Website Getting Traffic but No Leads? How to Tell a Traffic Problem From a UX Problem
A step-by-step diagnosis for a site with visitors but no enquiries: check tracking, traffic intent, device gaps and user behaviour before buying more traffic.
Traffic problem or UX problem: the short answer
If your commercial pages convert reasonably but get few relevant visits, you have a traffic problem. If they get a fair number of well-targeted visits and still produce almost no enquiries, you have a page, offer or UX problem. Before deciding either, prove that your form actually delivers and that your analytics records leads, because a broken form or a missing tracking event looks exactly like a website nobody wants to contact.
This guide is for founders and marketing leads of service businesses, B2B companies and SaaS products whose sites get visitors but few enquiries. It sets out six checks in the order that saves the most time, then a verdict table to decide whether to spend on marketing or on the site. If you are more concerned with the quality of the leads you do get, read how to improve enquiry forms without attracting low-quality leads.
Key takeaways
- Test the plumbing first: submit real test enquiries and confirm the GA4 generate_lead key event fires once per submission.
- Split traffic by landing page and channel. Informational blog traffic rarely converts and should not drag down your view of the service pages.
- Separate volume from rate. Few visits to commercial pages is a traffic problem; plenty of visits with no enquiries is a page problem.
- Compare devices on the same landing pages. A mobile-only gap points to layout, form or speed issues.
- Use behaviour signals such as rage clicks, dead clicks and quick-backs to find specific broken or confusing elements.
- Watch five people try to enquire. It is cheap and usually explains the numbers.
Step 1: Prove the form delivers and the lead event fires
Many "no leads" problems are measurement or delivery problems. A form that sends to an old address, a notification email filtered as spam, or a tracking event that stopped firing after a template change will all show traffic with no leads. Spend an hour ruling this out before anything else.
Google Analytics 4 has a recommended event for this, `generate_lead`, which Google describes as measuring "when a lead has been generated (for example, through a form)". To count it as a conversion in reports, mark it as a key event; Google describes key events as a way to "measure actions that are important to your business".
- Submit a test enquiry on a phone and on a desktop. Confirm each arrives in the inbox or CRM, and check spam, shared inboxes and any auto-forwarding rules.
- Confirm the person who should reply actually receives the notification, not just a generic address nobody watches.
- Open GA4 DebugView (via Tag Assistant or debug mode) and check that generate_lead fires once per successful submission: not on page load, not on a failed validation, and not twice.
- In GA4 Admin, check that generate_lead is marked as a key event. Newly marked key events can take up to 24 hours to appear in standard reports.
- Compare GA4 lead counts with the enquiries actually received over the same month. A large gap means the data cannot be trusted yet.
- If people also enquire by phone, email links or WhatsApp, track those clicks as key events too. Otherwise the site will look as if it converts less than it does.
Pro tip
Once leads are reliably tracked, GA4's lead-lifecycle events (qualify_lead, disqualify_lead with a reason, close_convert_lead) let you measure lead quality, not just volume. That matters when a fix increases enquiries but they turn out to be poor fits.
Step 2: Check traffic quality by landing page and channel
Total sessions mix very different visitors. Someone reading a how-to article has different intent from someone landing on your pricing or service page from a search for your service. Judge each group separately.
In GA4, open Reports > Engagement > Landing page, which shows the first page of each session with sessions, engagement time and key events by default. Add Session source / medium as a secondary dimension to see which channels land on which pages. Then sort your landing pages into two groups: informational (blog posts, guides, glossary pages) and commercial (service, pricing, product, case-study and contact pages).
Read engagement with GA4's own definitions. An engaged session is one that "lasts longer than 10 seconds", "has a key event" or "has 2 or more screen or page views". Engagement rate is the percentage of engaged sessions, and bounce rate is "the percentage of sessions that were not engaged". So a high bounce rate on a short article that answers a question is not necessarily a failure. A high bounce rate on a service page that paid traffic lands on deserves attention.
| Pattern on the same landing page | What it suggests |
|---|---|
| Low engagement from one channel, normal from others | That channel's targeting or promise does not match the page (a traffic quality issue) |
| Low engagement from every channel | The page itself: message, layout, speed or relevance |
| Good engagement but almost no key events | The offer, the call to action or the form is the blocker |
| Most sessions land on informational pages | Little commercial traffic; the overall conversion rate is diluted rather than broken |
Step 3: Separate volume from rate
Once traffic is split, ask two questions about the commercial pages only. How many sessions do they get? And what share of those sessions produce a lead? A low number of leads can come from either, and the fixes are different. Low volume is a marketing problem: search visibility, ad targeting, partnerships. A low rate on decent volume is a page and offer problem.
