Your analytics say people are visiting. The phone says otherwise. For a small business, that gap is one of the most frustrating states to be in: the site is clearly doing something, because the visits are there, but the calls and form submissions that pay for it never arrive.
Most advice for this situation is written for two other people: the person building a website from scratch, and the person chasing more traffic. This guide is for the third person, the one who already has a site and already has visitors. It walks through a sequence of checks that identifies which of five failures is actually yours, because each failure has a different fix and only one of them is visual.
Key Takeaways
- A conversion problem is a diagnosis, not a mood. Rule out small samples, junk traffic and broken forms before blaming the design.
- The median landing page converts 6.6% of visitors across 41,000 analysed pages. Most visitors never convert, even on healthy sites, so small traffic volumes produce zero-enquiry months by pure chance.
- The wrong visitors cannot be fixed by design. Search intent decides conversion more than layout does.
- 97% of consumers read reviews for local businesses. A service site without visible proof is asking for trust it has not earned.
- Almost every Australian adult now uses a phone to go online. Contact friction on mobile quietly kills enquiries that desktop testing never reveals.
- A redesign changes how a site looks. Most conversion failures are not visual, and a redesign can make things worse.
First, Prove You Actually Have a Conversion Problem
Before changing anything, make sure the problem exists. Three common situations look exactly like a conversion failure but are not, and each one sends owners to the wrong fix.
The first is small numbers. Unbounce’s Conversion Benchmark Report, an analysis of 41,000 landing pages, 464 million visitors and 57 million conversions, puts the median landing page conversion rate at 6.6% across all industries. Sitewide figures for small service businesses are usually lower; 2 to 3% is the range practitioners commonly cite, though published benchmarks for sitewide enquiry rates are thin. Run the maths on your own traffic: at 300 visits a month, a perfectly healthy site produces six to nine enquiries, and plain variance will hand you a zero month now and then. One founder on Indie Hackers documented three weeks of selling with 16 real visitors and zero customers; at that volume the funnel numbers are noise, not signal. If your site gets under a few hundred relevant visits a month, the honest first fix is more of the right traffic, not a new page.
The second is junk traffic. In a Shopify Community thread on heavy traffic with no sales, the opening advice from practitioners is to check visitor geography for bot and test traffic, and to check Search Console queries for intent, before assuming the site is at fault. Sudden spikes from countries you do not serve, or from one odd referral source, inflate the visit count without adding a single potential customer.
The third is the boring technical failure: the form that silently stopped delivering, the notification email that lands in spam, the phone link that dials nothing. An enquiry that never reaches you looks identical in your inbox to a visitor who never enquired. Once a month, from your own phone, submit your form and tap your phone number. You would be surprised how often this test is the whole diagnosis.
Set a baseline before you touch anything: pull the last 30 days from your analytics and write down three numbers: total visits, visits from sources and regions you actually serve, and genuine enquiries received. Every check below compares against those three numbers, and every fix you try afterwards gets judged against them too.
The Five Places a Website Loses an Enquiry
Once the noise is ruled out, a genuine conversion problem sits in one of five places. Each has a different signature in your analytics and a different fix, which is why “10 reasons your website isn’t converting” listicles so rarely help: they hand you all ten repairs when you need one.
| Failure | What you see | What it usually means | First move |
|---|---|---|---|
| Wrong visitors | Traffic steady or growing, enquiries flat | You rank or get shared for curiosity, not buying intent | Review queries and sources before touching pages |
| Page does not say what they came for | Visits end within seconds, little scrolling | The first screen fails the five-second read | Rewrite the hero in the customer’s words |
| Not trusted enough to contact | Visitors read and browse, then leave | Proof is missing, thin or hidden | Put reviews and real work next to the ask |
| Contacting is too hard | Mobile dominates, forms start but do not submit | Friction on the contact path | Cut fields, surface the phone number |
| Promise mismatch | One channel converts, others collapse | The click made a promise the page does not keep | Align each landing page to its traffic source |
The rest of this guide takes each failure in turn, with the two message failures handled together in the five-second test because they are diagnosed and repaired the same way. Work through them in order, because the earlier failures hide the later ones: fixing trust does nothing if the wrong people are arriving, and smoothing the form does nothing if nobody trusts you enough to start it.
Check the Traffic Before You Touch the Design
Design cannot convert people who were never going to buy. If a large share of your visitors arrived out of curiosity rather than need, every other fix in this guide is working on a minority of your traffic.
