A tutoring startup with a brand-new domain entered a market owned by giants. Its organic Google impressions ran 5× above the prior year in July, 24× in August, and the first twenty-five days of September brought 4.7× the impressions and 5.5× the clicks of the whole of August. In AI search, over the 30 days to 27 September, the engines that mention Genie Teacher place it at an average position of 1.5. Here is the work behind that curve, with the numbers and the caveats.
A note on scale first. Genie Teacher is early-stage. The base is small, so the multipliers run hot. We flag below what is momentum, what is branded demand, and what is still to be earned.
The client
Genie Teacher connects families with certified teachers for tutoring. Not gig tutors, not anonymous marketplaces: actual certified teachers. The audience is parents in Canada, heavily Ontario, searching for help with the school system their kids are actually in: report-card levels, the OSSLT, EQAO, grade 8 math.
The problem
The tutoring category in search belongs to a handful of global marketplaces with decade-old domains and millions of pages. A new brand cannot outrank them on "online tutoring," and the AI engines answering "best tutoring platform" name the same incumbents every time. Genie Teacher started from effectively zero: no rankings, no citations, barely any search demand for its name.
The strategy
The same dual-track SEO and AEO program we run for B2B clients, scaled to an early-stage brand. Every page has to rank in Google and be liftable by an AI engine. Two moves defined it.
Pick the wedge the giants ignore. Global marketplaces sell tutors. Genie Teacher sells certified teachers, to Canadian parents, for the Canadian school system. So the content answers the questions those parents actually type: what Ontario report-card levels mean, how the OSSLT works, what to do when EQAO results land, how to find a tutor who is a certified teacher. The incumbents cannot write this credibly at province level. A focused brand can.
Measure AI visibility from day one. We track 90 buyer prompts in Peec AI across the tutoring category, spanning competitor comparisons, certified-teacher tutoring, subject and exam prep, and local queries, so every content decision is checked against whether the engines start citing the brand rather than just whether a page ranks.
The results
The curve: a spring of doubling, then vertical
Monthly Google impressions since the spring (Google Search Console, root domain, to 25 September 2026):
| Month (2026) | Impressions vs the April baseline |
|---|---|
| April (from 13 April, when Search Console coverage starts) | baseline |
| May | 2.2× |
| June | 5.1× |
| July | 12.9× |
| August | 62× |
| September, first 25 days | 291× |
Through the spring each month roughly doubled the one before. In August the curve broke upward: the month closed at 62× the April baseline, and the first twenty-five days of September reached 291×. Measured per day, impressions rose 28× from May to August and 164× over 1 to 25 September; the latest day Search Console reports, 25 September, sits at 206× an average May day. Against the same month last year, July was up 5× and August 24×, and twenty-five days of September out-earned the whole of September 2025 by 84×.
Clicks are following, as the order of this kind of program says they should. The week ending 25 September brought 27× the clicks of an average May week, September's first twenty-five days brought 5.5× the clicks of all of August, and the average position across the site moved from 11.5 in August to 7.0 in September (Google Search Console). That position shift is the mechanism: the pages entered far more results over the summer, and are now climbing high enough in them to be clicked.
Leads and bookings: the curve reached the business
Search visibility only matters if parents act on it. Two actions are measured: a lead request, which is a parent asking to see a teacher's availability or to contact them from that teacher's profile (PostHog), and a lesson booking in the student app (Google Analytics). Over the first three full weeks of September (6 to 26 September), lead requests averaged 5.8× the weekly average of the seven tracked weeks before them (19 July to 5 September); week by week that was 7.8×, 5.8× and 3.8×, and they came from nearly four times as many different visitors a week. Lesson bookings over the same three weeks averaged 3.6× the average week from April to August. The counts stay private; the chart on this page shows the weekly shape, indexed to the pre-September average. Both series are short, so we will keep updating them here.
