What the numbers mean, and how the data is produced — so you can trust the rankings.
Every day, we ask ChatGPT, Claude, Perplexity and Gemini to recommend the best Shopify apps in each tracked category, across several phrasings per category. We call the official APIs with our own keys — no scraping, no residential proxies. Each answer's app mentions and their order are recorded verbatim.
Assistants name the same app many ways. We collapse variants to one canonical name at analysis time (e.g. “Klaviyo: Email Marketing & SMS” → “Klaviyo”), normalizing punctuation, dashes, and trailing “by Vendor” tags. Because raw mentions are kept verbatim, improving the rules just re-runs the analysis — no re-collecting.
“Agreement” is how much the four models' top picks overlap in a category (mean pairwise overlap of their top three). Lower means they disagree more — an app can be #1 in one model and invisible in another.
An app is “new” when it enters the cross-engine consensus in the recent window and holds for at least two days, so a one-off sampling blip doesn't count.
Shopify's in-admin Sidekick has no public API, so we can't read what it recommends. The readiness score estimates how well-positioned an app is against the factors Shopify states drive it — the Built for Shopify badge, rating, and review volume — plus its AI cross-engine presence as a discoverability proxy. It is readiness versus stated factors, not a prediction of Sidekick's output.
NextSASI is an independent measurement project. It is not affiliated with or endorsed by OpenAI (ChatGPT), Anthropic (Claude), Perplexity, Google (Gemini), or Shopify.
A short monthly email: the new apps ChatGPT, Claude, Perplexity and Gemini start recommending, the biggest movers, and where your app stands.
Free · no spam · unsubscribe anytime