Why do my Meta ads get clicks but no customers? A break-point test for Inner West businesses

Six links in a chain, and one of them is weak

A paid click is only the first link in a longer chain. The chain runs Ad, Click, Landing page, Enquiry, Booking, Customer. A dashboard that reports on the first two links can say "great" while the phone stays quiet, because the problem lives in the other four.

The useful habit is to stop asking "are my ads working?" and start asking "which link is weak?". An ad that earns clicks is doing its job. If nobody enquires, the page, the form, the offer or the price is failing the person the ad found. Spending more on the ad will push more people into the same leak.

The rest of this guide gives you a test to find the weak link, then walks through the usual causes in order. We work from the front of the chain to the back, because a fault early in the chain hides what is happening later.

The break-point test: read the ratios, find the biggest drop

The test has three steps. First, write down the count at each link for the last 30 days. Second, divide each count by the one before it. Third, look for the ratio that is out of line with the others or with what you expected when you set the campaign up. That is your break point.

Two readings do most of the work. The first is the gap between link clicks and landing page views. A link click is counted when someone taps the ad; a landing page view is counted when the page has loaded. A large gap is the tell for a slow or heavy page: people tapped, saw a blank screen and went back. The second is cost per result against cost per booked customer. A cheap lead that never books is not a cheap lead. The campaign is optimising for the wrong thing.

Use this test sparingly on tiny numbers. With 40 clicks and two enquiries, a single phone call changes the ratio dramatically; wait until each link has a few dozen events before you act on it. That is a rule of thumb, not a statistical threshold.

Objective and tracking: what is Meta being asked to find?

Traffic, Leads or Sales

The campaign objective tells Meta what to look for. In broad terms (check Meta's current descriptions of each objective), a Traffic objective seeks people likely to click. A Leads objective seeks people likely to submit an enquiry. A Sales objective seeks people likely to buy. If you ask for clicks, you get clicks, including from people who tap everything and buy nothing.

That is the most common cause of "clicks but no customers": a Traffic campaign pointed at a business that needs enquiries. Switching to a Leads or Sales objective is only useful if Meta can see the lead or the sale, which brings us to tracking. Third-party analysis (Jon Loomer, March 2025) notes that conversion-led lead optimisation works with the Leads objective and Instant Forms, using data sources such as a CRM, calls or messages. We could not check Meta's own documentation on this, so treat the detail as indicative.

Pixel and Conversions API

For a website, Meta needs to be told when an enquiry happens. The Meta Pixel does that in the browser. The Conversions API sends the same event from your server and can recover events the browser misses. If you use both, Meta's documentation says the server event_id must match the Pixel eventID, and the event_name must match too, and events are only deduplicated if they are received within 48 hours of the first one. Without that, you may be counting each enquiry twice and optimising against inflated numbers. See the tracking checklist to test yours.

Count the right event. A "page view of the thank-you page" is not a lead if the page can be reached any other way, and a "click on the phone number" is not a booking. Choose the event closest to a genuine enquiry, and make sure the number you read in Ads Manager is the same event you count in your own log.

Health and sensitive categories: a privacy caution

This is not a reason to stop measuring. It is a reason to decide what you collect, to avoid sending clinical details to an ad platform, and to check your consent notices. It also affects the remedy for a clinic with clicks but no customers: do not respond by adding more tracking before you have taken advice. Advertising to health patients also carries Ahpra rules; see the dental and medical guide.

Audience, placement and the creative-to-page match

Audience. A broad audience can work if the creative qualifies people on its own. A narrow audience can fail if it excludes the people who buy. For a local business the radius is usually the strongest filter. For a shopfront, test one tight radius (a short walk or drive) against a wider catchment before narrowing by interests, and watch cost per enquiry, not CTR.

Placement. Ads run in feeds, Stories and Reels, and each has a different habit. A tap from a fast-scrolling Story can be accidental. Break results out by placement (Breakdown, then Placement in Ads Manager) and see whether one placement is producing clicks at a low cost and no enquiries. Cut that placement, not the whole campaign.

Message match. The ad promised something. The first screen of the page has about a second to confirm it. If the ad says "Free first consult for Leichhardt physio" and the page opens with a generic homepage banner, a share of visitors leave. Echo the ad's headline, image and offer at the top of the page, and send each ad to a page built for it, not to the homepage.

