Ten weeks ago, AI assistants had never heard of Cartwheel and Google half-confused you with a mental-health company. This report answers four plain questions: are AI tools starting to recommend you, are more people seeing you on Google, what do shoppers do when they see you — and is any of it turning into business where you actually sell: Walgreens and Amazon.
| The question | How we measure it | What happened | Answer |
|---|---|---|---|
| Are AI systems reading & citing your site? | Times AI cited gocartwheel.com (Bing) | ~2/week in May → 21 last week, a record | ▲ yes, accelerating |
| Do AI assistants recommend Cartwheel? | Mentions in 200 lice-shopping questions | 0 in 200 (May) → ~8 in 200 (Jul) | ▬ started, not grown since Jun |
| Do you show up for questions, not just your name? | Non-brand Google appearances / day | 9 → 19 — doubled since the blog articles | ▲ yes, new |
| Are more people seeing the brand on Google? | Brand searches shown · your position | Appearances up · position 5.5 → 3.6 (mid-page → top 4) | ▲ yes, best ever |
| Where does the click go? | Clicks to your site vs the results around you | ~10 clicks/wk now land on Walgreens & Amazon listings | → to retail — where you sell |
| Do AI assistants send visitors directly? | Site visits arriving from ChatGPT etc. | 1 visit in 60 days | ▬ no — and that's expected |
| Is it producing retail sales? | Walgreens / Amazon sell-through | We can't see this yet — we need your retail data | ? the missing piece |
Starting, yes — and the leading indicator is surging. Before an AI assistant recommends a product, its systems have to read and trust the material behind it. That reading is measurable, and it's at record levels.
Read it bottom-up. AI assistants mention Cartwheel in about 4 of 100 lice conversations — up from zero in May, but flat since June (top chart's surge hasn't reached the answers yet). Meanwhile the citations — AI systems quoting your site as a source — jumped right after the five expert articles went live in late June, and are still climbing. Citations come first; recommendations follow. The Reddit and YouTube work now underway feeds exactly this pipeline.
| Week of | AI citations | Brand appearances/day | Clicks to your site/day |
|---|
They Google you. For 8 days your brand was visible inside ChatGPT (4,738 views). Almost nobody clicked through — but Google searches showing your brand ran at more than double the normal rate, then fell back within days of the ad ending.
This is the pattern the whole strategy is built on: people don't click out of AI conversations — they see a brand, then go search for it. So the payoff of AI visibility shows up in your Google numbers, not your visitor counts. (Honest footnote: the CafeMom article ran the same week, so we say "consistent with," not "proved." But the shape — rise with exposure, fall when it stops — is exactly the fingerprint.)
| Period | Days | Brand appearances/day | Clicks to your site/day |
|---|---|---|---|
| Normal weeks (late May–mid Jun) | 25 | 10.9 | 1.72 |
| CafeMom + ECRM press days | 4 | 19.2 | 2.25 |
| ChatGPT ad running | 8 | 25.6 | 1.12 |
| After the ad (clean) | 10 | 12.9 | 0.90 |
| Newsletter week | 6 | 15.5 | 2.17 |
More people see you; fewer click your site — and for your business, that's mostly good news. Your Google position improved from mid-page to top-4, yet clicks to gocartwheel.com fell. Those searchers didn't vanish: the results above and around you are Walgreens and Amazon — your own shelves.
We estimate about 10 brand-searches a week now end at a retail listing instead of your site. Since you sell through 6,000 Walgreens doors and Amazon — and you've told us DTC isn't the goal — that's intent arriving where you convert. Your site does its own routing too: we now track outbound clicks to retailers (Walgreens, Amazon, Instacart), and the "find it in stock near you tonight" article is built for exactly this. One improvement worth making: retail listings show star ratings in Google and your own result doesn't yet — adding your real reviews to the product page lets you decide how that shelf-moment looks, rather than leaving it to default.
| Period | Shown/day | Clicks to site/day | Google position |
|---|---|---|---|
| Before the work (May 3–18) | 11.9 | 2.19 | 5.5 |
| After first release (May 20–Jun 13) | 10.9 | 1.72 | 6.1 |
| After press weeks (Jun 27–Jul 6) | 12.9 | 0.90 | 3.6 |
Steadily, from brand recognition — not yet from AI. Most direct visitors land on your homepage: people who know the name from a shelf, a friend, or the press. The line to watch is the other one — people arriving directly on a product page, which is what it looks like when an AI told someone exactly what to buy.
The June bump in the deep-page line is the ChatGPT ad's clicks arriving unlabeled; it fell back when the ad stopped. The July homepage surge is your newsletter. The signal we're waiting for is the teal line rising in an ordinary week — that's when AI recommendations start walking in the door. It hasn't happened yet, which matches AI mentions still sitting at 4%.
| Week of | To homepage | To a deep page |
|---|
We can say: AI systems went from never citing you to citing you at record rates. AI assistants went from never mentioning you to mentioning you in ~4% of relevant conversations. You appear for real customer questions, not just your name. Your Google position is the best it's ever been. And brand-searchers are landing on the shelves where you sell.
We won't say yet: that AI mentions grew this month (they've been flat since June — the new citations haven't converted into recommendations yet); or that any of this produced retail sales — that claim waits for your Walgreens and Amazon data, because you deserve the real answer, not the convenient one.