# Conversion Tracking Checker

> Free self-audit of Google Ads, GA4 and Meta conversion tracking: 14 yes/no/not-sure questions, a weighted score out of 100 and a list of likely gaps, each linked to a fix.

Source: https://sterling.company/conversion-tracking-checker · Machine-readable spec: https://sterling.company/conversion-tracking-checker/spec.json · Publisher: Sterling Lab

## Short answer

The Conversion Tracking Checker is a free self-audit of Google Ads, GA4 and Meta conversion tracking. You answer 14 yes, no or not-sure questions and get a score out of 100 and a list of likely gaps, biggest first, each linked to a fix. It does not scan your website or your ad accounts, and your individual answers are not sent anywhere.

Nothing is scanned: the checker only scores the answers you give. The weights and band cut-offs are our own definitions, not a benchmark.

## How the score works

Each check has a weight of 1, 2 or 3. Yes earns the full weight, Not sure earns half, No earns nothing. A check marked not applicable is left out of both sides of the sum.

**score = round((sum of weights answered yes + 0.5 × sum of weights answered not sure) ÷ sum of weights of the applicable checks × 100). A check answered na is left out of both sums.**

The checks with weight 3 are core checks. A no or not sure on any of them keeps the band out of solid, even when the score is 80 or higher.

The fix list is sorted by points lost, biggest first: a No loses the full weight of the check, a Not sure loses half of it. Checks that lose the same number of points keep the order in which they are listed below.

## Bands

- **80 to 100**: Solid on paper. The remaining items are tidy-up work. Next step: Fix the remaining items when convenient, then re-run the check after any change to tags, consent or the shop.
- **50 to 79**: Gaps worth fixing. Some numbers that bidding relies on are partial or guessed. A failed or unconfirmed core check holds you here even with a score of 80 or more. Next step: Start with the first item in the fix list. If a core check is named, fix that one before anything else.
- **0 to 49**: Hard to trust. Fix the top items before you change bids or budgets. Next step: Do not change bids, budgets or targets yet. Fix the top three items, then run the check again.

## Worked example

A service business that closes deals by phone and does not run Meta ads: the two Meta checks (capi, eventid) are not applicable. A test enquiry fired the conversion tag twice, so "One real event, one conversion" is No. Leads are sent without a value, so "Real value and order ID" is No. The consent banner check and the tagged-links check are Not sure. Everything else is Yes.

1. Applicable weight: 28 − 2 − 2 = 24
2. Points lost: 3 + 2 + 1 + 0.5 = 6.5, so earned is 24 − 6.5 = 17.5
3. Score: 17.5 ÷ 24 × 100 = 72.9, shown as 73
4. Band: 73 is in 50 to 79, so gaps worth fixing
5. Gaps in order: One real event, one conversion (3 points lost); Real value and order ID (2); Banner updates Google consent (1); Tagged links keep their source (0.5)

Steps to compute any result: remove not-applicable checks from the total; lose the full weight for each No and half for each Not sure; earned = applicable weight − points lost; score = round(earned ÷ applicable weight × 100); then apply the band and core rules; sort the gaps by points lost, biggest first.

## The 14 checks (28 points in total)

### What gets counted

**Primary conversions are sales or leads** (weight 3, core)

- Question: Only actions that mean money (a purchase, a quote request, a booked appointment, a real call) are primary conversions in Google Ads.
- How to check: Goals → Conversions: page views, button clicks and newsletter sign-ups should be secondary or off.
- If No or Not sure: Bidding may be optimizing toward non-sales. When page views or clicks are primary, bidding chases whichever action is cheapest to get, which is rarely the one that pays.
- Fix: [Primary vs secondary conversions](https://sterling.company/blog/primary-secondary-conversions-smart-bidding)

**One real event, one conversion** (weight 3, core)

- Question: One real submission or test order is counted once: the conversion tag fires a single time, and no second primary action in Google Ads counts the same event.
- How to check: Preview your tag container or open Google Tag Assistant, submit a test and count the fires. Then open Goals → Conversions in Google Ads and look for two primary actions on the same event, such as a Google tag plus a GA4 import. Two fires or two actions mean double counting; no fire suggests the tag is not firing.
- If No or Not sure: Conversions may be double counted or missing. One submission can be counted twice, for example by a Google tag and by a GA4 import set up as two primary conversions for the same event. A tag that is installed is not a tag that works until a test proves it.
- Fix: [Google Ads conversion tracking setup](https://sterling.company/guides/google-ads-conversion-tracking-setup)

