Website Traffic But No Sales or Leads? Fix This First

Your report says hundreds of people visited the site last month, maybe thousands. The phone barely rang, and the inbox held a couple of messages, one of them spam. When you have website traffic but no sales, it is natural to suspect one of three things: the visitors are fake, the site is broken, or the money you spend on SEO is going nowhere. Any of those can be true, and they call for very different fixes. The good news is that you can usually tell which one you are dealing with in an afternoon, and the first part takes about fifteen minutes with your own phone. The order you check things in matters more than any single number on the report. Start with the check that costs nothing and can rule out the most expensive mistake.

Your first fifteen minutes belong to the form and the phone number

A form that silently fails, a number that does not dial, or an inbox rule that files every enquiry as junk produces exactly the report you are looking at: real visitors, no leads. It is also the cheapest thing to rule out, so it goes first. Do it from a phone, on cellular data rather than office Wi-Fi, because that is closer to how a visitor away from your office meets the site. Give the test a unique subject line so you can find it later.

If the test form never arrived, stop there. Fix the recipient address, the spam filtering, the CRM connection or the form itself, then send another test until it works end to end. Nothing else on this page matters until that loop is closed, because you cannot judge traffic quality while the route to you is broken.

This check finds obvious faults with the route, the tracking, the intent and the mobile experience. It cannot prove why a visitor left. Treat what it shows as the place to look next, not a verdict.

A zero lead count can be a tracking fault, not a customer problem

Sometimes the leads are arriving and the report simply cannot see them. The form works, the owner answers the phone, and the dashboard still shows no conversions because nothing tells it what a lead is. That makes the site look like a failure when it is really a blind spot.

GA4's enhanced measurement can record two form events on its own. According to Google's enhanced measurement documentation, form_start fires on the first form interaction in a session and form_submit fires when someone submits. Google also cautions owners not to collect personally identifiable information in those events. The catch is that form_submit counts every form, including a newsletter box or a search field. Google's guide on measuring key events explains how to count a specific lead form as its own key event, which is the number you actually want to watch.

Phone calls are harder. A tap on a phone link is not a completed call, and a call is not always a new customer. A simple call log where someone writes down each new caller and how they found you closes that gap without any new software. If your analytics shows almost no visitors at all, or numbers that look plainly wrong, the problem sits one step earlier, and our post on Google Analytics showing zero or wrong visitors covers that check.

GA4 Traffic acquisition shows which visits deserve your worry

Once the route works and the tracking counts real leads, the next question is whether the traffic itself is real and relevant. GA4 can separate a suspicious cluster from the rest of your visitors, as long as you read the right columns side by side. The Traffic acquisition report groups sessions by where they came from, which is the view you want here. Google explains the difference from the User acquisition report in its note on user acquisition versus traffic acquisition.

Know what an engaged session means

GA4 counts a session as engaged when it lasts longer than 10 seconds, includes a key event, or has at least two page or screen views, as described in Google's page on engagement rate and bounce rate. Bounce rate in GA4 is simply the opposite of engagement rate. That definition is mechanical. It does not measure buying intent, and a short visit where someone found your opening hours and left satisfied still counts as a bounce.

Compare the suspect source with everything else

Set the report to session source and medium, then add landing page and device. Look at sessions, engaged sessions, engagement rate and key events together. A source or landing page with a large number of sessions, an engagement rate near zero and no key events stands out against the rest. Google's page on how GA4 defines a session is useful if the counts look odd, and its entrances and exits page shows how to see which pages people land on first.

Treat a dead cluster as a flag, not proof of bots

GA4 automatically excludes known bots and spiders using a list maintained by the IAB. Google does not say that every automated visit is removed, so unknown bots and junk referral traffic can still inflate a report. A cluster that never engages is worth investigating, but it could also be real people who landed on the wrong page. Look at where it came from before deciding which it is.

Search Console shows what visitors were actually looking for

GA4 tells you where visitors came from. Search Console tells you what they typed into Google before they arrived, which is often the clearest answer to why they did not buy. Open the Performance report, select clicks, and filter by the page that gets the most traffic. Then read the queries.

You will usually see a mix. Searches that name your service, your problem or a buying action are the ones that should turn into enquiries. Searches that start with "what is" or "how to" can be perfectly legitimate, but those people are learning, not hiring yet. Searches that have nothing to do with what you sell point to a mismatch between the page and the promise that brought people there. That last group will not become customers however persuasive the page gets.

Expect the totals not to match GA4. Search Console counts clicks from Google Search only, while GA4 counts sessions from every tagged source with its own rules for collecting and attributing visits. Google also notes that filtering by query or page can change totals because some queries are anonymized. Use each tool for what it answers rather than trying to reconcile them to the last visit.

Relevant visitors who stall usually hit friction on a phone

If the route works, the tracking counts real leads, and the queries match what you sell, the problem is more likely friction or doubt on the page itself. This is where speed, mobile layout, form errors and trust come in. The research here is real but dated, and each study measured something narrower than your own result.

The speed figures come from the 2017 Google and SOASTA mobile benchmarks, and the current thresholds from Google's explanation of Core Web Vitals thresholds, which judges real-user field data at the 75th percentile. Use them to decide whether speed deserves a closer look on your busiest mobile entry pages, not to predict how many leads you are losing.

