Perth SEO benchmark
The technical problems that turn up most often on Perth small-business websites, how common each one is, and what each one costs the business that has it.
67 %
Images that make the page jump about as it loads
With no width and height set on an image, the browser cannot tell how much room to leave for it, so everything around it shifts as it arrives. Google measures that shift, and a reader who has just started a sentence loses their place.
- Measured across 6 Perth small-business websites
- Measured on 17 Aug 2026
- Method Argrow's own technical audit, read from each site's own public pages
- Privacy floor Groups of fewer than 5 sites are not published
These are our own measurements. Every figure comes from the same technical audit engine Argrow runs on customer sites, read from each website's own public pages - not survey answers, not a vendor's published figures, and nothing estimated.
What else turned up.
Ordered by the share of sites in this group where each one was found.
- 02
67 %
Images larger than they need to be
An image saved larger than the space it's displayed in adds load time to every visit, on every page it appears on. It's usually the fastest fix on this list, and it's holding back the page speed both visitors and Google notice.
- 03
67 %
Pages that show no preview when someone shares them
These are the tags a social platform or messaging app reads to build the preview card when someone shares a link. Without them the share posts as a bare address with no title and no picture, which very few people click.
- 04
67 %
Server missing some basic security settings
Security headers such as HSTS and a content-security policy are a one-time setup that tells browsers how to handle the site safely. Most website platforms don't turn them on by default, which is why they turn up missing so often.
- 05
50 %
Pages with no main heading
The H1 is the one heading on a page that states plainly what the page is about. Without one, a reader skimming the page and a crawler indexing it are both left to guess.
- 06
50 %
Pages with no search description
With no meta description, Google writes the search snippet itself, usually by lifting a fragment of body text. It's rarely the sentence that would have made someone click.
- 07
50 %
Addresses that redirect more than once before they land
An address that redirects to another, which redirects again before it lands. Every hop adds waiting time for the visitor, and chains tend to grow quietly each time a site is restructured.
- 08
33 %
Links to pages on the site that no longer exist
A link to a page that no longer exists sends a visitor to a dead end, usually at the moment they were trying to go somewhere useful. It also wastes the budget Google spends crawling the site.
- 09
33 %
Pages sharing the same main heading
Pages sharing a main heading read as the same page to anyone skimming, and to a crawler deciding which of them to rank. It is usually a template heading that was never changed per page.
- 10
33 %
Pages sharing the same search description
A description written once and reused across pages describes none of them accurately. Each page is answering a different question, and the snippet is where that difference has to show.
Not faults, but worth knowing.
These are measured the same way as everything above, but nothing here is broken. They are things most sites simply have not done yet, usually because the standard is new.
- 01
100 %
No AI agent catalogue published yet
An agent catalogue is a small file a site can publish to tell AI assistants what the business offers and how to deal with it. The specification is new and entirely optional, so almost nobody has one yet. Nothing is broken without it, which is exactly why it is worth knowing about while it is still early.
- 02
50 %
Business markup that could tell Google more
Schema markup is the structured data under a page that tells Google and AI answer engines what the page actually is, not just what it says. Incomplete markup makes a page harder to summarise correctly, in search results and in an AI answer.
More on how this is measured, and why some groups are held back: the full methodology.
Common questions.
How many sites is this based on?
This group is measured across 6 Perth small-business websites, as of 17 Aug 2026. Groups with fewer than 5 sites are never published, so no single business's website can be identified from the aggregate.
Where does this data come from?
These are Argrow's own measurements, taken with the same technical audit engine we run on customer sites and read from each website's own public pages. Nobody was surveyed, no business took part, and no figure is lifted from another vendor's published data.
How often is this updated?
This page rebuilds from the latest measured data each time the site is published, so the date shown above (17 Aug 2026) is when this group was actually last measured, not a fixed schedule.
Why isn't my city or industry listed?
A group is only published once enough sites in it have been measured to clear the k-anonymity floor and keep individual businesses unidentifiable. As the measured set grows, more cities and industries clear that bar and get their own page.
Is everything on this page a problem?
No, and the page keeps the two apart. Problems are faults: something is set up in a way that costs the business something. Opportunities are things most sites have not done yet, usually because the standard is new, and a site is not worse off for not having one. They are measured the same way and reported separately, because counting a brand-new optional standard as a fault would inflate the problem numbers.
See how your own site compares.
We show you what worked and what didn't. No smoke and mirrors, just real results.
7-day trial · cancel anytime
Free SEO audit - Free report, no payment details needed
- Approve every change
- No lock-in contract
- Proof on every update