The cleanest evidence for how much intent matters is the gap between channels. In LocaliQ’s 2026 search advertising benchmarks, drawn from thousands of customer campaigns, the average Google Ads conversion rate across industries is 8.18%. That is several times the 2 to 3% a typical small business site converts overall, and paid clicks are not magic: they are simply pre-filtered for intent. The searcher typed a problem into Google and chose a result that promised an answer. Compare that with traffic from a viral social post or a generic directory listing, where the visitor arrived to be entertained and your page is an interruption.
A founder’s postmortem on Hacker News in January 2026 shows the same thing from the inside. In six months of real numbers, a Product Hunt launch brought about 200 users and not one purchased, with visitors leaving after 30 to 40 seconds, and $12 of Reddit ads produced nothing. The author later reworked the landing page copy and added features before sales began; the product was only taken up once it sat in front of an audience already looking for that kind of tool. Same offering, different visitors, different outcome.
Two checks tell you whether this is your failure. First, open Search Console and read the actual queries bringing impressions and clicks: are they problems your customers have, or topics you happen to rank for? Second, in GA4, segment engaged sessions and enquiries by source and medium. If the traffic with intent converts and the rest evaporates, your problem is upstream of the website. For a locally traded service that usually means tightening your local SEO around the searches buyers actually make, and if you split budget across channels, our guide to choosing between SEO, Google Ads and AI search explains how the intent profile of each channel differs.
The Five-Second Test: Say What the Visitor Came For
Visitors give a new page seconds, not minutes. If the first screen does not confirm they are in the right place, the back button wins, and no amount of work further down the page ever gets seen.
A practitioner in a Shopify Community thread on why stores get traffic but no sales describes the check that matters: take the exact search term, post or ad that brought the visit, then read your landing page as that person. If the page does not answer the question that drove the click within a few seconds, the visit ends. The 30 to 40 second exits in the Hacker News story above are what that failure looks like in analytics.
The recurring pattern on small business sites is a first screen that talks about the business instead of the visitor. “Welcome to our website, established 2009, Adelaide’s leading provider of integrated solutions” tells a stranger nothing they came to learn. A homeowner with a blocked drain does not read “premier plumbing solutions” and feel found. They read “blocked drain cleared today, $0 call-out” and pick up the phone.
Three repairs cover most of this failure:
- Name the visitor’s situation first. The first screen should say who this is for, what you do for them, and where you do it, in words a customer would use to describe their own problem.
- Give the page one job. Quote, booking and phone number competing for attention means none of them wins. Pick the single action that matters and make everything else secondary.
- Make the button name the outcome. “Get a quote this week” tells the visitor what happens next. “Submit” tells them nothing.
One calibration from the founder community is worth keeping. In an Indie Hackers thread from a founder with 35,000 content views, 1,500 site visits and zero customers, the replies kept returning to positioning rather than polish, and one reframed the metric entirely: a first visit without contact is normal, because people compare; zero return visits is the real alarm, because it means nothing about the page was worth coming back for.
Trust Is the Gatekeeper for Enquiries
An enquiry is an act of trust. The visitor is about to hand a stranger their problem, their phone number and often their address, and before they do, they look for proof that other people already have.
That checking behaviour is now close to universal. In BrightLocal’s 2026 Local Consumer Review Survey of 1,002 US adults, published in February 2026, 97% of consumers read online reviews for local businesses, and 41% say they always read them when browsing, up from 29% the year before. Your visitors are looking for vouches before they commit. A service site with no visible proof is asking for trust it has not earned, and in 2026 that is a defect, not a missing extra.
The trust inventory for an Australian service business is short and concrete:
- Google reviews near the ask. Not a testimonials page nobody visits: the two or three strongest quotes sitting beside the contact form. With 71% of consumers using Google for reviews, in the same BrightLocal survey, your Google Business Profile is the de facto proof source, so keep it current and link to it.
- Real photos of real work. Visitors read stock imagery as “not us”. Actual jobs, actual team, actual ute in the driveway.
- Local anchors. A suburb-level service area, a local phone number, an ABN in the footer. Small signals that a real local business stands behind the page.
- A response promise. “We reply within one business day” turns a form from a black hole into a commitment.
Practitioners make the same point from the ecommerce side: in the Shopify thread above, the advice is to get the first 10 to 15 genuine reviews in place before expecting conversion. The sequence matters, proof first, then traffic converts. It also explains why copying a competitor’s layout rarely works: you inherit their look without inheriting their reviews.