Forms deserve a practical test, not a field count. Fill yours in on a phone with a deliberate mistake, such as a missing digit in the phone number. W3C's guidance on error identification says an automatically detected error should name the field and describe the problem in text, and a form that reloads without saying what went wrong does not meet that. The Baymard checkout research supports removing needless checkout steps, but it does not set a number of fields for a service enquiry. Keep the questions you need to qualify a lead and drop the rest.

Doubt is the other form of friction. A visitor with a high-cost decision may need to see who you serve, what happens after they get in touch, and real evidence of past work. If that evidence includes reviews or testimonials, the FTC's Consumer Reviews and Testimonials Rule questions and answers and its endorsement guidance require them to reflect genuine customer feedback, with any material connection clearly disclosed.

Your lead count needs a comparison that fits your business

A single benchmark is tempting because it seems to answer "is this bad?" in one line. Unbounce's 2024 conversion benchmark report found a 6.6% median conversion rate across 41,000 landing pages, 464 million visits and 57 million conversions. It is a useful dataset, and it also notes wide variation by industry.

That figure is a median among Unbounce's own landing pages, which are often built for a single campaign. Your site may count a lead differently, draw most of its visits from informational blog posts, or sell a service people buy once a decade. A low-volume, long-cycle or high-ticket service can produce a small number of leads from a lot of traffic and still be doing its job. Compare your site with itself over time, by landing page and by source, using the same definition of a lead.

Some fixes should wait until the contact path is proven

It is easy to reach for a redesign, a bigger SEO budget or more paid traffic when leads are low. Each of those can be the right call, but only after the cheaper checks. More visitors sent to a broken form produce more of the same silence, and a new design laid over a failing inbox rule changes nothing.

There is no national dataset that ranks these causes by how often they happen to small businesses. The order below is a diagnostic priority: cheap, testable checks first, subjective ones later.

None of these comes with a promised lift. If the check shows that most of your visitors want information unrelated to what you sell, better persuasion will not turn most of them into customers. The fix then is matching pages to the searches you want, or accepting that some of the traffic is simply readers. If you want a wider list of problems to look for once the basics check out, our post on signs your website is costing you customers walks through them.

One change at a time tells you what actually worked

You do not need an expensive testing tool to learn whether a fix helped. You need one clear defect, a baseline and patience. Pick a single problem you can observe, such as "the mobile phone link fails" or "the lead form does not reach the inbox."

Record the baseline first: the same landing pages, source, device and business outcome over a recent period. Make one change, note the date in GA4 as an annotation or in a simple log, and confirm it with a real submission and a real phone. Keep the old version available in case you need to roll back.

Then wait long enough for the count to mean something, and include a full business cycle of weekdays and weekends. There is no universal number of days or visits that works for every business. A page with five qualified visits a month cannot support a confident answer after one week. Watch qualified leads, not raw sessions, because more traffic after a change is not proof that more people got in touch.

Frequently Asked Questions About website traffic but no sales

Why is my website getting traffic but no sales?

The usual candidates, in the order to check them, are a broken form, phone link, inbox or tracking setup; visitors searching for something you do not sell or not ready to buy; and friction or doubt on the page, often on a phone. Test the contact path first, then read your GA4 sources and Search Console queries.

Can traffic be real and still produce no leads?

Yes. A real visitor may be researching, looking for a different service, or unable to complete the form or call on their device. A visit is not proof of buying intent.

How do I tell whether my GA4 traffic is bots?

Break Traffic acquisition down by source, landing page and device, and compare engagement and key events. GA4 automatically excludes known bots, but not every automated visit, so a cluster with many sessions and almost no engagement is a flag worth investigating rather than proof.

Is a high bounce rate proof my site is bad?

No. In GA4, bounce rate is the opposite of engagement rate, and a short visit where someone found what they needed still counts as a bounce.

What is a good website conversion rate?

There is no universal rate that diagnoses one site. Unbounce reported a 6.6% median for its own landing pages in 2024, with wide variation by industry. Compare your site with itself over time using one definition of a lead.

Do you get leads through your website if you only have a contact form?

You can, as long as the form works on a phone, reaches a monitored inbox, and appears on pages that match what people searched for. A tappable phone number and a clear next step give visitors a second way to reach you.

Moving Forward

When a report shows real traffic and a quiet phone, the answer usually starts with a fifteen-minute check you can run yourself. Prove the form and phone number work, confirm GA4 counts the lead you care about, then use Traffic acquisition and Search Console to separate bots and wrong-intent visitors from real people who stalled. Keep benchmarks in their place, and judge your site against its own history.

Once that loop is closed, the work list gets shorter and cheaper. You stop arguing with visitor counts, missed enquiries stop disappearing, and every fix you pay for is aimed at a cause you can see.

If the check shows relevant visitors, working contact routes and still too few enquiries, that is the point where outside help earns its cost. At Web Leveling, our conversion rate optimization work starts by recording your baseline and reproducing the problem, then states what we will change and reports the result without promising a number. If your check turned up a broken form or irrelevant traffic instead, fix that first, and we will tell you so if you ask. We work with small and medium businesses across the country and overseas. Tell us what your traffic report is showing, and we will help you work out which fix comes first.

Terms