Year In Review
What happened to your store over the last twelve months, and what to fix next.
What it does for you
A full year of your own trading, already read, ending in the two or three things worth fixing next.
Nobody running a store has the time to look back twelve months properly. This page did it once, from your own data, and wrote down what it found.
It is a story you scroll, not a dashboard you query.
The heading reads AI Insights — Your Year, Explained, and the grey line under it begins Based on and gives the exact window.
Read that line before anything else. It is the only place the page tells you which twelve months you are holding.
How to read it
Two different kinds of thing sit on this page, and only one of them is judged.
The verdict card at the top is graded — each number was measured against a bar and passed or failed. Everything below it is written — prose about what those numbers did across the year.
Trust the grades to tell you where to look. Trust the See the data button to tell you whether it's true.
Read it in this order
- The verdict card at the top. A row of graded numbers and one sentence. If you read nothing else, read the sentence.
- The red icons only. Scroll the whole page once, looking for the sections badged needs work. Skip everything green on the first pass.
- The closing card, "Where to start". It's a numbered list. Number one is the whole point of the page.
- Then go back up and read the two or three sections you flagged, properly.
Read it top to bottom and you finish with twelve things to do. Read it in that order and you finish with one.
The verdict card
The only thing you need from this card is which vital failed.
Each vital shows the plain name with the industry term in brackets: Basket size (AOV), Ad efficiency (MER), Monthly sales (avg/mo), Come back (Repeat rate).
A tick beside the value means it cleared the bar for that measure. A red warning icon means it didn't.
Read the icons, not the values. The values are already graded, so what's left to learn is which one broke.
A failed vital is a claim, not a diagnosis. It tells you which section further down is worth your attention.
The sections
This is where the year gets explained, one topic at a time, most of them ending in something you can do.
Below "Let's walk through your year together" the page becomes a timeline with a blue rail that tracks your scroll.
The sections always run in the same order, whatever the review found: customers, products, geography, timing, discounts and shipping, then channels and stores when your setup has them.
A section that isn't there is a section the review had nothing to say about, not a zero.
Each one carries a status badge. Four words are possible: needs work, worth a look, solid, healthy. The first two are the only ones that should slow you down.
Inside a section, the points are tagged with a word: Grow means push on it, Steady means leave it alone, Wait means it's too early to judge, Good means it's already working.
The word carries the meaning. The coloured icon beside it is there so you can find it while scrolling.
Some sections end in a Do this box with a short impact phrase next to it. A section without that box had no single clear action, which is itself information.
See the data
This is the button that turns a sentence about your year into a number you can check.
Every section that can be checked ends in one. It opens the screen holding the numbers behind that section:
| Section | Opens |
|---|---|
| Customers, channels | Shopify → Overview |
| Products | Shopify → Products |
| Geography | Shopify → Region |
| Timing | Shopify → Breakdowns |
| Discounts and shipping | Shopify → More |
| Stores | Agency View |
Use it before you act on a section, not after. The review is one reading of a year; the page behind the button is the year itself.
Where to start
The closing card is the only ranked thing on the page. The sections above are ordered by topic, not by how much they cost you.
It's a numbered list of priorities, each with a short impact phrase. When it disagrees with the order you read the sections in, it wins.
Under it you may see a Next steps card. That one is not about your year at all: it lists the connections you haven't made yet, and it disappears once everything is connected.
Claude and GPT
Same year, same data, two writers. When buttons labelled Claude and GPT appear beside the heading, each model wrote its own review, and Claude is what you land on.
Switching between them fetches nothing and changes no number. It changes the wording and the emphasis.
If the two rank your problems differently, that is a difference in writing — not a second opinion on the data.
Sharing it
Share saves the whole page as a PDF, or sends it by WhatsApp or email. It is the one screen in the app worth handing to a partner, an accountant or a lender exactly as it is.
The share bar, the model buttons and the "See the data" buttons are left out of the file.
What it won't tell you
Anything that happened after the date in the grey line. The review was written once, from that window, and it does not move on behind you.
It won't follow your filters. The date picker, the comparison period and the country filter are all hidden here.
The window is fixed when the review is generated. Changing the date range on another screen does nothing to this page.
A missing section is silence, not zero. Sections drop out when there was nothing worth saying. Don't read an absent products section as "no product sales".
It won't confirm its own numbers. The footer reads "AI-generated insights — verify before acting", and it means it — this is written prose about your data, not the data.
Open the "See the data" page and check the figure before you change anything that costs money.
It won't be there the first time. A first build reads twelve months and takes minutes, not seconds. The page tells you to check back.
Do this first
Read the verdict card, find the one vital with a red icon, scroll to the section it points at and press See the data.
Confirm the number there, then act on that section's Do this box and leave the rest of the page until next month.
See yoursYour last twelve months, written out, with the two or three things worth fixing.
Open Year In Review →