Promo Drawer
Analytics
What the app measures, what the numbers mean, and what it deliberately does not report.
Every plan reports what the drawer did, including the free one. What paid plans add is how far back you can look, not whether you can look at all.
What is measured
| Event | When it fires |
|---|---|
| Drawer view | A drawer is selected and rendered on a page. Counted once per session per drawer |
| Drawer open | A shopper opens it |
| Card view | At least half a card is visible for at least a second |
| Card click | A card’s button is activated |
| Code copy | A shopper copies a discount code |
| Video play | A shopper taps play on a video card |
| Signup submit | The newsletter form is submitted |
| Follow click | A social link is activated |
Individual events are never stored. They are combined before they are written down, into daily counts per drawer, card, page type and device. There is no record of one shopper’s visit, because there is no need for one to produce any number on these pages.
What you see
The app’s Overview shows the last seven days against the seven before it: drawer opens, open rate, card clicks, click rate, and your best-performing card.
The Analytics page adds a date range, a daily trend line, a per-card table with views, clicks and click-through rate, a per-drawer table once you have more than one, and breakdowns by page type and device. Cards with a secondary action — codes, videos, signups — show that count too, because a coupon card’s real success is a copy rather than a click.
Orders that followed
Two kinds of attribution, with different windows, because they are different strengths of signal:
- Opened and purchased — orders placed in the same session after opening the drawer, capped at 24 hours.
- Clicked and purchased — orders placed within 7 days of clicking a card.
Both are labeled that way in the app rather than as a single “conversions” number, because they do not mean the same thing and the distinction is the whole point.
These are measurements, not claims of cause. An order counted against a drawer is an order that happened after a shopper interacted with it. Some of those would have happened anyway. Any tool that tells you otherwise is choosing a model and presenting it as a fact — this one reports what it saw and labels the window it saw it in.
Where a sample is too small to mean anything, the app says so rather than printing a confident percentage from twelve impressions.
Why your numbers may be lower than you expect
Analytics respects shopper consent, and fails closed. Before anything is counted, the app asks Shopify’s Customer Privacy API whether analytics processing is allowed for that visitor. If the answer is no — or if the question cannot be asked at all — nothing is collected.
In regions where you have configured consent to be required, that means visitors who have not accepted are not counted. Your drawer still works for them; it is only the measurement that stops. Expect the totals here to sit below your storefront’s own session counts, and further below in the EU and UK than in the US.
This is the deliberate trade. Counting people who declined would produce a better-looking chart and a worse product.
Why a rate sometimes shows a dash
An open rate needs opens and views to describe the same population, and that relies on browser storage the app uses to avoid counting one shopper twice. In a private window, some in-app browsers, or anywhere storage is blocked, that deduplication is not possible.
When that happens the app shows a dash and the reason, rather than a percentage. A number nobody measured is worse than an honest gap, and clamping the rate to 100% would state something specific that never happened.
What is deliberately not reported
- Estimated revenue. Attributing money to a drawer requires a model, and merchants reasonably read whichever number is printed as truth.
- Individual visitor journeys, heatmaps and session recordings. Those need a persistent identifier and a much larger privacy surface than this app is willing to hold.
- Geographic breakdown. It needs IP geolocation, which the app does not do.
- Comparisons against other stores. “You are in the top 10%” is a claim about other people’s data.
What the app holds about a shopper, and for how long, is in the Privacy Policy.