Suburb pages: what happened when we published six at once

If you publish several suburb pages that say the same thing with a different place name, expect Google to treat them as one page, and expect the whole group to suffer. That is the short version of what happened to a client pest control site in August 2026. The site published six near-identical suburb pages on 1 August. By 7 August the searches it was ranking for in the top 30 had dropped from 43 impressions a day to 15. This post sets out what we measured, what we ruled out, and what we would do differently. It is one site and one cluster, so treat it as a case and not a rule.

What we published

The business serves a regional area. The site was rebuilt in June 2026 and already had a set of suburb pages under /pest-control/. On 1 August we added six more, built from the same template as the existing ones: the same section order, the same service cards, the same review block and the same short FAQ, with the suburb name and a few local details changed. Google crawled them between 2 and 6 August.

What moved on 7 August

We pulled the site's Search Console data by day and by page. Searches where the site held a position of 30 or better went from 43 impressions a day to 15 on 7 August. The drop was not spread evenly:

• Suburb A went from 142 impressions to 2 across the two comparison windows.
• Suburb B went from 82 to 6.
• Suburb C went from 91 to 10.

The searches did not move to another page on the site. They stopped appearing. One older suburb page, edited in the same commit as the new ones, held up. Suburb C, another suburb page and the home page fell even though nobody had touched them that week.

Across three consecutive 28 day windows ending on 4 September, total impressions went from 3,161 to 1,787 to 1,535. Clicks stayed at three or four per window throughout, which tells you how little of that traffic was ever turning into work.

What we ruled out

Before blaming the new pages we checked the other suspects.

• The 28 August outage. The site was caught up in the hosting problem we describe in our post on a host rate limiting a whole account. Impressions for 27 to 31 August were 129, 93, 66, 131 and 114 a day. There was no dip. The cliff was three weeks earlier.
• Seasonality. The same months in 2025 sat at 8 to 40 impressions a day, with no August drop.
• Our own answer block edits. A batch of page edits had gone out on 4 August. The cliff on 7 August hit pages in that batch and pages outside it, and the page that was in the batch and in the cluster held. That is not what an edit problem looks like. We cover how we measure this kind of thing in our post on answer blocks.
• Junk impressions. About a thousand of the earlier impressions came from national searches such as "ant control" plus a city name, at positions 70 to 92. Those were never going to produce a click, and they were leaving anyway. The part that mattered was the top 30 band, and that part fell.

What we concluded

The timing and the pattern pointed at the suburb cluster. Our commit message on 18 August already called the set a doorway cluster, and Google's own guidance treats pages that exist to funnel searches for slight variations of a place as a quality problem. We cannot see inside Google, so this is our reading of the evidence and not something Google confirmed. The facts we can state are the dates, the page-level numbers and the absence of any other explanation that fitted.

On 18 August we rewrote the template to strip the repeated blocks. Two pages then gave us a small test of whether the rewrite helped:

• One page came back on 27 August, the day after Google recrawled the new version.
• Another was recrawled on 22 August and did not come back.

So the rewrite helped one page and not the other, on the evidence we have so far. We would not promise anyone a recovery.

What to check on your own suburb pages

1. Open two of your suburb pages in separate tabs and compare them. If most of the text is the same, a search engine will see that too.
2. Look for repeated blocks that carry the duplication. On the trade sites we measured, the biggest single carrier was a review block pasted in full on every page, followed by an identical set of FAQs.
3. Ask what each page says that only applies to that suburb: the housing stock, the pests or problems you actually see there, jobs you have done. If the answer is nothing, that page has no reason to exist.
4. Publish in small batches. If we had added one page and waited, we would have had a clean read before committing to six.

Our post on the half of local SEO that happens on your website covers the principle, and the numbers above are what it looks like when the principle is ignored.

What this means for a monthly plan

A cliff like this one is easy to miss without someone looking at the data by day and by page. Spotting it, ruling out the other causes and deciding not to keep adding pages is the sort of work that sits under the SEO add-on. Our pricing page sets out the base plan and the add-on. If you want us to look at your own site, the way in is get started.