GA4 Ultimate Stack

Catch a key event drop before month-end, across every brand

Scan the exact key events you choose across every brand, judged against each event's own history, so real drops are flagged and normal quiet days are not.

ODDVXPractical guide

Why one threshold doesn't fit every event

A day with no purchase events on a busy shop is an emergency. A day with no quote_request on a small brand is a Tuesday. A single rule like "alert when it hits zero" either floods you with false alarms on quiet events or misses real breaks on busy ones.

Key Event Anomaly Alert judges each event against its own history on each brand. It is read-only.

A zero on a busy event is a break; zeros on a quiet event are normal. One rule can't tell them apart.

Two daily bar charts: purchase with a single real zero-day break, and quote_request with normal zero days.

Pick brands and the exact key events

  1. Select the brands. Free scans one brand per run; Premium can scan all of them.
  2. Choose the exact key events to watch, by name. No naming convention is assumed.
  3. Set the days of history to learn from (56 by default) and run the scan.

By default each brand's roll-up property is scanned. You can include its source property as well.

Brands, the exact key events and how much history to learn from.

A scope form with four brands, three key events and 56 days of history.

Choose a sensitivity

Three presets control how readily the scan alerts. Standard is the default. Conservative alerts less often and suits noisy or seasonal events; Aggressive alerts sooner. Advanced settings let you adjust each value.

The three presets side by side. Standard is the default.

A grid comparing conservative, standard and aggressive presets on drop percentage, window length, streak ceiling and minimum gap.

Critical: a dry streak beyond normal

For each event, the scan learns how long it normally goes without being recorded. Critical means the current run of zero days is longer than that event's own historical ceiling. An event that occasionally skips two days won't alert on a two-day gap, but will on six.

Two-day gaps are normal for this event. Six days isn't.

A daily bar chart of purchase with a normal two-day gap and a six-day zero streak flagged critical.

Warning: a window far below its own baseline

Warning means the event is still being recorded, but the total over the trailing window has dropped sharply against the event's own baseline: 80% or more on the standard preset. That's the pattern of a partial break, such as one form or one browser no longer tracking.

Still recorded, but 82% below its own baseline: the signature of a partial break.

A daily bar chart of generate_lead dropping well below a dashed baseline over the last seven days, flagged warning.

Read the findings

Findings are ranked by severity: Critical, then Warning, then OK. Click a row in the app for the full reasoning: the streak, the ceiling, the window sum and the baseline. On Premium, Export CSV keeps the whole audit trail. Then confirm the drop is real (a test journey, the site's own records) before you treat it as a tracking break; sometimes the business really did go quiet.

Critical first, then Warning. Each row explains its own reasoning in the app.

A findings table with one critical, one warning and two OK rows across brands.

Which property is scanned

Each brand's roll-up property is scanned by default, because that is usually what clients report from. Add the source property when you need to see whether a drop is specific to one site.

Roll-up by default; source when you need to narrow it down.

A brand linked to its roll-up property, scanned by default, and its source property, optional.

Keep going

ODDVX for Windows

Every property, one screen.

Start free on one GA4 brand per run. Premium adds every brand in one run, batch writes, exports and scheduled checks.

Download FreeSee the plans