**Real value and order ID** (weight 2)

- Question: Purchases are sent with their real order value and order ID (leads with the value you assigned them).
- How to check: In Tag Assistant, your container preview or GA4 DebugView, open a test purchase and check the value, currency and transaction_id parameters against the real order.
- If No or Not sure: Conversion value is missing or wrong. ROAS divides credited revenue by spend. Without a real value the revenue half of that formula is empty or made up.
- Fix: [How ROAS is calculated](https://sterling.company/guides/how-is-roas-calculated)

**Forms, calls and bookings tracked** (weight 2)

- Question: Every way a lead reaches you that matters (forms, phone calls, bookings) is tracked, including forms that show a message without loading a new page.
- How to check: List the ways customers contact you, then check each one has its own conversion action.
- If No or Not sure: Some leads never reach the ad platform. A form that confirms on the same page has no thank-you URL to fire on, and calls need their own tracking. Leads that are not recorded do not exist for bidding.
- Fix: [Google Ads conversion tracking setup](https://sterling.company/guides/google-ads-conversion-tracking-setup)

### Google tags and consent

**GA4 key events are real conversions** (weight 2)

- Question: The key events in GA4 match your list of real conversions, with no scrolls or button clicks counted next to real leads.
- How to check: GA4 Admin → Key events. Write down what each one is in real life.
- If No or Not sure: GA4 counts activity as conversions. A button click or scroll marked as a key event makes the total look healthy while real inquiries stay flat, and the reverse happens when the one event that matters was never turned on.
- Fix: [A GA4 audit you can finish in one afternoon](https://sterling.company/guides/ga4-audit)

**Enhanced Conversions sends data** (weight 2)

- Question: Enhanced Conversions is on, and the conversion tag actually sends the hashed customer data.
- How to check: Google Ads → Goals → Conversions → Summary → open the action → Diagnostics tab shows whether Enhanced Conversions receives user-provided data.
- If No or Not sure: Enhanced Conversions may be switched on but empty. Turning the setting on in Google Ads is not enough: the tag on the confirmation step has to include the customer data object when it fires.
- Fix: [Enhanced Conversions on WooCommerce and Shopify](https://sterling.company/blog/enhanced-conversions-stack-woo-shopify)

**Banner updates Google consent** (weight 2)

- Question: After a visitor clicks Accept on your cookie banner, Google's tags receive a consent update that covers all four Consent Mode v2 signals.
- How to check: DevTools → Network → filter “collect”, click Accept. gcd carries the signals: after Accept its 3rd and 4th letters should be r, t, v or n (granted); p, q, m or u mean still denied, l means no consent signal at all. gcs shows only ad_storage and analytics_storage (G111 = both granted).
- Not applicable when: No banner needed
- If No or Not sure: The cookie banner may not pass consent to Google. A banner can save its own cookie and never call the consent update. GA4 can then keep behaving as if consent was refused, even for visitors who accepted.
- Fix: [Consent Mode v2 and missing GA4 conversions](https://sterling.company/blog/consent-mode-v2-ga4-attribution-gap)

**Tagged links keep their source** (weight 1)

- Question: Paid, email and social links carry UTM tags (auto-tagging stays on in Google Ads), so their visits appear in GA4 under their own source and medium.
- How to check: GA4 → Traffic acquisition: your campaigns should appear under their own source and medium. A large or growing Direct or Unassigned row next to active campaigns suggests untagged links.
- If No or Not sure: Paid and email visits may land in Direct. A link with no tags gets a guess, and the guess is often Direct. Credit then moves away from the channel that earned the visit.
- Fix: [UTM parameter best practices](https://sterling.company/guides/utm-parameter-best-practices)

**Source survives external checkout** (weight 1)

- Question: When payment or booking happens on another domain, the visitor's original source survives the trip back.
- How to check: Place a test order from a tagged link and check which source GA4 credits it to.
- Not applicable when: All on our domain
- If No or Not sure: External checkout may reset the source. A payment or booking page on another domain can start a fresh session, so the sale can be credited to the payment provider as a referral instead of the ad.
- Fix: [A GA4 audit you can finish in one afternoon](https://sterling.company/guides/ga4-audit)

