Measurement
See where shoppers stop. Follow the order in your own analytics.
Rekotto records what happens inside the quiz: who started, who finished, which question people reached and which product they clicked, so you can see where undecided shoppers drop off and which products they go to. The Shopify app also shows an estimate of purchases that followed a quiz (see below). Orders themselves happen in your cart, so your own analytics stay the record of them. Every product link on the result screen is tagged so you can find those visits there.
Live on the Shopify App Store for shops in the UK and EU, in English, German, French and Italian.
What is recorded
Five moments, recorded as they happen
Five events, one session ID, no names.
Each event carries a session ID, which is one shopper in one visit. It carries no name and no email address.
Quiz started
quiz_startRecorded when the shopper taps the start button, not when the page opens. Someone who looks and leaves is not counted as a start, so they do not drag your completion rate down.
Question reached
question_viewRecorded once for each question a shopper gets to. Going back and forward does not count a question twice. This is how you see where people stop.
Quiz finished
quiz_completeRecorded when the shopper submits the last answer.
Result seen
result_viewRecorded when the shopper taps through to see the results. It is the starting point for the click rate.
Product clicked
product_clickRecorded when the shopper taps a product on the result screen, with the product and its position in the results.
Quiz started
quiz_startRecorded when the shopper taps the start button, not when the page opens. Someone who looks and leaves is not counted as a start, so they do not drag your completion rate down.
Question reached
question_viewRecorded once for each question a shopper gets to. Going back and forward does not count a question twice. This is how you see where people stop.
Quiz finished
quiz_completeRecorded when the shopper submits the last answer.
Result seen
result_viewRecorded when the shopper taps through to see the results. It is the starting point for the click rate.
Product clicked
product_clickRecorded when the shopper taps a product on the result screen, with the product and its position in the results.
What the report shows
A few counts, each with its sum written out
Seven figures. Each rate shows its count.
The report is built from the events above and from nothing else about the shopper. Every percentage is shown with the count it was worked out from, so a rate based on a handful of sessions looks like one.
| Figure | How it is worked out | Counted as |
|---|---|---|
| Starts | Number of quiz starts in the period | Sessions |
| Completion rate | Quizzes finished ÷ quizzes started | Sessions |
| Product clicks | Every tap on a product | Taps. A shopper who opens three products counts three. |
| Click rate | Sessions with a product click ÷ sessions that saw results | Sessions. That same shopper counts once. |
| Questions reached | How many sessions got to each question | Sessions. The drop between two questions shows where people stop. |
| Where sessions ended | Each started session placed in one outcome: clicked a product, saw results without a click, finished without opening results, or left part-way | Sessions. The four add up to the starts. |
| Daily trend | Starts and finishes for each day | Sessions |
This page describes what the report counts, not how a screen looks. The report itself is in the Rekotto dashboard, in English.
Orders
Follow an order in your own analytics
Tagged links, read in GA4 or any tool.
Every product link on the result screen carries tags. Your analytics reads them, so a visit that came through the quiz and any order that follows it sit in the reports you already use. This works if your analytics already records orders.
Find the quiz in GA4
- Open the traffic report
Go to Reports, then Acquisition, then Traffic acquisition.
- Show source and medium
Change the table’s dimension to Session source / medium.
- Find the quiz row
By default it reads
diagnosis / quiz. That row is every visit that arrived through a Rekotto result link. Read its orders against the other rows, over the same dates.
In another analytics tool, look for visits whose UTM source is diagnosis. The idea is the same.
The tags on each link
| Tag | Default value | What it tells you | Use it to |
|---|---|---|---|
utm_source | diagnosis | Every visit that came from the quiz. Filter on this first. | Filter |
utm_medium | quiz | Confirms the visit came from a quiz result. Use it with the source. | Filter |
utm_campaign | Your store’s identifier | A label for your store’s quiz. It tells stores or quizzes apart if you run more than one. | Break down |
utm_content | Quiz ID, then the position | Which product in the results was chosen. The final digit is its position, so rank1 is the top recommendation. | Break down |
Visits that lose their source, and shoppers who change device between the quiz and the order, cannot be matched. The figures will read lower than what happened.
