113 impressions. Zero clicks.
Aroc reports to its clients every month. This is the same report, about Aroc, published in full. It covers 20 days — 4 to 23 August 2026 — not a calendar month, and it contains no good news that survives inspection. That is the point of publishing it.
Why 20 days and not a month.The reporting cadence set on 5 August defined Month 1 as 4 August to 3 September. This is published on 26 August, and Search Console data lags about three days, so 23 August is the last complete day. Every figure below is a 20-day figure and none of them should be quoted as the month’s final number. The full window is reported in September.
The short version
Impressions were 113 over the 20 days, against 112 in the equivalent 20 days before the site was rebuilt. Effectively flat. Clicks were zero — they were also zero in the prior period, and the site has recorded three clicks in the last ninety days, all of them before the new site went live.
The one real change is compositional. The two retired Phuket pages that used to carry almost all of Aroc’s visibility finished disappearing during this period. Underneath them, eight pages that had never appeared in search at all began to.
The average position improved from 61.77 to 52.66. That is real arithmetic and a misleading summary.
The position improvement is a composition change, not a ranking gain
A badly-ranked group of pages shrank. That moves an average without anything improving.
| Page group | Prior 20 days | This period |
|---|---|---|
| Retired Phuket pages | position 70.42 over 104 impressions | position 68.30 over 56 impressions |
| Every other page | position 20.91 over 22 impressions | position 39.59 over 67 impressions |
| All pages | position 61.77 over 126 impressions | position 52.66 over 123 impressions |
The retired pages fell from 104 impressions to 56, and they sit around position 70. Removing them lifts the site average. Meanwhile the pages Aroc actually wants to rank moved from position 20.91 to 39.59 — the opposite direction.
That second number is a mix effect too, and saying so matters. Eight pages became visible for the first time and most entered weakly — an article at position 90.2, the method page at 70.7, the diagnostic page at 58.0. Adding weak new entries to a small strong set lowers the average without anything having fallen.
This is what an SEO audit cannot tell you. Every figure in the paragraph above is a legitimate SEO metric, and the summary they produce is false. What an SEO audit tells you — and what it cannot sets out three more of the same class.
Two pages did genuinely get worse, and both gained impressions while doing it: the homepage went from position 23.7 over 14 impressions to 42.0 over 22, and the about page from 25.0 over 2 to 43.1 over 9. More impressions at a worse position is a page being shown for a wider, weaker set of queries — what you expect when a site’s positioning has just changed and Google is still working out what the pages are for.
The old traffic cleared, and what is underneath is small
Aroc retired its Phuket pages at the rebuild and wrote down in advance that impressions would fall as a result. They did, inside this period.
- 15 July – 3 August: the largest retired page carried 87 impressions
- 8 – 13 August: 31
- 17 – 23 August: zero
In the final seven days, no Phuket query and no retired page recorded a single impression. The redirects have been reprocessed. That is the cleanest result in the report, and it is a prediction closing rather than a gain.
Those seven days produced 17 impressions and zero clicks across nine pages. With the old traffic gone, that is the honest floor.
The first queries that were neither Phuket nor the brand
The prior 20 days contained none at all. This period contains eleven impressions across five of them. Eleven impressions is noise, and positions 71 to 95 are pages eight to ten, which produce nothing. It is reported because the direction is the first evidence that the repositioning is reaching search at all — not because it is a result.
The query mix moved: Phuket terms fell from 97% of query-dimension impressions to 73%.
The figures
| Metric | Prior 20 days | This period | |
|---|---|---|---|
| Impressions | 112 | 113 | 20 days each |
| Clicks | 0 | 0 | and 0 the period before |
| Distinct queries | 9 | 20 | +11 |
| Pages appearing | 6 | 13 | +7 |
Why the totals in this report do not add up to each other. Search Console counts by date, by query and by page, and the three do not reconcile — it anonymises rare queries and can attribute one appearance differently in each view. This period is 113 by date, 77 by query and 123 by page. Dividing a number from one view by a number from another is wrong, and it produced two incorrect figures in Aroc’s own earlier records before the rule was written down. Every percentage here names the view both its numbers came from.
Week by week
| Window | Impressions | Position |
|---|---|---|
| 4 – 9 August (6 days) | 65 | 63.86 |
| 10 – 16 August | 31 | 44.87 |
| 17 – 23 August | 17 | 41.88 |
Impressions are falling week on week, and that is the intended outcome rather than a failure — it is the retired pages leaving, which was written down before it happened. The three windows sum to 113. The first is six days because the period starts at the rebuild, not on a week boundary.
Where Aroc ranks
Every position is shown with its impression count, because a position without one is the most common honest-seeming lie in this industry.
| Query | Impr. | Position | Can this buy? |
|---|---|---|---|
phuket seo | 13 | 64.2 | Cannot buy |
seo phuket | 13 | 64.4 | Cannot buy |
search engines marketing phuket | 7 | 59.4 | Cannot buy |
aroc | 4 | 50.3 | Brand |
audit of website | 4 | 95.0 | Fits the offer |
growth system | 3 | 82.0 | Fits the offer |
audit website | 1 | 71.0 | Fits the offer |
56 of 77 query-dimension impressions — 73% — are for a category Aroc no longer sells. That is not traffic to be recovered. It cannot buy anything, and it should keep falling.
The line worth acting on is aroc at position 50.3 over 4 impressions. A business that cannot reach page one for its own name has an entity-recognition problem, and that is the one class of query where a small site should win outright.
Being straight about it
- Zero clicks. Not a low number — zero, across this period, the period before, and every week since the rebuild.
- Enquiries are unmeasured.Contact tracking fires, but Aroc’s own analytics property identifier is not recorded anywhere the reporting tooling can reach it. Traffic, source, behaviour and contact events are absent from this report entirely. Stated here rather than quietly left out.
- No conversion target exists, so nothing here can yet be called underperformance against a goal.
- Twenty days is not a month, and 17 impressions in the final week cannot support a trend. Two to three months is the honest window, and this is week three.
- There is no Google Business Profile, so there is no map surface at all.
- No prediction accuracy section, because this is the first report. From September, every prediction made here is revisited — including the wrong ones.
What happens next
- Publish the September report on the full 4 August – 3 September window, so this partial one is superseded rather than left standing.
- Read the canonical-fix experiment to a conclusion. The diagnostic page has held at 3 impressions and position 58 across four consecutive reads. If that holds, it is recorded as a null result rather than extended until it says something nicer.
- Record whether the eight newly-visible pages hold their impressions now that the old traffic has gone. This is the first month that question has a clean answer available.
Two experiments on other pages run to October and nothing will be changed on them before then, because changing a page inside a measurement window destroys the measurement.
Why publish this at all
Every agency says it reports honestly. The claim costs nothing and nobody can check it, because client reports are private.
This is the report Aroc sends its clients, run on Aroc, in public, in a month where the answer is zero. If the format cannot make a bad month legible and honest, it is not ready to carry anyone else’s.
The same discipline applied to a business that is not Aroc is Assess, and the method behind both is the Growth System. The one client month already published in this form is Evergreen Spa.