx. 19,800 | approx. 7,800 |
| 2% | approx. 21,100 | approx. 9,800 | approx. 3,800 |
| 3% | approx. 13,900 | approx. 6,500 | approx. 2,500 |
| 5% | approx. 8,200 | approx. 3,800 | approx. 1,500 |
A worked example: your landing page gets 2,000 visitors a month and converts at 2 percent. To reliably detect a rise to 3 percent, a 50 percent uplift, you need around 3,800 visitors per variant, about 7,700 in total. That is almost four months for a single test, and only for a large effect. Smaller improvements of 20 percent would take almost two years.
For most B2B sites this means A/B tests only pay off at several thousand visitors a month on the same page, and then for bold changes such as a different offer or headline, not button colours. Below that, you are better off with before-and-after comparisons over longer periods, with the questions from real enquiries, and with five conversations with customers who tell you what they were looking for on the page.
Why more traffic moves the problem
The maths is uncomfortably simple. At one percent conversion, doubling enquiries costs double the budget. If the rate goes from one to two percent, enquiries double at the same budget. One of those is a permanent increase in cost, the other a one-off effort.
That is why the order is almost always the same: page first, reach second. Running a campaign onto a page that hides the next step means paying a lot for something that gets lost afterwards.
It works the other way too. A very good page with no visitors stays a very good page with no visitors. Both sides of the equation have to hold, just in that order, because the second does not cure the first.
What we took from it
We build pages backwards, from the next step. One main path, as few fields as possible, evidence next to the button, and the calendar on the page itself rather than behind another click. Since 2023 we have booked over 2,000 meetings for clients, and the lesson is always the same: you win where the next step takes the least effort.
If you have an existing site and do not know which of the four places is losing people, twenty minutes is usually enough to narrow it down. What we take on when building sites, from measurement to the embedded calendar, is on our websites page. For a first conversation about your numbers, get in touch.
Frequently asked questions
What is a good conversion rate for B2B websites?
There is no universal number. For B2B services that need explaining, enquiries and bookings in our experience mostly land in the low single-digit percentage range, lower for cold ad traffic and clearly higher for visitors from brand search. Comparing with your own previous period, split by traffic source and page type, says more than any industry figure. And the rate alone decides nothing: one percent with large deals can be worth more than five percent with small ones.
How do I calculate my website's conversion rate?
Divide the number of conversions by the number of visitors and multiply by 100. Example: 30 enquiries from 2,000 visitors gives 30 ÷ 2,000 × 100 = 1.5 percent. Decide beforehand whether you count visitors or sessions, because the session-based rate comes out lower. Also calculate separately by page type and traffic source, and for each macro conversion on its own, so do not lump enquiries and booked meetings together.
Why is my conversion rate so low?
It usually comes down to one of four places: the ad promises something other than the page, the page loads too slowly, after thirty seconds it is unclear who the offer is for, or the next step is too much effort, for example because of long forms. Check the measurement too: a low consent rate or badly configured key events make the rate look worse than it is. So always compare GA4 with your inbox or CRM.
Do newsletter sign-ups count as conversions?
Yes, but as a micro conversion, not a macro conversion. A newsletter sign-up shows interest but is a different promise from an enquiry or a booked meeting. Report it on its own line, and do not mark it as a key event in GA4 if you are optimising for enquiries. Adding sign-ups and enquiries together produces a number that looks good but no longer steers any decision.
How many visitors do I need for an A/B test?
More than most B2B sites have. At a 2 percent conversion rate you need around 3,800 visitors per variant to detect a 50 percent uplift, and around 21,100 for 20 percent. That assumes 95 percent confidence and 80 percent power and is an approximation. Below a few thousand visitors a month on the same page, before-and-after comparisons, customer conversations and analysing real enquiries are usually the better choice.
Does an embedded booking calendar improve conversion?
Often yes, because it removes the wait between enquiry and meeting and the visitor sees straight away when things move forward. That is why we usually embed the calendar on the page itself rather than behind another click. Two conditions apply, though: it has to load fast without slowing the page down, and it has to show real open slots in the next few days. A calendar that offers nothing for three weeks tends to put people off.
Sources
- Google Analytics Help: Mark events as key events
- Google for Developers: Consent mode overview
- Google for Developers: Server-side tagging with Google Tag Manager
- web.dev: Web Vitals
- web.dev: Milliseconds make millions (study by 55 and Deloitte, commissioned by Google)
- Evan Miller: Sample size calculator for A/B tests