The proof is
the audit.

askmagno.com, scored like any client. Baseline: 47/100.

On July 24, 2026 we put this site through the exact onboarding we sell: a full GEO and SEO audit, a Search Console baseline, a frozen panel of 15 buying prompts, and a 90-run measurement across Perplexity and Google AI Overviews with screenshots for every run. Then we froze the numbers. They are not flattering, and that is the point: this is what an honest day-one baseline looks like.

Same day, we shipped the first fixes we would ship for you: a complete entity graph with schema on every page, real 404s, images cut from 11.5 MB to 169 KB, the guarantee written into our own terms with its measurement method, and citable pages for each vertical we serve. The panel gets re-run monthly with the same method, and this section will show the deltas, good or bad, with dates.

Same-day update, and the honest part. After shipping that work we re-scored the site twice with the identical rubric: 47 climbs to 62, then to 69 once the guides, the published methodology and the entity work were live. Schema went 37 to 93 and content and trust signals went 40 to 78, because those are what on-site work moves in a day. Brand authority went from 3 to 4 and stopped there, because it measures the footprint we have off our own site, and nothing off-site changes in an afternoon. Our AI share of voice is still 0.0: engines need weeks to re-crawl and re-rank, which is exactly why the guarantee is measured at day 90 and not at day 1.

Every category, so you can rebuild all three numbers

The formula and the weights are on our methodology page. Multiply each category by its weight, add them up, round down. If your arithmetic disagrees with ours, we want to hear about it.

Brand authority is the one that refuses to move, and it is the honest centre of this whole exercise. It weighs 20% and it measures our footprint on other people's websites: reviews, directories, knowledge bases, mentions. A day of on-site work cannot touch it, so it went from 3 to 4 and stopped. That single category is holding roughly nineteen points of composite hostage, and no amount of writing on our own site will free them.

Day five: what actually happened

Numbers, not adjectives. Five days after the work went live, Google had crawled and indexed the new pages: the guides hub, the methodology page, the vertical pages and the first guides all show as indexed in Search Console, with crawl dates between July 24 and July 27. Impressions in those five days: eleven. Clicks: zero.

Eleven impressions is nothing, and we are not going to dress it up. What matters at this stage is which queries they came from, because before the sprint the site barely surfaced for anything: "ai search optimization for gyms" at position 14, "geo for local businesses", "dental seo cost", "get dental practice recommended by chatgpt". Those are buying questions from the panel we froze on day one, and the pages answering them now exist and are indexed. That is the part you can control in week one. Rankings and clicks come later, or they do not, and we will publish either way.

Indexation and impressions read from Search Console on July 29, 2026 · the frozen 15-prompt panel gets its first full re-run in late August

A correction, since we ask you to audit us. We first published this baseline as 48. Re-checking the arithmetic against our own published rounding rule, the six weighted categories sum to 47.95, which rounds down to 47, not 48. We were wrong by one point in our own favour and we are fixing it in public rather than quietly. The frozen evidence file still records what we captured at the time.

Baseline frozen July 24, 2026 at 47/100 · same-day re-scores 62 and 69 · next full measurement late August 2026 · method: fixed 15-prompt panel, 3 runs per prompt per engine, screenshots kept for every run.