Hypothetical example: a consultancy's site has 6,000 sessions a month and 10 enquiries, a site-wide rate of about 0.17%. Split by landing page, 5,200 sessions land on blog articles and produce 1 enquiry, while 800 land on service pages and produce 9, a rate of about 1.1%. The service pages are not obviously broken; there are simply not many visits to them. The better spend is on bringing relevant visitors to the service pages and linking blog readers through to them, not on redesigning the service pages first.
Be careful with small numbers. When a page produces a handful of enquiries a month, one extra lead changes the rate a lot, so look at several months and do not over-read week-to-week movement. The same limit is why most small lead-generation sites cannot run reliable A/B tests; a CRO plan for low-traffic websites explains what to do instead.
Step 4: Compare landing pages by device
Take your top commercial landing pages and compare engagement rate and key-event rate on mobile against desktop. If desktop converts and mobile barely does on the same page with the same traffic sources, the cause is almost always on the page: a form that is awkward on a small screen, a call to action pushed far down, a sticky element covering content, or a slow mobile load.
For speed, use field data rather than a one-off lab test. Google's Core Web Vitals treat a page as good when Largest Contentful Paint is 2.5 seconds or less, Interaction to Next Paint is 200 milliseconds or less and Cumulative Layout Shift is 0.1 or less, measured at the 75th percentile and split by mobile and desktop. If mobile fails where desktop passes, see why a website is slow and how to find the cause.
Step 5: Read behaviour signals
Analytics shows where people stop; session tools show what they were doing when they stopped. Microsoft Clarity is a free tool that flags several behaviours automatically, and its documentation defines them precisely (Microsoft Clarity: semantic metrics, vendor documentation for its own product). Filter recordings to your commercial landing pages and to sessions with these flags, then watch a sample rather than relying on the counts alone.
| Signal | Clarity's definition (paraphrased) | What it often points to | What to check |
|---|---|---|---|
| Rage clicks | Multiple clicks in a clustered area in rapid succession | Something looks clickable but is not, or responds too slowly | Images, cards, icons and text styled like buttons; slow form submit |
| Dead clicks | A click that gets no feedback in a reasonable time: no visual change and no navigation | Clarity lists "broken elements, high latency requests, or misleading UX" | Broken links and buttons, unstyled disabled states, slow scripts |
| Quick-backs | Leaving for another page and returning faster than that page's dwell-time threshold | The link or menu item promised something the destination did not deliver | Navigation labels, link text and the pages they lead to |
| Excessive scrolling | More vertical scrolling than expected for the session | People are hunting for something: price, proof, contact details | Whether key information is easy to find; page order |
Worth noting
Session recording tools collect behavioural data. Mask form inputs, mention the tool in your privacy notice and make sure the tool respects your cookie consent settings.
Step 6: Watch five people try to enquire
Numbers tell you where; watching people tells you why. Give a small number of people who match your buyers a realistic task ("find out whether this firm could help with X and get in touch") and watch without helping. Ask them to think aloud.
Jakob Nielsen's well-known guidance at NN/g is that testing with five users finds about 85% of usability problems, but that figure rests on a model assuming each user uncovers about 31% of problems. The article also says it holds only "for comparable users who will be using the site in fairly similar ways"; with two or more distinct audiences, test three or four people from each. It recommends several small rounds with fixes in between rather than one large study. Five-user tests are qualitative: they reveal problems, not conversion rates. For how much design and testing to do before building a new site, see how much UX design to do before development.
Is it a traffic problem or a UX problem?
