Measurement
Know whether AI recommends you.
rankyn asks the buying questions your shoppers actually ask, on the AI surfaces they use. You see who gets recommended, who wins instead, and why.
“Best magnesium for better sleep?”
ChatGPTRecommended
- 1
Halvor & CoMagnesium Bisglycinate Powder - 2
Lumae LabsMagnesium Glycinate Capsules - 3
OstrelHigh Absorption Magnesium
Not recommended
How a measurement runs
Every measurement starts from a buying intent and ends with a read on whether the AI recommends you. Five steps, repeated on every surface.
01
Intent
An intent is a buying situation your customers are in, such as finding a gentle magnesium for sleep. You choose the intents that matter for your catalogue.
02
Prompt
A prompt is one concrete question a shopper asks inside an intent, in their words and in your market. Each intent holds several prompts.
03
AI surfaces
Each prompt runs on the AI surfaces you choose, such as ChatGPT and Google AI Overview. One answer from one surface is one measurement.
04
Answer
rankyn captures every answer in full: which products the AI names, in which order, and what it says about them.
05
Evaluation
rankyn reads each answer against your products and your competitors, and turns it into a recommendation rate you can act on.
Your prompts, not a generic list
rankyn starts you from suggested intents and prompts, but the questions are yours to shape. Nothing is measured until you confirm it.
Suggested by AI at onboarding
rankyn proposes buying intents and prompts from your catalogue and category, so you start with questions that fit your products.
You confirm or edit
Nothing is activated without you. You confirm the suggestions you want, edit them, or write your own before the first measurement runs.
Shopper language, your market
Prompts are written the way your customers ask, in your language and your market, so the measurement matches real buying behaviour.
Your catalogRead from your store
yourstore.myshopify.com128 productsMagnesium Glycinate PowderSupplements · Sleep127 more productsSuggested intents and promptsProposed by rankyn
Magnesium for sleep3 prompts“Best magnesium for better sleep?”Suggested“Which magnesium helps you fall asleep faster?”Suggested“Magnesium glycinate or citrate for sleep?”SuggestedMagnesium for muscle cramps3 promptsYou confirm or editNothing is measured before this
“Best magnesium for better sleep?”Confirmed“Which magnesium helps you fall asleep faster?”Confirmed“Is magnesium glycinate better than citrate for sleep?”EditedWrite your own prompt
Measured every day
Every plan is measured daily, one sample a day. Each prompt runs on each surface on every measuring day.
- Measuring days a week
- 7One sample a day on every measuring day.
- AI surfaces to choose from
- 5Growth measures 5 of them at once; every prompt runs on each one.
- Measurements a week
- 5,250Growth: 150 prompts × 5 surfaces × 7 days.
Your recommendation rate
The recommendation rate is the share of answers on one surface where the AI recommends your product. Every surface has its own rate, and every rate comes with who was recommended instead.
Defined per surface
rankyn counts how often the AI puts your product forward, out of all the answers captured on that surface in a week.
Who is recommended instead
Every rate names the competitors the AI chose over you, so you see who wins when you do not.
The number that matters
The rate is your measure of whether you win the recommendation. A fix is proven when a re‑test moves it, not when a report sounds better.
Google AI OverviewAnswers this week33%
4 of 12 answers recommend your productRecommended instead most often: Lumae Labs
Every surface has its own rate
- ChatGPT17%Halvor & Co
- AI Overview33%Lumae Labs
- AI Mode0%Halvor & Co
Perplexity50%Ostrel
Your position in the answer
Position is where your product stands among the products an AI answer recommends, per surface. It is the second figure beside the recommendation rate, which stays the main one.
Only where you are recommended
An answer that does not recommend your product has no position for it. That answer already counts against your rate.
Per surface
Each AI surface orders its answer in its own way, so every surface shows its own position.
Beside the rate, not instead of it
Being recommended comes first. Position tells you how prominently, once you are.
“Best magnesium for better sleep?”Recommended in this answer
- 1Lumae LabsMagnesium Glycinate Capsules
- 2Your productMagnesium Glycinate Powder
- 3Halvor & CoMagnesium Bisglycinate Powder
Position per surface
- #3ChatGPT
- #2AI Overview
- –AI ModeNot recommended
- #1Perplexity
Share it as a PDF
Export a report with your recommendation rate, position, findings and re‑tests as a PDF and pass it on, to your team or, if you run an agency, to your client.
The whole picture in one file
Recommendation rate and position per surface, the open findings with their fixes, and the latest re‑test results.
Made to pass on
A PDF opens anywhere, so the people who decide see the same numbers you do, without an account.
PDF export is included in Growth and Scale.
Magnesium Glycinate Powder
- Recommendation rate
- 33%Google AI Overview
- Position
- #2among recommended products
- Open findings
- 31 high priority
- Re‑tests
- 2latest: rate up
See where you stand.
Start measuring this week with a free trial, or see where your store stands with a free audit first.
Questions about measurement
Cadence, prompts, surfaces and the recommendation rate in short. Can’t find your answer? Write to hello@rankyn.ai.
How often does rankyn measure?
rankyn measures every prompt on every surface daily, one sample per measuring day, so your recommendation rate is never built on a single answer.
What is a recommendation rate?
A recommendation rate is how often the AI recommends your product on one surface, out of all the answers rankyn captured there that week. It is your measure of whether you win the recommendation, and each surface has its own rate.
Who decides what rankyn measures?
You do, with a head start from rankyn. rankyn suggests buying intents and prompts from your catalogue and category, then you confirm, edit, or write your own. Nothing is measured until you activate it.
Which AI surfaces does rankyn measure?
rankyn measures the surfaces your shoppers actually ask, chosen from a pool of 5. Your plan sets how many you measure at once, and every surface is measured the way it really works, as a real interface or an API.
What counts as one measurement?
For rankyn, one measurement is one answer from one AI surface to one prompt. Your weekly total is prompts times surfaces times measuring days, so the plan you choose sets how wide and how often rankyn measures your catalogue.
How much will my plan measure?
rankyn sizes each plan by prompts times surfaces times measuring days, so more prompts, surfaces, and days means more measurements. The recommended plan runs 5,250 measurements a week across 5 surfaces.