How to read it
Read the counts before the rates
Five habits that keep the figures honest.
- Look for the question, not the average
- A steep drop between two neighbouring questions points at the second one: its wording, how many answers it offers, how much thought it asks for. A low click rate beside a healthy completion rate points at the result instead: the products shown and the reasons given.
- Check the count behind the percentage
- With few starts, one busy afternoon moves every figure. The count under each rate tells you how much weight it can bear.
- Compare like with like
- Use the same quiz and the same number of days before and after a change.
- Sessions are visits, not people
- One shopper who takes the quiz on a phone and again on a laptop is two sessions.
- Days run on Tokyo time today
- The report currently cuts each day on Tokyo time, so a day may start or end at an odd hour in your own. Time zones set per shop are planned.
What Rekotto will not say
Three things the report will never claim
No cause, no uplift, no body claims.
- That the quiz caused a sale
Shoppers choose to take the quiz, and people who choose it are often already close to buying. Rekotto does not split your visitors into a group that sees the quiz and a group that does not, so it cannot tell the two apart.
- An uplift, a revenue figure or a benchmark
No percentage lift, no average from other shops, no “shops like yours”. The report holds the counts from your own quiz.
- Anything about a shopper’s skin, hair or body
The quiz classifies and recommends. It does not promise an outcome. How the wording is checked is on the claims check page.
Data and consent
What this page does and does not say about data
No cookies. Consent before any ID.
- Cookies
- The quiz itself sets no cookies.
- Session ID on the cart
- When a shopper reaches the results, the session ID is added to the product links and written onto their Shopify cart, which Shopify carries onto the order. For stores with a UK or EEA address this happens only when the shopper has allowed analytics in your cookie settings; otherwise no ID is written. Your own shopper privacy notice needs to cover it.
- Consent
- Rekotto also installs a Shopify Web Pixel. It respects the shopper’s consent choice and sends nothing until analytics and marketing are both allowed.
- Ownership
- The answers shoppers give, and the numbers the quiz produces, belong to you.
- Where it is held
- In a database in Tokyo, Japan, with the app running on servers in Tokyo. The UK and EU privacy notice and the Data Processing Addendum say who handles the data and for how long.Read the legal documents →
FAQ
Questions about measurement
What does Rekotto record?
Five moments: the shopper starts the quiz, reaches each question, finishes, opens the results and clicks a product. Each event carries a session ID, which is one shopper in one visit, and no name or email address.
How do I see which orders came from the quiz?
The Shopify app shows an estimate of purchases that followed a quiz, built from two sources. One is your Shopify order records, matched to a quiz session through a session ID the quiz writes onto the cart; for stores with a UK or EEA address that ID is written only when the shopper has allowed analytics, and the source keeps only the amount, the currency and the day. The other is a measurement pixel, which counts only shoppers who allowed analytics and marketing. Each purchase is counted once. It will read lower than Shopify’s own sales report, because orders with no quiz session cannot be matched. Your own analytics are the record of orders. Every product link on the result screen carries UTM tags (by default source diagnosis and medium quiz), so you can filter on them in GA4 or another analytics tool. Visits that lose their source, and shoppers who change device, cannot be matched.
If a shopper opens three products, is that three clicks?
The product-click total counts every tap, so yes. The click rate counts sessions, so that shopper counts once.
Can the report show that the quiz raised my sales?
No. Shoppers choose to take the quiz, and Rekotto does not split your visitors into a group that sees it and a group that does not. The report shows what happened in the quiz, not what caused an order.
Does the quiz set cookies?
The quiz itself sets no cookies. It does write a session ID onto the shopper’s Shopify cart so an order can be matched to a quiz. For stores with a UK or EEA address it does so only when the shopper has allowed analytics in your cookie settings. The Shopify Web Pixel is separate and sends nothing until analytics and marketing are both allowed. The data is held in Tokyo, Japan; the UK and EU privacy notice says who handles it and for how long.
Who owns the data?
You do. The answers shoppers give, and the numbers the quiz produces, belong to you.
Give every visitor an answer, on your own store.
Install it from the Shopify App Store, start from a ready-made question set for your shelf and preview it before it goes live. Billing runs through Shopify, and you can stop at the end of any month.