SEO reports can be persuasive and still mislead. A rising line in a third-party tool, a single screenshot of a ranking or a traffic number with no source can all look like results without proving much. This is how we measure and report SEO work, and how you can read any report you receive.
Know what each source actually measures
Google Search Console: measured data from Google
Search Console reports real impressions, clicks, click-through rate and average position for your site in Google Search. It is the closest thing to ground truth for Google organic search. Limitations: data is sampled and anonymised for rare queries, positions are averages across locations and devices, and it only covers Google.
Google Analytics 4: what visitors do on your site
GA4 shows sessions, engagement and conversions by channel. It tells you whether organic visitors actually enquire or buy. It depends on correct tracking and, in many regions, on visitor consent, so totals are often lower than reality.
Ahrefs and Semrush: estimates from outside
Third-party tools estimate rankings, traffic and traffic value by combining their own keyword databases with ranking checks. They are excellent for comparing sites, spotting trends, researching competitors and tracking keywords, but their traffic figures are modelled, not measured. In our case studies we label them as estimates every time, for example “Ahrefs estimate, October 2026”.
AI visibility tools: samples of AI answers
Tools that track citations and mentions in AI answers (Ahrefs and Semrush both offer this) run sets of prompts and record which sites appear. Answers vary between sessions, users and locations, so treat these numbers as indicators of direction rather than exact counts.
What we report, and how
- The source and the date for every figure. “#1 on Google UK, Ahrefs, October 2026” means something. “#1 on Google” does not.
- The before and the after, where we have both. Where we do not have a starting value, we say so rather than implying one.
- Measured and estimated kept apart. Search Console clicks and Ahrefs traffic estimates go in different rows, never added together.
- Rankings with context. Search volume, search type (brand or non-brand) and location. Ranking first for your own brand name is useful, but it is not the same as ranking first for the service.
- Business outcomes. Enquiries, calls, sales or sign-ups from organic search, wherever tracking allows.
- What we do not know. Other marketing, seasonality, site changes and Google updates all influence results. Good reports name the factors that could also explain a change.
You can see this approach in our case studies: for example, the Houst study shows positions in October 2026 and notes that starting positions weren’t documented.
Red flags in an SEO report
- Traffic figures with no source.
- Screenshots cropped so the date, location or tool is missing.
- Rankings for long, obscure phrases nobody searches for, presented as wins.
- “Guaranteed” rankings or AI citations.
- A single cause claimed for a change that had several possible causes.
Setting up your own measurement
- Verify your site in Google Search Console and check it monthly.
- Set up conversions in GA4 for enquiries, calls and sales.
- Tag campaign links with UTM parameters so you can separate sources (our UTM link builder makes this easy).
- Pick 10 to 30 priority searches and track them consistently in one tool, in the right location.
- For AI visibility, keep a list of the questions that matter to your business and record the date and exact prompt each time you check.
Every Rank Roots service includes reporting that separates measured data from estimates. If you want a second opinion on a report you have received, send it to us with your website.