Contact Friction: Mobile, Forms and the Phone
Most of your visitors are holding a phone. ACMA research published in February 2026 found 97% of Australian adults accessed the internet via mobile phone during 2025, with 92% doing so daily. Your contact path therefore has to work one-handed, on a small screen, possibly standing in a carpark. A form that is merely annoying on a desktop can be impossible there.
On speed, the thresholds worth aiming at are Google’s Core Web Vitals, measured at the 75th percentile of real visits: Largest Contentful Paint under 2.5 seconds, Interaction to Next Paint under 200 milliseconds, and Cumulative Layout Shift under 0.1. Any page can be checked free in PageSpeed Insights. One honest caveat: the confident statistics linking each second of load time to an exact percentage of lost sales come mostly from vendor case studies, not controlled experiments. The direction of the effect is well supported; the exact magnitude on your site is not predictable from someone else’s numbers.
On the form itself, every field is a decision the visitor can say no to. To reply to an enquiry you need a name, one contact channel, and a sentence about what they need. Everything else, suburb, budget range, “how did you hear about us”, is a tax on the people most ready to talk. Collect it on the call instead.
And do not bury the phone. For most trade and service businesses the call is the highest-quality enquiry, yet many sites hide the number on a contact page. Put it, tap-to-call, in the header of every page. Then run the monthly test from the first section: submit the form, tap the number, confirm both reach a human.
If your diagnosis lands here, the fix list gets specific quickly. Our guide to mobile UX fixes that lift conversion walks the mobile contact path element by element.
Why a Redesign Probably Will Not Fix It
When enquiries are flat, a redesign feels like decisive action. It is also the most expensive way to avoid finding the actual cause. Look back at the five failures: four of them are not visual. Wrong visitors stay wrong after a redesign. Missing reviews stay missing. A seven-field form stays seven fields. Repainting the shopfront does not change what is on the shelf.
A redesign can also subtract. One store owner in a CRO question repository on Shopify Community reported conversion dropping after a migration and redesign, and CRO practitioners argue the same from the other side: a redesign alone does not lift conversion, because it does not change traffic quality, offer clarity or trust. A rebuild that changes URLs without careful redirects can also cost you the search rankings that supply the traffic, turning a conversion problem into a traffic problem as well.
Sometimes a rebuild is the right answer, and the diagnosis is what tells you: when the site is genuinely unusable on phones, when there is no room in the layout for proof, when the platform will not let you edit what needs editing, or when the messaging is so scattered it needs re-architecting rather than rewriting. If that is where you land, scope the work against what a conversion-ready site actually includes in our small business web design guide, and price it against what professional web design costs in Adelaide in 2026 before signing anything. Diagnosis first, then the smallest fix that addresses it, redesign last.
Frequently Asked Questions
What is a good conversion rate for a small business website?
For a service business measuring enquiries against all visits, 2 to 3% is the range practitioners commonly cite, and the median landing page across 41,000 analysed pages converts 6.6%. Treat both as context, not targets: a site converting 2% of high-intent local search traffic may be healthier than one converting 4% of mostly irrelevant visits. The trend against your own 30-day baseline matters more than any published average.
Why did my enquiries drop after launching a new website?
Check in this order: the form and phone links still work and deliver, your search rankings and traffic survived the launch, and only then whether the new messaging converts as well as the old. Broken forms and lost rankings account for most post-launch drops, and both look like a conversion problem until you test them.
How long should I wait before judging a change?
Long enough to collect a few hundred relevant visits, which for most small business sites means weeks, not days. Judging a rewrite after 40 visits is the small-sample trap in reverse: you will credit or blame the change for what is really just variance.
Will a website redesign increase conversions?
Only when the diagnosis shows the page itself is the blocker: unusable on mobile, no room for proof, or a structure you cannot edit. A redesign does not change who arrives, whether they trust you, or how hard it is to make contact, and a poorly managed one can reduce conversion and search traffic at the same time. Run the five checks first; they cost nothing but an afternoon.
Where to Start This Week
The order of operations matters more than the effort. Pull your 30-day baseline. Rule out junk traffic and test your own contact paths from a phone. Then walk the five failures in sequence: intent, first screen, trust, friction, promise match, and stop at the first one that is clearly yours. That is the fix to make, and it is usually cheaper and faster than the fix you were about to buy.
The short version: a website that gets traffic but no enquiries is not one problem, it is five possible problems wearing the same symptom. Prove the problem is real, find which failure is yours, fix that one, and let a redesign be the last resort rather than the first reaction.
