What we found auditing 31 Australian trade businesses

What we found

Over a few weeks I went through 31 Australian trade and service businesses properly. Not a scrape and not a survey. I opened each website and clicked the links, read the page source, checked the Google profile, went through the Facebook page and the ad library, then searched the terms their customers would actually type and looked at who came up instead. About two hundred findings came out of it.

What surprised me is how far down the stack the problems sat. Most marketing writing, including a fair amount of ours, argues about the lead form settings that decide whether your leads are real→, offline conversion tracking→, follow-up sequencing and retargeting windows. Middle of the funnel. Of these 31, exactly 2 had a tracking problem worth naming, and both had the same one.

Meanwhile 13 were running two or more conflicting business names, addresses or phone numbers across their own assets. Fourteen couldn't be found in a search for the thing they sell. Five had a phone or email link that didn't work at all, including a cabinetmaker whose header number carried an extra digit while the footer number on the same site was correct.

That isn't a funnel problem. It's plumbing. And it means a good deal of our own material is written for a business a rung or two above where most of this batch actually sits, which is worth saying out loud even though it costs us the easier pitch.

How we picked them, and what got checked

Twenty six of the 31 were running paid ads, and I came across nearly all of them the same way: scrolling Facebook and stopping on an ad. That's the single most important thing to know about this sample and there's a whole section on it below. The other five I found with no ad involved at all. Two were vans I walked past on Melbourne streets. One was a comment answering a build cost question inside a builders group. One turned up on a Master Builders award listing, and one came through a referral.

Every business got the same pass. The website end to end, including the actual href behind each phone and email link rather than the number printed beside it. The Google Business Profile: review count, category, which services are filled in, how many photos. The Facebook page, the ad library, Instagram where it existed. Then the business name, address, phone and email domain compared across all of those at once, which is a check almost nobody runs on themselves. Then two or three searches a real customer would type, to see who turns up.

None of this comes from an industry report or a category average. Every finding was verified against the live thing on the day. Where something couldn't be confirmed either way it isn't in the count, which is why the review numbers further down rest on 10 businesses rather than 31.

What was actually broken

Fourteen patterns came out of the two hundred findings. Two of them can only apply to a business that's already advertising, so those two are measured against the 26 advertisers rather than all 31. The bars are the share of the businesses each pattern could apply to, and the number beside each one is the raw count.

The creative is the cheapest thing to fix

In 18 of the 26 running ads, the creative fails before you get anywhere near targeting or budget. One business was running a screenshot of its own Google Business Profile dashboard as the ad image, still carrying the words 'Only managers of this profile can see this'. Another was paying to boost a logo on a black background with the page bio pasted underneath as the copy. A third was running a lathe spec sheet, which sells the machine rather than the business, with every contact detail rendered inside the image where nothing is clickable.

Now, these aren't people refusing to spend on marketing. They're spending. The money is going behind an asset that couldn't work no matter how well the account was managed, and nobody has told them, because the agency conversation starts at targeting and the creative arrives as a given. Ten of those same 26 were pointing the spend at a Messenger thread on top of it, which I've written about on its own in A Messenger conversation isn't a lead→.

Conflicting business details was the most common finding of the lot, at 13 of 31, and it's the one nobody discusses. A welding business carried three different addresses across its website, Facebook and Google, a footer email link pointing at an address with no domain suffix on the end of it, and a Facebook contact under a fourth trading name. A plumber's homepage title tag named a different plumbing company entirely. That costs nothing to fix, takes an afternoon, and suppresses local ranking the whole time it sits there, which is most of what Local SEO for tradies→ is about.

Then the pattern worth being generous about. In 8 of the 31, the strongest marketing asset the business owns already exists and earns nothing. One earthmoving contractor's best piece of content was an unpaid post about clearing land before fire season, his own machine working in the scrub, tagged for four towns. It got one like. The ad budget was going behind the logo. The talent isn't missing in these businesses. The distribution is.

Who actually owns the website

Twelve of the 31 have a site that works normally or better, and four of those are genuinely well built. The other 19 have something missing, unfindable, unbuilt, rented, branded as somebody else's, or blocking the crawlers that would find it, and these days that includes the crawlers behind AI answers↗.