### Meta

**Conversions API next to the pixel** (weight 2)

- Question: Meta receives your key events from a server (Conversions API) as well as from the browser pixel.
- How to check: Events Manager: each key event should list both the browser and the server as sources.
- Not applicable when: No Meta ads
- If No or Not sure: Meta sees only what the browser lets through. Browsers block scripts, drop cookies and close tabs before a request finishes. Sending the same event from a server fills some of those gaps.
- Fix: [Meta Conversions API setup](https://sterling.company/guides/meta-conversions-api-setup)

**Matching event name and ID** (weight 2)

- Question: The pixel and the server send each event with the same event name and the same event ID, so Meta merges the two copies and marks one as Deduplicated.
- How to check: Events Manager → open the event and look at its deduplication details, or Test events if your server sends a test_event_code: the browser and server copies should be merged, one marked Deduplicated. A different event name or event ID is the usual reason they are not.
- Not applicable when: No server events or no Meta ads
- If No or Not sure: Meta may count each order twice. Unless the event name and event ID both match, Meta cannot merge the browser and server copies, so conversions look inflated and optimization learns from numbers that are too high.
- Fix: [Conversions API and the duplicate-event trap](https://sterling.company/guides/facebook-conversion-api-gtm-server-side)

### Back to real sales

**Click ID stored with each lead** (weight 2)

- Question: Each lead or order stores a click ID (gclid, or gbraid and wbraid for some iOS traffic; fbclid for Meta) or contact details, so a closed sale can be traced back to the ad click.
- How to check: Open your last five sales and try to trace each one to the click that began it.
- Not applicable when: All sales close online
- If No or Not sure: Closed sales cannot be matched to ads. An offline conversion import is a matching exercise. Without a stored click ID or contact details the platform has no way to place the sale.
- Fix: [Offline conversion import](https://sterling.company/guides/offline-conversion-import)

**Monthly check against real sales** (weight 3, core)

- Question: In the last month someone compared the conversions the ad platforms report with real orders or leads in your CRM, inbox or call log.
- How to check: Compare last month's conversions in Google Ads and Meta with the orders or leads in your CRM, inbox or call log. A steady gap is normal; a sudden drop, or a number that stops moving, means something broke.
- If No or Not sure: Ad numbers are not checked against real sales. Tracking usually breaks later, during ordinary site maintenance, and nobody connects the two events. A monthly comparison catches it sooner.
- Fix: [Setting up conversion tracking that stays working](https://sterling.company/guides/setting-up-conversion-tracking)

**Accounts under your business login** (weight 1)

- Question: Your Google Ads account, GA4 property and tag containers are under your business login, with you as admin.
- How to check: Check the user list in each account for your own business email with admin rights.
- If No or Not sure: Your tracking history belongs to someone else. Conversion history lives in the account and leaves with whoever owns it. Anyone helping you should be added as a user, not the other way around.
- Fix: [Setting up conversion tracking that stays working](https://sterling.company/guides/setting-up-conversion-tracking)

## FAQ

**Does the conversion tracking checker scan my website?**

No. It asks questions and scores your answers. Nothing is loaded from your site or your ad accounts, so the result is only as good as the answers. Each question has a short note on how to check it yourself.

**Why does “Not sure” count as half?**

An unknown is not a confirmed failure, but it is not a pass either: nobody is watching that check. Half the weight keeps it visible in the score and on your list.

**Is the score a benchmark against other businesses?**

No. The weights and the 80 and 50 thresholds are our own definitions, chosen to rank what matters most for bidding. They are not drawn from industry data, and the score does not compare you with anyone.

**What happens to my answers?**

They stay in your browser tab, are not sent anywhere, and disappear when you leave or start over. Our site analytics receives only an aggregate: that the check was started and finished, with the final score, the band and the number of gaps, and the score again if you click the audit link. The individual answers are never sent.

**Why can a high score still stay in the middle band?**

The three checks with weight 3 decide what Smart Bidding and your reports learn from. A No or a Not sure on any of them keeps the result out of the top band, even when the score is 80 or higher, because 25 out of 28 points can hide a wrong primary conversion, and a Not sure means nobody is watching that check. This rule is our own definition, like the weights and thresholds.

**What does the free audit add on top of this?**

We look inside your Google Ads, GA4 and Meta accounts and test the same points directly, instead of relying on answers, then send a written report with the fixes in order.

## Next step

If you want the same points tested inside your Google Ads, GA4 and Meta accounts, request the [free conversion tracking audit](https://sterling.company/services/conversion-tracking-audit).
