How the work gets done
Four steps: measure where you stand today, diagnose what is holding the site back, prioritise by what is capping it and how long each fix takes to show, then improve and measure again against the same baseline.
GetBorvis / Organic visibilityPositive signals appear to add together. Demotions behave like multipliers applied afterward. If something is capping the site, everything added underneath it is multiplied by close to zero, and the site does not move no matter how good the work was.
That is the honest explanation for the most common complaint in this industry, which is that a business paid for SEO for a year and saw nothing. Often the work was fine. It was just filed behind a cap nobody had looked for.
So we look for the cap first, every time, before recommending a single piece of content or a single link.
We plot your traffic against the published core, spam and site-reputation update calendar, and name the system that most plausibly moved you. Then we record where you appear across Google and the AI engines your buyers actually use, and who is being recommended in your place.
A dated baseline you can argue withWe sweep for anything capping the site first, then check the quality gate, then ask which class of signal is sitting at zero. A drop on a spam-update date is a different problem from gradual erosion, and they need opposite responses.
Causes with their mechanism attachedEvery finding is tagged with the clock it sits on, then ordered by what is capping the site, what it is worth, the risk and how long it takes to show. Slow work starts early, because site-level changes can take months to register.
A sequence, not a listWe fix the highest-value problems, then re-run the same measurement with the same prompts at a fixed interval, with the Google update calendar marked on it. The comparison decides what happens next, including whether to stop.
Work checked against the baselineGoogle does not read every change at the same speed. Sequencing by what is fastest to ship is the intuitive order and usually the wrong one, because it pushes the slowest work to the back of the engagement, where it has the least time to show.
| Order | Work | Why it goes here |
|---|---|---|
| 1 | Remove what is capping the site | Until this clears, other gains are held back. Site-level changes can take months to show, so starting late usually means seeing results late |
| 2 | Quality gate and structure | Also slow, often months. Pruning thin sections usually does more than adding pages, which is the opposite of what most proposals sell |
| 3 | Relevance, answers, sub-question coverage | Roughly a ten-week cycle on our estimate. Started third, it tends to show around the same time as the slower work |
| 4 | Titles, snippets, freshness | Fastest to show movement, four to six weeks, so it goes last and still reports first |
We plan sprints to land before the next expected update and measure at update boundaries rather than at arbitrary month ends, because a month end is not a thing Google recognises.
This is written into scope rather than discovered halfway through. If these do not arrive, the timeline moves, and we will say so at the time rather than at the end.
Prices, outcomes, timelines, process, measurements, named examples. The things only you can state. We cannot invent these and a model cannot either, which is precisely why they work.
Search Console, analytics, and whoever can deploy changes. Read-only is fine for the audit.
Forty minutes to extract facts we can turn into pages. This is the input most engagements drop, and it is the one that decides whether the work compounds.
A ranking can improve while the business gets nothing, and a page can be cited by an AI engine with no ranking at all. So the report leads with four things: how much of the question you cover rather than just the headline query, impressions in Google's generative AI report, branded search over sixteen months, and enquiries that actually arrived.
Position tracking goes in the appendix, on the same query set and the same method every month, with the update calendar drawn on the chart. If a number moved the week of a core update, you should be able to see that yourself rather than take our word for it.
The measurement method, published ↗Not every warning deserves your time or money. Plenty of audit findings are real, technically correct and worth nothing. We separate genuine constraints from harmless noise and say which is which.
If the check finds that your visibility problem is not an SEO problem, or that the smallest useful fix is an afternoon of work rather than a retainer, that is what the report will say. We would rather lose the project than sell a plan we do not believe in.
Start with a dated baseline, the one constraint most likely holding the site back, and the clock it sits on.
Agency rebuilding a client’s site? See the launch handoff, from $1,200. Building something else? See how we partner.
…it was worth gold!!Founder of Notum Learn, a SaaS company we audited