Use this once Step 1 has confirmed your data is trustworthy.
| What you see | Likely verdict | Who usually owns the fix |
|---|---|---|
| Leads in the inbox do not match leads in GA4, or the test enquiry never arrived | Measurement or delivery problem. Fix this before judging anything else | Developer or whoever manages tracking and email |
| Commercial pages convert reasonably but get few sessions | Traffic problem | Marketing: SEO, paid search, partnerships, internal links from content |
| Most traffic lands on informational pages and rarely moves on | Traffic mix problem, plus weak routing from content to services | Marketing and content, with small page changes |
| Poor engagement from one channel only, on pages that work for other channels | Traffic quality problem for that channel | Whoever runs the channel: targeting, keywords, ad copy |
| Poor engagement from all channels on the same commercial page | Page problem: message, relevance or speed | Design and content |
| Good engagement, very few form submissions | Offer, call-to-action or form problem | Design, copy and development |
| Desktop converts, mobile does not, on the same pages | UX or performance problem on mobile | Design and development |
| Many rage clicks, dead clicks or quick-backs on key pages | UX problem at specific elements | Design and development |
| Leads arrive but rarely become customers | Lead quality or follow-up problem, not a website-volume problem | Sales process, form qualification, targeting |
If conversions dropped after a redesign or migration
A sudden fall right after a launch is usually something specific that broke, not a sudden change in demand. Check in this order.
- Tracking: the GA4 tag loads on every new template, there is only one copy of it, and generate_lead still fires once per submission. Rebuilt forms and new thank-you pages often break the old trigger.
- Key events: the event name is unchanged and still marked as a key event, and any Google Ads conversions built on it still import.
- Delivery: the new form sends to the right inbox or CRM, and notification emails are not landing in spam after a hosting or email setup change.
- Consent: a new or reconfigured cookie banner can change how many sessions analytics records. Compare against enquiries actually received, not GA4 alone.
- Redirects: old URLs that had search traffic, backlinks or ad landing pages redirect (301) to the closest matching new page, not to the homepage or a 404.
- Indexing: no leftover noindex tags or robots.txt rules from the staging site. Check Search Console's page indexing report.
- Removed content: testimonials, client logos, phone numbers, pricing guidance, FAQs or service detail that existed before and quietly disappeared.
- Changed paths: calls to action, menu items and internal links still lead to the right pages, and the contact page is as easy to reach as before.
- Mobile and speed: the new templates pass Core Web Vitals on mobile and the form works on a real phone.
- Fair comparison: compare the same weeks year on year, or the same number of weeks before and after, and allow for seasonality and campaign changes.
Fixing the page: message, trust, form and follow-up
If the verdict points at the site, these are the usual fixes, roughly in order of effort.
Message and message match. Within the first screen, a visitor should understand what you offer, who it is for and what to do next. If they came from an ad or a search result, the page should repeat that promise in recognisable words. Generic headlines that could belong to any competitor are a common reason a page "looks fine" and still does not convert. A purpose-built page usually beats sending paid traffic to the homepage; see what a well-built landing page needs.
Trust and the call to action. People hesitate to contact an unfamiliar business. Real evidence helps: named people, credentials, specific work you can show, clear contact details and an explanation of what happens after they get in touch. The website trust and credibility checklist covers the signals worth checking. Give one clear primary next step per page rather than several competing ones.
The form. Ask only what you need to respond well, explain why you ask for anything sensitive such as a phone number, and keep labels visible above the fields. Removing fields can raise volume but lower quality, so make the trade-off deliberately; the guide to enquiry forms works through each field.
Follow-up. If enquiries arrive but are answered slowly or not at all, the website is not the bottleneck. Decide who owns new enquiries, how quickly they should be acknowledged, and how outcomes are recorded so you can tell good leads from poor ones.
An older companion article covers common reasons a website is not converting from a broader angle.
Common mistakes
- Redesigning the whole site before testing the form and the tracking.
- Judging the site on one blended conversion rate that mixes blog readers with buyers.
- Treating GA4 bounce rate as a verdict on its own, without comparing like-for-like pages and channels.
- Buying more traffic to a page that does not yet convert the relevant visits it already gets.
- Changing many things at once, so nobody knows which change helped or hurt.
- Running A/B tests with too few enquiries to reach a reliable result.
- Counting leads without recording which ones turned into customers.
Limits of this diagnosis
This process finds where enquiries are lost; it does not fix a weak offer. If the price, positioning or the service itself is wrong for the people visiting, better pages will not compensate, and customer conversations or sales-call notes will tell you more than analytics. Very low-traffic sites may not have enough data for Steps 2 to 4 to be meaningful; lean more on Steps 1, 5 and 6. And some businesses win most work through referrals or outbound sales, where the website mainly reassures people who already intend to contact you. In that case the number of form leads is the wrong measure of whether the site works.
