Most businesses respond to a disappointing month by buying more traffic. It is the obvious move and it is usually the expensive one. If a hundred visitors produce one enquiry, getting to two enquiries per hundred is worth exactly as much as doubling your ad spend, and it does not recur as a monthly cost.
Conversion rate optimization has a reputation for being about button colours, which is unfortunate, because the real work is far less exotic. It is finding the specific point where people decide to leave and removing whatever made them decide.
Work out what your conversion rate actually is
A single sitewide conversion rate is close to useless, because it averages people who arrived ready to buy with people who landed on a blog post from a search about something adjacent. Split it at minimum by traffic source, by device, and by landing page.
Define what counts as a conversion before you measure anything, and be strict. A form submission, a phone call over thirty seconds, a booking, a quote request. Newsletter signups and brochure downloads are worth tracking separately as soft conversions, but folding them into one number hides the thing you actually care about. Setting this up properly is covered in our guide to Google Analytics 4 and conversion tracking.
Find the page where people are actually leaving
Before changing anything, find the leak. Look for the pages with high traffic and low conversion, the steps in a multi page process with the sharpest drop off, and the difference between mobile and desktop performance on the same page. One of those three usually contains the whole problem.
The common finding is not a bad page. It is a mismatch: an ad promising one thing and a landing page describing another, a search phrase about price landing on a page that never mentions price, or a mobile layout that pushes the enquiry form below four screens of content. Fixing a mismatch is faster than redesigning anything.
Answer the three questions every visitor has
Within a few seconds of landing, a visitor is deciding three things: is this what I was looking for, can this business do it for someone like me, and what do I do next. A page that answers all three above the fold converts. A page that opens with a mission statement answers none of them.
In practice that means a headline naming what you do and who for, immediate evidence in the form of a result, a named client or a review, and one obvious next action. Everything else on the page is supporting material for somebody who has already decided to keep reading.
Cut the friction out of your forms
Forms are where the most conversions are lost and where the fastest gains live. Every field is a small tax on completion, and most forms carry fields nobody ever uses. Remove anything you could ask for later in a phone call, make optional fields genuinely optional, and never ask a question the answer to which changes nothing about what happens next.
Beyond field count: validate as the person types rather than dumping five errors after submission, keep the whole form on one screen where possible, use the right keyboard type on mobile so a phone number field opens a number pad, and say what happens after submission. The sentence we reply within one business day removes more hesitation than any design change.
Make the call to action a decision, not a label
Submit, Learn more and Get started describe the mechanism rather than the outcome. Get my quote, Book a site visit and See pricing describe what the visitor receives. The second kind converts better because the visitor is deciding whether the trade is worth it, and the button is the only place you state the trade.
One primary action per page. If a page offers a call, a form, a download and a newsletter with equal weight, you have asked the visitor to make a decision about which decision to make. Pick the one that matters, make it visually dominant, and demote the rest.
Put proof next to the ask
Trust signals work where the hesitation happens, which is beside the button, not on a separate testimonials page nobody visits. A specific review from a named customer near the form does more than a carousel of five star quotes in the footer.
Be specific. Rated 4.9 across 180 Google reviews beats Trusted by hundreds. A result with a number in it beats a promise. Licences, insurance, guarantees and response times are worth stating plainly for service businesses, because they answer the unspoken question about risk.
Give a price, or explain why you cannot
The most common reason a visitor leaves a service website is that they could not work out whether they could afford it. Hiding pricing entirely does not delay that question, it moves it to a competitor who answered it. Every quote request you receive from somebody with a budget a third of your minimum is a cost you paid for withholding the number.
If genuine pricing depends on scope, publish the shape of it: a starting figure, a typical range for a common job, or three worked examples with what each included. That filters out the enquiries you did not want and reassures the ones you did, and it is usually the single highest impact change available on a service site.
Fix the mobile experience separately
Most local service traffic is mobile and most sites are still designed on a desktop. The result is predictable: tap targets too close together, a phone number that is not a tap to call link, a sticky banner covering the form, and a hero image that pushes the first useful sentence below the fold.
Test on a real phone on a real mobile connection, not just a browser window resized. Then check the loading side of it, because a page that takes six seconds to become usable has lost a share of its visitors before any of your copy gets read. Our guide to speed and Core Web Vitals covers the technical half.
Watch real sessions before you redesign
Heatmaps and session recordings answer questions analytics cannot. Where do people stop scrolling. What are they clicking that is not a link. Where do they hesitate in the form. Do they reach the pricing section at all. Twenty recordings of real visitors on your highest value page will produce more actionable findings than a month of dashboard reading.
Add one exit question on the pages that matter: what nearly stopped you from enquiring today. The answers are frequently embarrassing and always useful, because they are written in the words your customers use rather than the words your industry uses.
Test properly or do not call it a test
A change that coincided with a good week is not a validated improvement. If you want to test, run one change at a time, keep both versions live at once so the traffic is comparable, and work out the sample size you need before you start rather than stopping the moment the numbers look good.
Low traffic sites usually cannot run valid split tests, and pretending otherwise produces confident decisions based on noise. If you get two hundred visitors a month to the page, skip testing and apply the changes that are obviously right: fewer form fields, a clearer headline, proof beside the button, a faster page. Save formal testing for the pages with the volume to support it.
The order to work in
Fix tracking first, because everything after it depends on the numbers being real. Then fix the mismatch between where traffic comes from and what the page says. Then cut form friction. Then rewrite the headline and the call to action. Then add proof beside the ask. Then look at speed and mobile layout. Only after all of that is testing worth the time.
Working in that order matters because each step changes the measurement of the next. Rewriting a button before the tracking is trustworthy means you will never know whether it helped, and that is how businesses end up redesigning the same page every eighteen months.
How Social Signals Marketing approaches conversion work
We treat the site as the final step of the campaign rather than a separate project, because the ad, the video and the landing page either say the same thing or they cancel each other out. Tracking is configured before any design work starts, the enquiry path is built as the shortest honest route from arrival to contact, and changes are made one at a time so we can attribute the result.
See how it comes together in our website design services, read website design that converts visitors into leads, or contact Social Signals Marketing for a conversion audit.
Frequently asked questions
What is a good website conversion rate?
It depends entirely on the business and the traffic source. Local service sites commonly run between two and five percent on enquiry forms, ecommerce sits lower, and a well matched paid landing page can run much higher. The useful benchmark is your own rate last quarter, split by source and device, because a sitewide average across mixed traffic tells you almost nothing.
How do I find out why visitors are not converting?
Combine three sources. Analytics tells you which pages and sources underperform, session recordings and heatmaps show you what people actually do on those pages, and a short exit question tells you what they were thinking. Analytics alone shows the drop off but never the reason, which is why most conversion work that relies on dashboards only ends up guessing.
How many fields should a contact form have?
As few as you need to have a useful first conversation, which for most service businesses is name, contact method and a short description of the job. Every additional field reduces completion, and information like budget, timeline and address can be gathered in the reply. If a field does not change what happens next, it does not belong on the first form.
Is A B testing worth it for a small business?
Only if the page gets enough traffic to reach a valid sample size in a reasonable time, which usually means several hundred conversions rather than several hundred visits. Below that, a test will produce a result that looks convincing and is actually noise. Small sites get far more from applying the changes that are known to work than from trying to measure them.
Does page speed affect conversion rate?
Yes, and the effect is largest on mobile. Visitors abandon pages that take several seconds to become usable, and that loss happens before any of your copy or design gets a chance to work. Speed is worth treating as the floor rather than as an optimisation: get the page fast, then improve what it says.