Melbourne Plumber Website Audit 2026

What we checked

We took every plumber that came up in a Google search across ten inner-north suburbs: Preston, Brunswick, Coburg, Reservoir, Northcote, Thornbury, Fawkner, Pascoe Vale, Heidelberg and Bundoora. After removing duplicates, the list held 293 unique businesses.

For each business we checked three things. Does the Google listing link to a website? If it does, does that website load as a working page? And if the site loads, can a customer find a way to make contact?

The checks were passive. We loaded each site once and recorded what we found. We did not submit contact forms and we did not call any business. This is a point-in-time snapshot from June 2026.

One in seven had no website at all

Forty-one of the 293 businesses had no website linked from their Google listing. That is 14%, or roughly one in seven plumbers in the sample.

The practical result is blunt. A customer searches, finds the business, taps through, and lands on nothing. Most people do not investigate from there. They go back to the results and try the next plumber.

That enquiry disappears before the business knows it existed.

Nearly one in five had no working website a customer could reach

The no-website group was not the whole problem. Another sixteen businesses had a listed website that did not work for a reason a customer would also hit: an unresolving domain, a browser security warning, a parked or for-sale domain, a server error, a timeout, or a not-found response.

Counting those alongside the businesses with no site gives 57 of 293. That is 19.5%, or nearly one in five plumbers with no working website a customer could reach from the listing.

An unscreened automated count put the figure higher, at 25%. We do not use that as the headline because it included sites our automated browser could not load because of bot-detection challenges or request blocking in the test environment. A real visitor would most likely load those, so we treated them as not assessable rather than broken.

Working sites still lost enquiries

The part that stood out was not visual design. Of the 221 sites that loaded as working pages, many still made contact harder than it needed to be.

On 21% of working sites, a customer landing on a phone saw nothing on the first screen to tap to call or enquire. The contact details may have been somewhere on the page, but the customer had to scroll or hunt for them. On 14%, the phone number was plain text rather than a tap-to-call link.

Forty-six per cent had no contact form. That sounds larger than it is. Plenty of plumbers take work by phone and may not want a form. A missing form is not automatically a lost enquiry. The stronger signal is the mobile customer who is ready to make contact and has nothing obvious to tap.

What the better sites had in common

The sites that worked well did a few plain things. They loaded. The phone number could be tapped from the first screen on a phone. There was an obvious way to make contact without scrolling or searching.

That was the whole difference. Not a heavier design. Not more copy. Just a shorter path between a person with a plumbing problem and the business that wants the call.

A five-minute check for your own site

If you run a service business, these checks catch the same enquiry leaks we saw in the audit.

• Open your Google business listing and check that the website link is present.
• Tap the website link and make sure the site loads without a browser warning.
• Open the site on your phone. Before you scroll, check whether there is something obvious to tap to call or enquire.
• Tap the phone number. If it does not start a call, the number is probably plain text instead of a phone link.
• Send a test message through your contact form and check that it arrives.

Method

The sample was every plumber returned by a Google search across the ten named suburbs, deduplicated from 600 raw results to 293 unique businesses. This is the set of plumbers a customer discovers through Google in these areas. It is not a census of every plumber that exists.

The checks were passive and automated. We loaded each site once and recorded objective signals. We did not submit forms or call businesses.

"No working website" counts the 41 businesses with no website listed, plus 16 listed websites that failed to load for a reason a customer would also hit: an unresolving domain, a broken security certificate, a parked domain, a server error, a timeout, or a not-found response. Together that is 57 of 293, or 19.5%.

Excluded from that count are 14 sites that could not be loaded by our automated browser because of bot-detection challenges or request blocking in the test environment, plus one site that navigated away mid-check. A real visitor would most likely load these, so we treat them as not assessable rather than broken. Counting them as failures produces the unscreened figure of 25%, or 72 of 293, which we do not use as the headline.

The mobile, phone, and contact-form figures are based on the 221 sites that loaded as working pages. Websites change, so the value is in the fixed method as much as the point-in-time count. The same checks can be rerun on another list of businesses and compared with this result.