There is a structural problem with how most SEO reporting works, and it makes good work look like no work at all for the first three months of almost every campaign.
Here is the situation. You take on a site that ranks nowhere for its important terms. You fix the technical problems, publish better pages, earn a few links. Over three months a keyword moves from position 74 to position 23.
That is real, substantial progress. It is the hardest part of the journey, and it is the part that predicts everything after it. And in a tool that only reports the first page of results, it shows up as: not ranking, then three months later, still not ranking.
Why tools do this
Not laziness — cost. Reading one page of search results is one request. Reading ten pages is ten requests. A tool tracking a million keywords across its customer base has a very direct financial interest in stopping at page one, and most of them do, quietly, without saying so on the pricing page.
The result is a reporting layer that is accurate about the top ten and blind below it. For a site already ranking well, that is fine. For a site trying to get there — which is every site that needs SEO — it removes exactly the information you need.
What you lose
Four specific things:
- Direction. A keyword at 74 and a keyword at 23 are wildly different situations. One needs a fundamental rethink; the other needs a nudge. Reported as "not ranking", they are indistinguishable.
- Early warning. A term that slips from 8 to 14 has left page one. If your tool stops there, you see it vanish and have no idea whether it went to 11 or to 90. Those need completely different responses.
- Prioritisation. The single best use of a limited budget is finding the terms closest to breaking through — the ones at 11, 12, 13 that need one good page or one decent link. You cannot find those if you cannot see them.
- Client confidence. Three months of reports saying "not ranking" is how good engagements get cancelled just before they start working.
The part nobody mentions
There is a second-order effect worth understanding: tracking depth changes what a keyword costs to check.
If you find your URL on page one, you stop after one request. If it is on page seven, you have made seven. If it is not in the top 100 at all, you have made ten requests to establish that.
Which means keywords you rank well for are cheap to track, and keywords you rank badly for are expensive. That is a slightly uncomfortable incentive — it costs most to measure exactly the keywords you most need to measure — and it is the real reason page-one-only tracking is so common.
We took the other side of that trade. Our tracking reads to the top 100 and prices at one credit per results page read, so the cost is visible rather than hidden in a feature limit. Tracking gets cheaper as your rankings improve, which is at least an incentive pointing the right way.
What to ask your current tool
Three questions that will tell you where you stand:
- What depth do you scan? If the answer is not a number, it is page one.
- Do you store the URL that ranked? Knowing you are fourth is useful. Knowing that the page ranking fourth is a 2019 blog post rather than the product page you optimised is what changes what you do next.
- What happens when a keyword drops out of range? "Not ranking" and "we stopped looking" are different statements, and a tool that conflates them will eventually have you chasing a problem that does not exist.
That third one matters more than it sounds. A tracker that reports a data-source failure as "position: none" will quietly write a fabricated history of a collapse that never happened — and you will spend a week investigating it.
