The leak is in the journey.

Your website is the most important digital thing you own: the one place you control, where a stranger becomes a customer. If that site "brings in no enquiries", it can be two things: too few of the right people arrive (a visibility problem), or the people who do arrive drop out before taking the step (a conversion problem). This is the guide about the second, because it’s more often the cause than people think and it gets far less attention. So look at your visitor figures first: if almost nobody comes, start with being found. If plenty come but nothing happens, you’re in exactly the right place.

This is the complete guide to fixing that. First the funnel itself, stage by stage, so you know exactly where you lose customers. Then the tools to close each leak: the words, the psychology, the trust and the speed. Then a way of working that improves it measurably instead of guessing. Take your time, or jump through the index to the stage where you suspect it goes wrong.

Your conversion is the share of visitors who take the step you want: calling, emailing, asking for a quote, booking, buying. Most sites sit between 1 and 4 percent. It differs by trade:

These are international figures (summarised by Lucky Orange from sources including IRP Commerce, WordStream and Unbounce), so read them as a reference point rather than a rule. If you sit at 1 percent where your trade sits at 3, you lose two of every three customers you could have had, and you don’t fix that with more visitors, but with a better site.

That’s where the leverage is. Improving conversion is the cheapest growth there is, because you don’t pay a cent more for the traffic. Work it out with your own numbers. Half a percent more sounds like nothing, but in revenue it’s often the difference between a site that pays for itself and one that doesn’t:

A visitor goes through a series of moments, and at each one a share falls away. Dave McClure, founder of 500 Startups, summed up that journey in 2007 as the "Startup Metrics for Pirates", five stages you pronounce as AARRR: Acquisition, Activation, Retention, Referral and Revenue. Devised for tech companies, but the logic holds for any site that wants to sell something. I walk through them in the order a visitor experiences them (with Revenue in its natural place in the middle):

1. Acquisition. Do you attract the right people? The wrong audience never converts.
2. Activation. Does a stranger understand and trust you within seconds, and stay?
3. Conversion. Does that engaged visitor take the step to an enquiry?
4. Retention. Most don’t buy on the first visit. Do you come back into view?
5. Referral. Do satisfied customers pass it on?

A chain is as strong as its weakest link. A great offer nobody finds brings in nothing; mountains of traffic on a confusing page just as little. That’s why "a prettier site" is almost never the answer: if the weak link wasn’t the looks, nothing changes.

How you get found (Google, adverts, your business profile, word of mouth) belongs here, and there’s a separate guide about that. For conversion what counts is the quality of your traffic, not the quantity. A thousand visitors looking for something other than what you offer bring in less than a hundred who want exactly your service.

The hinge is the promise. What a visitor expects when they click has to match what they find. If someone searches "emergency plumber Rotterdam" and lands on a general home page, the promise is broken and they click back. Marketers call this "message match": the scent of the trail has to carry from the search to the page they land on. Make every important entry point line up exactly with what people come for, ideally a dedicated page per service or place.

This is where most service businesses bleed. The visitor is in, and now you have literally seconds to make three things clear: what you do, that you can be trusted, and what the first step is. If that fails, they’re gone.

The figures are merciless. People form a first judgement of a page within about 50 milliseconds, faster than they can read anything (research by Lindgaard and colleagues, 2006). And within about ten seconds they decide whether to stay. Eye-tracking by the Nielsen Norman Group found that visitors spend roughly half their attention on what’s visible without scrolling. If your core message sits below that, for most people it doesn’t exist.

• One clear promise, at the top, in human language. A stranger should be able to say, without effort, what you do and who for. How you word that promise matters so much that there’s a whole section on it further down.
• Visible proof that you’re real and good. An address, a face, a few reviews, real work. Trust is the crux for service businesses; there’s a section on that too.
• Calm and direction. Not ten menu items, banners and pop-ups all fighting for attention at once. One glance, one message, one obvious next step.
• Speed and mobile. Slowness costs you visitors before they see anything, and most arrive on their phone. The concrete measures are in the Core Web Vitals section.

Buying a service is riskier than buying a product. You’re buying a promise from someone you have never met. That’s why trust is almost always the heaviest link for service businesses, and the most rewarding to strengthen. Stanford University researched with thousands of people what makes a site credible and poured it into ten guidelines. Translated to your site:

• Real photographs of you, your team and your work. Not smiling stock models who appear on ten other sites. Real images convince; stock photographs make you anonymous.
• Reviews and customer stories, shown properly. Recent weighs more than old, specific more than vague, with a name and a face more than anonymous, and on an independent platform (Google) more credible than only on your own site. A page with reviews converts on average 270 percent better (Spiegel Research Center, Northwestern, 2017); the effect levels off after about five.
• Address, phone number and company registration number, visible. It proves there’s a real business behind it that will still exist tomorrow.
• A face and a story. People do business with people, certainly in services.
• Take the risk away. A guarantee, money back, a no-obligation introduction, "a reply within a day". Every promise that puts the risk on you instead of the customer lowers the threshold.
• A secure connection and a flawless page. A padlock, a privacy statement, and no typing errors. Small things, but they undermine trust the moment they’re missing.