AI visibility: a 28.39% peak, settling near 10%
Peec AI's visibility tracking shows AI visibility rising from 5.26% in the week of 25 May 2026 to a peak of 28.39% in the week of 20 July 2026, a 5.4× increase across eight weeks with tracking paused in between, before settling between 8% and 11% from August into September (9.75% in the week of 21 September). Two things sit behind the settle. The peak came in the week 30 new prompts were added, a first-week reading rather than a settled one, and it shows what the ceiling looks like once the engines start citing the brand. Then the tracked prompt set grew by half after July, from 59 prompts to 90, so the answers Peec samples each week roughly doubled, from about 1,000 in the peak week to about 1,850 from August. Both lower any share-of-answers figure. The current level is where it has stabilised. Over the 30 days to 27 September, Genie Teacher stands 3rd of 7 tracked brands by share of voice and 5th of 7 by visibility. The strength is not the overall ranking, it is the rank when it is named: AI names Genie Teacher first or near-first (average mention rank 1.5 over those 30 days).
Broken down by engine over the same 30 days: Gemini gives Genie Teacher 12.3% visibility, and ChatGPT and Google AI Overviews 9.0% each, with the mention rank between 1.4 and 1.5 on all three, rounded. No single engine carries the number.
Share of voice: 2.26% to 13.0%, mention rank holding
Peec AI's share-of-voice tracking, run alongside the visibility numbers above, shows the same story from a different angle: 2.26% in the week of 25 May 2026 to 13.04% in the week of 21 September 2026, with an average mention rank between 1.0 and 1.6 every week since July. The May reading came from an earlier, smaller prompt set; the current 90 prompts were built between 29 May and 24 August, so the later figure is measured against a wider and harder set, not an easier one.
AI search: when the engines cite Genie Teacher, they cite it first
Across the 90 tracked prompts (Peec AI, 30-day window ending 27 September 2026), Genie Teacher's average position when cited is 1.5. Overall share of voice is 13.8%, which is honest for a young brand in a category the giants dominate. The shape underneath is the encouraging part: on competitor-comparison prompts Genie Teacher appears in 59% of answers at an average position of 1.1, and on the certified-teacher wedge it has climbed from 10.8% in the window ending 19 August to 16.4%. On the branded prompts, tracked from May to August, the engines named Genie Teacher in 99% of answers; those prompts are no longer in the set, so every prompt now tracked is one where the brand has to earn its place. The play is the same one that worked for our fintech clients: own the wedge prompts first, let the generic ones come later.
Branded demand is forming
People now search the name and click it. Over the 28 days to 25 September the query "genie teacher" runs a 77% click-through rate at position 1.0. Branded search is the trace that content, word of mouth, and AI mentions leave behind, and it is appearing on schedule.
The content pages doing the pulling
The Ontario-parent hub is what bent the curve. Over the 90 days to 25 September the report-card levels explainer has drawn the most impressions of any page in the hub, at position 5.9, followed by the OSSLT guide at 7.0, the periodic table quiz at 8.3, the teacher-certification guide at 6.9, and the grade 12 chemistry guide at 6.1; the periodic table quiz and the EQAO grade 6 practice test bring the most clicks. Most of the hub now sits on page one, where the click volume lives. None of them mention the brand in the query, which means this is non-branded, buyer-intent demand the content earned on its own.
What made this different
Early-stage SEO usually fails one of two ways: the brand chases head terms it cannot win, or it publishes thin content nobody asked for. The wedge solved the first. Writing to the exact questions Ontario parents ask solved the second. And measuring AI citations from day one means the next stage, becoming the engines' certified-teacher answer the way our fintech clients became the stablecoin payroll answer, has a scoreboard from the start.
The curve is young and we will update this page as it grows. That is the point of publishing it now: the interesting part of a hockey stick is watching the blade form. Updated 28 September 2026 with Search Console data to 25 September, Peec AI weekly readings and its 30-day window to 27 September, PostHog to 26 September, and Google Analytics bookings to 26 September.
Engagement: SEO and AEO program for an early-stage education brand, ongoing.



