This is our third engagement together: the Care Package through 2023 and 2024, the year of Google Maps work through 2025 (monthly updates, wrapped with the Maps performance snapshot in February), and then the Content Refresh Challenge: 80 pages across a 15-cluster topic map and 30 priority keywords, built to restart a lead pipeline that had gone quiet.
Delivery ran in batches through to July: landing page and blog refreshes, then 47 new articles as Google Docs, which you loaded, merged and published yourself, confirming the full set was live on 13 July. The May status report marked the midpoint: 61 of 80 delivered, average position across the set already up 14 places.
Two more numbers from the delivery window worth having on record: organic users roughly 2.5x'd while the content was landing (242 in March to 618 in June on our tracking), and across the last sixteen months 475 visits arrived from AI assistants, with ChatGPT more than half of them. One fair caveat on the impressions story below: the site also went through your own URL restructure in February and March, so the year-on-year lines carry that change as well as ours.
Between February and June 2026, with all 80 pages loaded, the tracked keyword set moved like this:
| Measure | February 2026 | June 2026 |
|---|---|---|
| Keywords ranking (of 30 tracked) | 11 | 21 |
| Keywords in the top 10 | 2 | 8 |
| Average placement across the set | 42.3 | 21.2 |
| Pages earning impressions on Google | 181 | 554 |
And on the current site structure, impressions went from 43,390 in January to 123,888 in August, up 186%, with clicks up 61%. The content did what it was supposed to do.
Delivery finished and tracking stopped in June. The live check on 2 September shows the cost of that:
| Measure | June 2026 | Live, 2 September |
|---|---|---|
| Keywords ranking | 21 of 30 | 15 of 30 |
| Keywords in the top 10 | 8 | 4 |
Eight keywords have dropped out of the top 100 entirely since June. Search positions are not a state you reach, they are a position you hold, and nothing has been holding them.
Two things need saying plainly, because they change how the headline numbers read.
First, one article was doing a lot of work. A single Windows troubleshooting post drove 853 clicks, 27% of all clicks, in May to August last year. It has since been removed and redirected. Correcting for it takes the year-on-year decline from a frightening 45% to a more honest 25%.
Second, the non-branded footprint was always thin. Excluding that one article, non-branded clicks fell from 320 to 112 across the same four months. The engagement built ranking positions across a much wider set of pages, and that width has not yet converted into commercial search traffic.
20 of your 30 tracked searches now show a Google AI answer above the results, and you are cited in one of them: "smb1001 certification", where you also hold organic position 5.
That is a real foothold, and it is worth knowing what produced it, because it points at exactly what a next engagement should scale.
| Move | Why |
|---|---|
| Page two to page one on the commercial cluster | "IT support christchurch" sits around 19, "IT services christchurch" around 21, "IT support" around 26. These are the terms that turn into enquiries and they are all one push from page one. |
| Authority | DR 27 with no link building in three years. It is the ceiling on everything above. |
| An AI citation programme | You are cited in one AI answer of twenty. Modelling the rest on the page that already earns one is the most direct route. |
| Monthly measurement | The decay above happened because nothing was watching. Whatever comes next should be measured monthly rather than reviewed at the end. |