And watch out for the AI-slop trap

There’s a new, creeping trust-breaker: sites that look as if an AI or a stock template spat them out. The same typefaces, the same purple-blue gradients, the same meaningless headlines ("We take care of everything for you"), the same stock smile. Such a site looks like every other, and that’s exactly what undermines the trust that turns a visitor into a customer. For a business owner who has already been let down once, generic reads as cheap and anonymous. Being distinctive and well made isn’t vanity, those are trust signals.

The biggest conversion control on your site isn’t a button, it’s your words. A service business sells with text: what you promise, in whose language, and how clearly. Yet this is the most neglected part of nearly every site.

Start with one sharp promise. At the top there should be one sentence saying what you do for whom, and why you. A simple frame: I help [who] to [what result], without [the familiar pain].

• Write about the customer, not about yourself. Count how often "I" or "we" appears on your home page against "you". Not "we have 15 years of experience", but "you get someone who has done this daily for 15 years".
• Sell the outcome, not the feature. Behind every feature sits a "so what?". "Fibre" is a feature; "never another stuttering video call" is the outcome. Do the "so what?" test.
• Use a proven structure. PAS (Problem, Agitation, Solution) or AIDA (Attention, Interest, Desire, Action). If your page follows that order, it leads to the button by itself.
• Write your buttons as a promise, not as a command. Not "Submit", but "Ask for a no-obligation quote". Small word choices count: conversion specialist Michael Aagaard raised clicks considerably by changing "Start your free trial" to "Start my free trial".
• Handle objections where they arise, right next to the button: "what if it disappoints?", "roughly what does it cost?", "how soon will I hear something?".
• Cut the jargon. "We unburden", "total solution", "synergy": write the way you speak.

People decide with their gut and justify it afterwards. A handful of principles explains nearly everything that does or doesn’t work on your site. Use them to remove thresholds and to tell the truth well, never to mislead, because a business owner who feels played is gone for good. Psychologist Robert Cialdini boiled the best known down to seven principles:

• Social proof. People do what others do: reviews, numbers, familiar logos.
• Authority. People follow whoever seems expert: knowledge of the trade, certifications, and being the one who explains it.
• Liking. People buy more readily from someone they like: a face, a story, the same language.
• Reciprocity. Whoever receives something valuable first wants to give something back: a free check, a useful guide, honest advice.
• Commitment. A small first yes makes a bigger one easier. Ask for a low threshold first.
• Scarcity, but only when it’s real. Fake scarcity and countdown clocks cost you trust instead.
• Unity. "For business owners like you" works harder than "for everyone".

On top of that, a few thinking errors you can use honestly: loss aversion ("what you’re leaving on the table every month" hits harder than what you would gain), anchoring (the first number colours the rest), cognitive fluency (what reads easily feels more trustworthy), and choice overload (too many options paralyse, Hick's law).

The visitor understands you and trusts you. Now clear away every last threshold between "interesting" and "done".

• One unmistakable next step. More choices means more hesitation. Give one main action per page and repeat that button; the strongest landing pages keep an "attention ratio" of 1 to 1 (Unbounce).
• Proof next to the button. Put your strongest review or guarantee right where someone clicks, not on a separate page.
• Remove the last doubt with a price indication, a guarantee, "a reply within a day", no obligations.

The enquiry moment, field by field

Most enquiries die on the form, right on the finish line. The Baymard Institute found that 18 percent drop out because they had to create an account and 17 percent because it took too long. Those are shop figures, but the principle holds just as hard for a contact form:

• Ask only what you need to call them back. Often a name and one way to get in touch. Every field you cut raises your enquiries.
• Don’t force anyone into an account. Nobody should have to invent a password to send an enquiry.
• Offer the route the customer picks: a clickable phone number at the top (on a phone a tap dials), WhatsApp if that suits you, or a short form.
• Show errors straight away, per field, not only after sending.
• Make your thank-you page more than "sent". Confirm what happens now and when ("I’ll call you within one working day").

Speed is invisible until it’s too late. According to Google research, the chance of someone bouncing rises by 32 percent when your page goes from 1 to 3 seconds of load time (Think with Google, 2017). Research by Portent (2022) found conversion peaks at 1 to 2 seconds. Google measures this with three figures, the Core Web Vitals:

You can check them free with PageSpeed Insights, and see the figures from your real visitors in Google Search Console. Note: mobile is judged separately and more strictly, and that’s exactly where most traffic comes from. The usual culprits are images that are too heavy, too many scripts and slow hosting; jumping (CLS) you fix by reserving space for images in advance.

