How often should we run it?

Match it to how often your catalogue actually changes. A marketplace onboarding sellers weekly wants weekly. A specialist retailer with two seasonal ranges wants monthly or even quarterly. Running it more often than your catalogue changes just produces identical numbers.

What triggers an alert?

A drop in a category's readiness beyond a threshold you set. The alert names the shelf, the attribute responsible and the products that introduced the gap, so it is actionable without opening a dashboard.

Does this need an integration?

No. It reads the same export the audit read. If you can schedule that export, monitoring works, and nothing on your site or in your stack changes.

Can we monitor categories we have not audited?

They need an audit first to set the baseline. After that, monitoring keeps measuring against it.

Decision Monitor

Catch a category slipping before a buyer does.

Catalogues decay. New products arrive unrecorded, supplier feeds change shape, and a shelf that was ready last quarter quietly stops being ready. Monitor re-measures as your catalogue changes and tells you when one slips, naming the cause.

Category readiness, week by week

One quarter

Monitor says so in week 2

Noticed at the quarterly review

Laptops fell from 71 to 62 while 340 products arrived without processor data. No single week looked like a problem, which is exactly why nobody caught it.

Where you are

The repair holds until the next few hundred SKUs arrive.

A supplier changed their exportand the fields quietly stopped mapping
A range was onboarded in a hurrybecause the deadline was the launch, not the data
The person who knew the taxonomy leftand the convention left with them

-9

Readiness, one quarter

Nobody decides to let a category decay. It happens because a supplier changes their export format, or a buyer onboards a new range in a hurry, or the person who understood the taxonomy left. Each of those adds products that skip the fields your shoppers rely on.

By the time anyone notices, the shelf has been quietly worse for a quarter and you are back where you started, except now the enrichment work has already been done once and asking for it again is a harder conversation.

Laptops fell from 71 to 62 because 340 new SKUs arrived without processor data. That is the sentence you want, in the week it happens, not at the quarterly review.

Why it gets expensive

The same defect costs four different things depending on when you find it.

Nothing about the gap changes between week one and month six. What changes is how many products are wearing it, and who has to approve the fix.

Week 126One export to remap. Nobody to ask.
Month 1105A backlog item, done inside a sprint.
Month 3340A project, and a conversation about it.
Month 6680A second budget request for work already paid for once.

Products carrying the gap, at the arrival rate in the example above. The defect does not change; only the number of products wearing it, and who has to approve the fix.

What it looks like

The alert, not a dashboard.

It names the shelf, the attribute and the cause, so it is actionable without opening anything.

Shelf alertLaptops

Readiness fell 71 to 62

340 new SKUs arrived without processor data on 14 March.

Readiness, 12 weeks

What happens

How does monitoring work?

01Once
Set the cadence
Weekly, monthly or on a feed. It reads the same export the audit read, so there is nothing new to build and no integration to maintain.
02Ongoing
We re-measure
The same method on every run, so the numbers are comparable over time rather than a series of unrelated snapshots.
03When it matters
You get told, with the reason
Not a dashboard you have to remember to open. An alert naming the shelf, the attribute and what caused the drop.

What you get

What monitoring gives you

An alert that names the cause

Which category slipped, which attribute did it, and how many new products arrived without it. Enough to act on without opening anything.

A trend per category

Readiness over time per shelf, so you can see the slow drift as well as the sudden drop, and argue for the fix before it is urgent.

A cadence you choose

Some catalogues change daily and some change twice a year. You set how often, and you can change it.

Nothing new to maintain

It reads the export you already produce. No tag, no integration, no library to keep up to date as your platform changes.

Before you commit

What monitoring watches, and what it does not

What this doesDetect the drift while it is still one supplier feed, rather than a quarter of arrivals

Attribute the drop to a named cause, so the fix is assigned rather than investigated

Hold a threshold you set, so a shelf that moves a point does not page anybody

Keep running whether or not you act on it, so a shelf you chose to ignore is still measured

What it does not doFix the drop. It tells you what happened and where; the repair belongs where your data lives

Watch your traffic, your rankings or your competitors. It reads product data only

Alert on a category that was never audited, since there is no baseline to compare against

Questions

The ones that decide it.

How often should we run it?: Match it to how often your catalogue actually changes. A marketplace onboarding sellers weekly wants weekly. A specialist retailer with two seasonal ranges wants monthly or even quarterly. Running it more often than your catalogue changes just produces identical numbers.
What triggers an alert?: A drop in a category's readiness beyond a threshold you set. The alert names the shelf, the attribute responsible and the products that introduced the gap, so it is actionable without opening a dashboard.
Does this need an integration?: No. It reads the same export the audit read. If you can schedule that export, monitoring works, and nothing on your site or in your stack changes.
Can we monitor categories we have not audited?: They need an audit first to set the baseline. After that, monitoring keeps measuring against it.

Fix it once, then keep it fixed.

Monitoring starts after an audit, so the first step is the same one: send us one category.

Talk to us about monitoring

Catch a category slipping before a buyer does.

Set the cadence

We re-measure

You get told, with the reason

An alert that names the cause

A trend per category

A cadence you choose

Nothing new to maintain

Fix it once, then keep it fixed.

Catch a category slipping before a buyer does.

Set the cadence

We re-measure

You get told, with the reason

An alert that names the cause

A trend per category

A cadence you choose

Nothing new to maintain

Fix it once, then keep it fixed.