A rented front door is more common than a missing one

The middle band is the interesting part. One business's site is a folder on a website builder's domain that returns a 401 when you fetch it directly. Another's ad lands on a page under their parent builder's brand, where the only contact detail is the parent company's email and the footer social links point at the parent's accounts. A third pays for a domain and runs a free-tier advertising banner for the platform across the top of their own homepage.

Three more are on a .com where a .com.au would give them a local ranking signal they're currently giving away for nothing.

None of that is a design problem, and you can have the whole ownership conversation without saying a word about how the site looks. Template websites vs custom built→ goes further into what you actually own at the end of it.

The review numbers have no middle

Review counts could only be verified for 10 of the 31, so this is the weakest number in the piece. It's also the most striking, and How to get more Google reviews→ is the practical half of it.

Either you've never asked, or you're hiding them

Six of the ten sit at four reviews or fewer, two of those confirmed zero. The other four sit at 28, 30, 32 and 214. Nothing in between at all, which splits into two completely different conversations rather than one.

The 214 is my favourite finding in the batch. That business's own website hero says 180+, so they're undercounting themselves by 34 reviews, and there's no review widget anywhere on the site, which means all 214 exist only inside Google. They did the hard part years ago and the buyer comparing three quotes never sees it.

What this sample can't tell you

Thirty one businesses is small, and how I found them matters more than the number. Nearly all of them I came across on Facebook, already running ads, which means advertising was effectively the entry ticket. So this is a sample of businesses spending money on advertising and it can say nothing whatsoever about the ones that aren't. If 26 of 31 looks like proof that everyone advertises, it isn't. It's proof that I was looking where advertisers are.

There's a second layer under that, and it's the part I can't get around. I didn't choose which service businesses Facebook put in front of me. Facebook did. It has watched me stop on trade ads for months and drawn its own conclusions about what I'll engage with, so what I saw had already been filtered by a system optimising for my attention rather than sampling the market. Nobody can correct for that from the outside, including me. The honest description of this sample isn't businesses that advertise. It's businesses an algorithm chose to show one person who keeps looking at them.

The spread isn't representative either. Victoria 11, New South Wales 10, and Queensland, Western Australia, South Australia and Tasmania sharing the last 10 between them. That's where I was looking, not where the opportunity is.

Now the consequence, and this is the part I didn't expect. The five I found with no ad involved, the vans and the awards listing and the group comment, were consistently the stronger operators. Three of the four best businesses in the entire batch run no ads at all. One is a cabinetmaker with 32 five star reviews, 28 years on the tools and a genuinely well built website, and I only found him because I walked past his van.

Facebook showed me people spending money badly. The street showed me people doing excellent work and no marketing at all. That's a different audience and a different message, and it means every count above should be read as of the businesses currently advertising, not of trade businesses generally.

Run this on yourself

The version of this you can do on your own business takes about an hour and costs nothing.

1. Tap your own phone number

   Open your website on your phone and tap the number in the header, then the one in the footer. Five of the 31 had a link that didn't dial, and in every case the number printed on the screen was correct, which is exactly why the owner had never noticed.

2. Put your own details side by side

   Open your website, your Facebook page and your Google Business Profile in three tabs and compare the business name, the address, the phone number and the email domain. Thirteen of 31 were running at least two versions of themselves at once.

3. Search your own business name

   Then search the thing you actually sell plus your suburb. Fourteen of 31 couldn't be found for what they do, and in three of those somebody else was ranking for their own name.

4. Look at what your ad is a picture of

   If it's a document, a logo, a spec sheet or a screenshot, that's the 18 of 26. Then scroll back through your own posts and find the one that did well without a cent behind it. That's usually the ad.

5. Count your Google reviews, then look for them on your site

   Under five and it's the highest leverage free thing available to you this month. Over twenty and not on your website, and you've already done the hard part and you're hiding it.

One more thing

None of the list above needs us, and none of it needs you to buy anything or hire anybody. Most of it runs on software you already pay for or already have: the website, whatever the CRM is, the Google profile sitting there for free. What it needs is an hour and a willingness to look at your own business the way a stranger with no loyalty to it would, which is the genuinely hard part and the reason most of these findings were sitting in plain sight.

If the ads half is what's biting, Why ads stop working for trade businesses→ comes at the same ground from the other direction. And if you want the whole method rather than the audit, it's written out end to end in The Foundational Marketing Blueprint→, free, whether you ever hire anybody or not.