Mobile speed. Most of the traffic will arrive on a phone. ACMA reports that 97 per cent of Australian adults access the internet via mobile phones, so a page that is only fine on a laptop is failing nearly everyone you are paying to reach. Google's web.dev sets the "good" thresholds for Core Web Vitals at a Largest Contentful Paint within 2.5 seconds, an Interaction to Next Paint of 200 milliseconds or less, and a Cumulative Layout Shift of 0.1 or less, measured at the 75th percentile of page loads. Use these as a measuring stick. Run your landing page through a speed test on a mobile connection, and read the click-to-landing-page-view gap in Ads Manager as the real-world check.

After the click: forms, phones and response time

• Form friction. Each extra field costs enquiries. Ask for name, a way to contact, and one qualifying question. Make the phone and email fields use the right keyboard on mobile. Test the form on your own phone, in the rain, one-handed.
• Phone-first behaviour. For clinics and trades, many people would rather call than type. The Think with Google study of how Australians choose home services (conducted in 2016 and so dated) found 66 per cent said the ability to call a business directly from a mobile search results page was extremely or very important. Put a tappable phone number above the fold on mobile and track calls as their own event.
• Response speed. The same study found Australians consider three providers on average and contact two. If you take a day to reply, the other provider has probably replied first. Set a rule (for example, every enquiry gets a human reply within an hour in opening hours) and measure it.
• Opening hours and the phone. If your ads run on evenings and weekends and nobody is available to answer, the break is in your roster, not in Ads Manager. Schedule ad delivery for hours you can respond, or use a clear after-hours auto-reply with a promised callback time.

Worked examples (illustrative)

Both examples below are composite and modelled to show the arithmetic. They are not results from any real business.

An Enmore Road boutique running a Traffic campaign

A Leichhardt physio running a Leads campaign

When the problem is the offer, not the ad

Sometimes every link works and the chain still produces nothing, because the thing at the end is not wanted at that price, in that place, at that time. Ads do not repair a weak offer. They show it faster and more expensively.

Signs that the offer is the problem:

• Visitors read the whole page (long time on page, scrolling) and still do not enquire, even after speed and form fixes.
• Enquiries arrive, but they ask only about price, and then vanish.
• People say, in replies or calls, that they did not realise what you do, or that a competitor down the strip does the same for less.
• Your best customers did not come from ads in the first place, but from referrals, a GBP listing or a walk-in.

A cheap test: ask five recent enquirers who did not book what stopped them. Then show the page to five people outside your business for 10 seconds and ask what you sell and for whom. If they cannot say, no media plan will fix it. In the Inner West the stakes are higher than average, because the population compares options constantly: Census 2021 data shows 43.8% of residents rent and the median age is 38, so people are used to switching providers, and your nearest competitor is usually one tap away. A clear, specific offer (who it is for, what it costs, what happens next) is the cheapest conversion tool you have.

Limits of this test

We cannot see your accounts here, so every number above is a pattern to check against your own data. Three limits are worth stating plainly.

• Attribution is imperfect. People see a Meta ad, then search your name later, then book by phone. GA4 and Ads Manager will each credit that journey differently, and some of it will not be credited at all. A "how did you hear about us?" question on the form and on the phone fills part of that gap.
• Small samples mislead. Local businesses often have small numbers. Do not rebuild a campaign on the strength of two enquiries.
• Some benefits show later. A campaign that builds familiarity may produce bookings weeks later, through search or word of mouth, which the click-by-click chain will not capture.

On the audit side, an AI agent can check a page, a tracking setup and an ad account against the chain above in a few minutes. A human strategist then decides whether the weak link is a fixable fault or a sign that the offer needs rethinking.

What we would do first, this week

• Pull the last 30 days: impressions, link clicks, landing page views, enquiries (from your own log), bookings and customers.
• Calculate each ratio and circle the biggest unexplained drop.
• Open your landing page on a phone on mobile data. Time it, and check the phone number and form.
• Check the campaign objective. If it is Traffic and you need enquiries, plan a Leads or Sales campaign once your lead event is firing correctly.
• Test your Pixel and Conversions API events and make sure each enquiry is counted once.
• Write a rule for response time and measure the last 10 enquiries against it.
• If you are a health practice, check pixel and consent settings with a privacy adviser before changing anything.