Marketing / 02.2 Content Insight
Four decisions from the search data
Including one that argued against my own output.
Clicks
178Detail +
Across the window below.
Impressions
3,990Detail +
At 4.5% click-through.
Avg position
13.4Detail +
Across 284 ranking queries.
Branded clicks
69Detail +
39% of all clicks are people typing the company name.
Pages indexed
79Detail +
Against 77 not indexed, across six reasons.
Google Search Console · nextstarrealty.com.au · 15 Jul – 12 Sep 2026 · captured 15 Sep 2026
Decisions
What the data said, and what changed
01 · Fix before you build
0.8%Click-through on the site’s highest-impression page: 366 impressions, 3 clicks.
Decision +
Google already shows it to the right people. The result itself fails to earn the visit. Decision: rewrite titles and descriptions on high-impression, low-click pages before commissioning anything new.
02 · Know what to ignore
577Impressions from the United States. Clicks: zero.
Decision +
For an agency selling in Brisbane, that is content reaching an audience it cannot serve, and it drags down every site-wide average. Decision: exclude non-target markets from the measure and report Australia and Vietnam separately.
03 · Back what works
10.2%Vietnamese click-through, against 4.6% for Australia.
Decision +
Australia brings volume; Vietnam converts impressions at more than double the rate, because the competing content there is far thinner. Decision: stop treating Vietnamese as the secondary language.
04 · Argue against yourself
109Non-branded clicks. The homepage took 104 of the 178 total.
Decision +
I wrote those articles. The comfortable answer is to write more. The data does not support it: nothing is being clicked, the site has no external links, and the best-ranking pages convert worst. Decision: slow new production. Move to rewriting what ranks, building off-page from zero, and conversion.
Rigour
Three rules, each one paid for
01A summary report is not the truth
Search Console’s indexing report can lag over a week. Comparing it to a sitemap gave a confident, wrong answer about eighteen articles.
02A success code is not proof of a write
The lead pipeline returned success even when the receiving script rejected the record at its first line. The status described the transport, not the outcome.
03A counted event is not an intended one
The form-start event fired when a cursor touched a dropdown. Six of them did not mean six serious attempts.
Next quarter
In order
- Rewrite titles on every page above 100 impressions with sub-2% click-through.
- Start off-page from zero. No external links is the ceiling on everything else.
- Split Australia and Vietnam in reporting, permanently.
- Recover the 31 pages Google found and never indexed.
- Only then commission new articles, against clusters that already rank.
Danny Portfolio 