If you would rather have someone independent run the checks, a CRO audit follows broadly this order: measurement, segmentation, behaviour and user research, then a prioritised list of fixes.
Sources
Google Analytics: GA4 recommended events reference (generate_lead and lead-lifecycle events); GA4 Help: key events; GA4 Help: DebugView; GA4 Help: engagement rate and bounce rate; GA4 Help: Landing page report.
Behaviour and usability: Microsoft Clarity documentation: semantic metrics (vendor documentation); Jakob Nielsen, NN/g: why you only need to test with 5 users (2000).
Performance: web.dev: Web Vitals.
Accessed 14 October 2026. Analytics and tool interfaces change; menu names and report locations may differ from those described here.
Conclusion
Traffic without leads has a findable cause, and the order you look in matters. Prove the form and tracking work, split traffic into informational and commercial, separate volume from rate, compare devices, then watch behaviour and real people. The verdict table will usually tell you whether your next budget belongs in marketing or in the site. Start with Step 1 this week; it takes an hour and rules out the most embarrassing cause.
Want an independent read on where your enquiries are going?
ZSpace Labs is a remote-first technology studio that works with clients globally on CRO audits and UX and page redesign. If useful, we can talk through your analytics and the checks above, and tell you whether the evidence points at traffic or at the site.
Common questions.
It is rarely one cause. The usual suspects are a form or tracking setup that is quietly broken, traffic that arrives with research intent rather than buying intent, a page that does not make the offer clear, missing trust signals, or a form that asks too much. Checking them in a fixed order stops you fixing the wrong one.
Look at where the low numbers sit. If your commercial pages convert reasonably but get few visits, or poor engagement is confined to one channel, it is mostly a traffic problem. If commercial pages get decent, well-targeted visits from every channel and still produce no enquiries, or one device performs far worse than another, it is mostly a page or UX problem.
Usually not on its own. If the site is not converting the relevant traffic it already gets, more of the same traffic tends to scale the problem. The exception is when commercial pages already convert and simply receive too few visits; then acquisition is the right lever.
Not by itself. In GA4, bounce rate is the share of sessions that were not engaged, meaning they lasted 10 seconds or less, had no key event and viewed only one page. A blog post that answers a question quickly can have a high bounce rate and still be doing its job. Compare bounce and engagement rates for similar pages, channels and devices instead.
Use the GA4 Landing page report with session source / medium added as a secondary dimension. If most sessions land on informational pages, or engagement is poor for one source while other sources do fine on the same page, the traffic is the issue rather than the page.
Yes. If a visitor cannot quickly tell what you offer, who it is for and what happens next, they leave without acting. A mismatch between the promise in an ad or search result and the page it lands on has the same effect.
It can. A slow page loses visitors before they see the message or the call to action. Check Core Web Vitals field data separately for mobile and desktop, because a mobile-only slowdown can look like a general conversion problem.
First prove the form delivers by submitting test enquiries on mobile and desktop. Then compare form starts with submissions if you can track both, watch a few session recordings of people using it, and question every required field. Our guide to enquiry forms covers field-by-field decisions.
Not necessarily. One clear, well-placed next step usually works better than several competing ones. The problem is more often an unclear or poorly placed call to action than too few of them.
Yes. If a meaningful share of traffic is mobile and the mobile layout, form or speed has friction the desktop version lacks, leads can suffer in a way that desktop-only testing never reveals. Splitting results by device is one of the quickest ways to spot it.
A real one. If enquiries are captured but land in an unwatched inbox, the spam folder or a CRM nobody checks, or are answered days later, the website may be working and the process after it is not. Rule this out before redesigning anything.
Check measurement first: rebuilt forms and templates often stop the lead event firing or stop emails being delivered, which looks like a conversion drop. Then check redirects, accidental noindex or robots rules, and content that was removed in the redesign, such as testimonials, contact details or pricing information.
Long enough to cover normal weekly cycles and a reasonable number of enquiries, and compare like-for-like periods to allow for seasonality. On a low-traffic site you will rarely reach statistical certainty, so judge fixes directionally and alongside what you learned from watching real users.