AnalyticScan
KO

Sample result

Commerce GA4 diagnosis

All thirty verdicts and the evidence behind them, without signing in.

Run 2026-09 · over the last 28 days

Three things to look at first

  1. Critical failureCV-02Conversions not collecting
  2. Weakest categoryConversions (key events) — 38%
  3. Quick winPR-02Parameter missing rate — Worth 5 points

Data health score

73/ 100

Grade

C

Caution — some key metrics can't be trusted

1 critical failure(s) (CV-02)

Score by category

  • Event collection92
  • Conversions (key events)38
  • Parameters & custom definitions62
  • Property settings100
  • Tagging & data streams80
  • Integrations & governance100

All thirty checks

  • CV-02Conversions not collecting

    A conversion that is registered but never collected reads as zero. You end up switching off a campaign that was selling perfectly well.

    Key events with zero count: purchase, app_store_subscription_convert, event_1, event_2, event_3 — configured but not collecting

    I don't understand what this means

    In plain terms — we check whether an action you've flagged as a conversion is actually being recorded even once. Flagging it is the switch; recording it is a separate thing.

    Terms

    Key event
    — GA4's current name for what used to be called a "conversion": an event you've flagged as a business outcome, which is what ad platforms optimize toward.
    0/7Fail

  • PR-02Parameter missing rate

    When a field often arrives empty, every breakdown built on it is distorted. "(not set)" takes the top row and you cannot tell which segment is actually performing.

    Worst gap: dimension_1 (not set) 96%

    I don't understand what this means

    In plain terms — we check how often a key parameter arrives empty. When that happens a lot, any analysis broken down by that parameter can't be trusted.

    Terms

    Parameter
    — Extra detail sent along with an event — for example, a purchase event might carry the item name and price as parameters.
    (not set)
    — What GA4 shows in a report when it didn't receive a value for that dimension. A lot of it is a sign of a gap in tagging or consent.
    0/5Fail

  • CV-03Conversion value collection

    Without the purchase amount, you cannot judge performance by revenue. ROAS and revenue-by-campaign cannot be calculated at all, so budget gets split on conversion counts alone.

    purchase carries no value/currency at all — ROAS can't be computed

    I don't understand what this means

    In plain terms — we check whether your purchase event carries what it was actually worth — an amount and a currency. Counts alone can't tell you anything about revenue.

    0/4Fail

  • TG-03Cross-domain tracking

    If checkout sits on a different domain, the visitor becomes a new person at that moment. The link between the ad that brought them and the purchase they made is broken.

    Own domain appears as a referral (self-referral): host_1 — sessions break and attribution skews

    I don't understand what this means

    In plain terms — we check whether moving to a different domain, like at checkout, is still recognized as the same visit rather than a brand-new visitor.

    Terms

    Cross-domain tracking
    — The setting that keeps a visitor recognized as the same person and session when they move to a different domain, such as during checkout.
    0/2Fail
Run the same diagnosis on my GA4 — freeAsk about implementation

A real diagnosis of a property the owner runs, anonymised. Property name, hosts, custom event and dimension names and internal filter names were substituted, and the country and time zone removed. The verdicts and the figures are untouched.

← Other verticals