By far the most first-time visitors don’t become customers straight away, certainly not for a considered purchase. They’re not lost, unless you let them disappear.

• Capture an email address before they leave, in exchange for something useful (a checklist, a price indication). Reciprocity in practice.
• Stay visible with remarketing: with a simple pixel you show adverts to people who already visited your site.
• Give a reason to come back. With a service, whoever comes to mind at the right moment often wins.

Word of mouth is often the strongest channel for a service business, and it feeds two stages: a review convinces the next visitor, and a recommendation brings in warm traffic.

• Ask for reviews systematically, after every finished job, at the moment satisfaction is highest, with a direct link. The cheapest conversion lift there is.
• Reply to your reviews, including the poor ones. That shows everyone reading along how you handle problems.
• Make passing it on easy with something shareable and a clear name.

You don’t have to guess. Three tools, free or cheap.

GA4: where in the journey you drop out

• First mark what a "customer" is as a key event: Admin, then Events, and switch on "Mark as key event". A phone click or booking button you create through Create event. (explanation)
• Build a funnel and read the drop-off: Explore, then the Funnel exploration template. Steps from arrival to submitted form. The step with the biggest fall is your bottleneck. (explanation)
• Find your weak entry pages: Reports, Engagement, Landing page. Many sessions plus low engagement equals your biggest opportunities. (explanation)
• Note: the bounce rate in GA4 is now simply the inverse of engagement (a session counts as engaged after more than 10 seconds, a key event, or two pages), not the old "single-page visit".

Microsoft Clarity: watch people over the shoulder

GA4 tells you where people drop out, Clarity (completely free) shows you why: recordings of real sessions, a scroll heatmap (where does the colour go cold, is your button above that?), a click heatmap (do people click your button or something that’s not clickable?), and frustration signals: rage clicks (something doesn’t work), dead clicks (broken or misleading), excessive scrolling (can’t find something) and quick backs (wrong page).

A/B testing, with an honest caveat

A/B testing lets the data decide which of two versions wins. But the honest truth: most small businesses have too little traffic to test reliably (roughly thousands of visitors and hundreds of conversions per variant needed). Below that: test big jumps instead of button colours, do a simple before-and-after, and lean on your recordings and heatmaps. Google Optimize was shut down on 30 September 2023; good starting points are Crazy Egg or the free open-source options PostHog and GrowthBook.

Loose tips change little; a fixed way of working does. Improving conversion is a loop you keep turning:

1. Measure. Where is the biggest drop? Your GA4 funnel points at it. Always start at the biggest leak.
2. Understand. Why do people drop out there? The Clarity recordings and heatmaps show that.
3. Form a hypothesis: "If I [change], then [effect], because [reason]."
4. Prioritise with ICE (Impact, Confidence, Ease, each 1 to 10; from Sean Ellis) or PIE (from WiderFunnel): take the highest first.
5. Change one thing, substantially, so you know what made the difference.
6. Verify (A/B if you have the traffic, otherwise before-and-after), and keep it or roll it back.

Advertising makes every leak more expensive, because you’re now paying for visitors who can still drop out. The click has to match the landing page seamlessly: send "emergency plumber Rotterdam" to exactly that page, not to your home page. Google judges that through the Quality Score (ad relevance and landing page experience, Google itself); if your page matches poorly, you pay a higher price per click.

And only advertise once your conversion is in order. Buying traffic to a page that converts badly is pouring money into a bucket with a hole in it. The other way round, advertising multiplies the return on every improvement: double your conversion and the same advertising euro brings twice as many customers.

Fair is fair: you can make your page perfect and still get no customers, because the problem isn’t in the site but in what you’re offering. A crystal-clear presentation of an offer nobody wants, or one that’s too expensive or too vaguely priced, still doesn’t convert. If you recognise this, the site isn’t your first concern:

• People do get in touch, but don’t become customers. Then your site works fine, and it sits in your quote, your price or your sales conversation.
• You look identical to cheaper competitors. Then it’s a positioning question: why would anyone choose you?
• The demand simply isn’t there, or not at your price. Then more traffic or a prettier site is a waste of your money.

This is a business question first, not a website question. Solve that, and only then is it worth fine-tuning the page around it.

Go through this, tick what holds, and copy it along. Every tick that’s missing is a leak.

Don’t want to guess where your site loses customers? Run the free website check. It measures your site for you, on first impression, speed, visibility and the technical side, without you having to look anything up. You see the outcome straight away, and you know what you would pick up first.