Website conversion: turning a visitor into a request
Working on conversion is the cheapest way to get more clients. Traffic costs money every month; a fixed form or an added price list works for free from then on. And yet this is usually the last thing anyone gets to.
For a service site a working benchmark is 2–5% of visitors leaving a request. Two things raise conversion fastest: adding prices, even just a "from" figure, and cutting the form down to name and phone. Both take a day and usually deliver more than a full redesign.
In this article8 sections
What conversion is and what counts as normal
Conversion is the share of people who take the action you want. 1000 people came, 20 left a request — conversion is 2%.
There is no universal norm: it depends on the niche, the traffic source and the price. But there are reference ranges.
| Situation | Typical range |
|---|---|
| Service site, search traffic | 2–5% |
| Landing page for advertising | 3–8% |
| Online store | 0.7–2% |
| Visits from maps | 5–15% |
Compare against your own past numbers, not someone else's. Going from 1.5% to 2.5% means two-thirds more requests on the same traffic budget.
Where people drop off most often
The first screen
A person decides in a few seconds whether they are in the right place. If it is unclear what you do and who for, they leave. A heading like «Welcome to our website» says nothing.
No prices
The most common reason for leaving a service site. People want to understand the order of magnitude. When they cannot find it, they go to the competitor who published theirs.
The form
Eight fields instead of two, a mandatory patronymic, a captcha, no explanation of what happens next. Each of these costs you a share of your requests.
Mobile details
A button you cannot hit with a finger, a phone number that does not dial when tapped, a popup covering the whole screen right after entry.
No proof
No examples of work, no reviews, no photos. The person cannot tell whether you are a real company.
What produces the fastest growth
- Add pricesAt least a «from» figure and the order of magnitude. This is usually the single most noticeable change on a service site.
- Shorten the formName and phone. Ask everything else in the conversation.
- Make the phone number tappableOn mobile a large share of contacts happens by call, not by form.
- Add messengersSome people will not call on principle. A WhatsApp or Telegram button removes that barrier.
- Show proofPhotos of real work, reviews with names, certificates. Not stock imagery.
- Explain what happens after the request«We will call within 15 minutes and give you a price» removes the anxiety before the click.
How to find the leaks on your own site
- Session recordings. Watch how people move through the page: where they stop, where they hesitate, where they close it.
- Click maps. They often reveal that people are clicking something that is not a button.
- The funnel in analytics. How many reached the form, how many started filling it in, how many sent it.
- Ask three clients. Ask what was unclear on the site. The answers are usually surprisingly simple.
- Test on a phone. Walk the client's path yourself, from the street, on a poor connection.
How to test changes properly
You changed five things at once and conversion went up — you do not know which one worked. And you do not know whether one of them spoiled the result of the others.
- Change one meaningful element at a time.
- Let data accumulate: at least a week, ideally two.
- Compare comparable periods: weekdays with weekdays, season with season.
- Look at request quality, not just conversion — sometimes more requests means less revenue.
The first screen: what must be visible immediately
This is the most important part of the page and most of the loss happens here. A person answers three questions for themselves in a few seconds.
| The person's question | What answers it | Typical mistake |
|---|---|---|
| Where have I landed? | A heading with the substance of the service | «Welcome» |
| Is this for me? | Specifics: who for, where, in which cases | Vague words about everything |
| What do I do next? | One visible button | Five equally weighted buttons |
The form: how many fields are really needed
Every field is a decision the person has to make. The rule: ask only for what you need in order to call back.
- Always: name and phone. That is it.
- Sometimes justified: a convenient time to call, the city — if you work in several.
- Almost never: patronymic, address, a detailed description of the task, budget.
- Never: a captcha, unless you have a real spam problem. It cuts off people too.
Everything you do not ask in the form gets asked in the conversation — and there people answer willingly, because contact has already begun.
Trust: what builds it
Doubt is the main reason someone leaves a form they were ready to send. It is removed by specifics, not by reassurance.
- Real photos of the work, the premises, the team — not from a stock library.
- Reviews with names and links to a platform where they can be checked.
- Prices, even if they are «from».
- Clear terms: timelines, guarantee, what is included.
- The company's legal details in the footer.
- A stated description of what happens after the request and when you will call.
Conversion rising while request quality falls is a standard trap. You strip every field from the form, requests double, but half are irrelevant and your manager spends the same hours on them.
Frequent questions
Where to start if conversion is low?
With prices and the form. These are the two most common sources of loss on service sites, and both are fixed in a day.
Do we need a redesign for conversion to grow?
Rarely. Targeted changes are usually enough. A full redesign without data lowers conversion more often than it raises it: what worked gets removed along with what did not.
How many requests do we need before drawing conclusions?
At least a few dozen per comparison period. On ten requests any difference is noise.
Do popups help or hurt?
It depends on the timing. A popup that appears immediately on entry is irritating and lowers conversion. One that triggers on exit intent or after the person has read the page works better.
Does live chat help?
In niches where people ask questions before buying, yes — if someone actually answers within a minute. An unattended chat lowers trust more than having none at all.
Should we show prices if they are above market?
Yes, but explain why. Someone who sees the price is already a qualified client. Someone who learns it by phone after waiting leaves more annoyed.
The materials answer general questions. We will look at your specific case — free and without obligation.