Skip to content
Bipin Kumar
Multi-accountMeta + Google Ads

I Audited My Own Ad Accounts Before Optimising Them

35% of the spend was reporting numbers I could not trust.

₹1.11L

documented spend

Documented

5

accounts

Documented

₹38.9K

associated with unreliable, unverified or inflated measurement

Calculated

34.9%

of documented spend

Calculated

The situation

Five ad accounts across Meta and Google were reviewed before treating performance numbers as optimisation truth.

The order matters. Optimising against a conversion column you have not checked is not optimisation — it is guessing with extra steps. So before any budget decision, the question was narrower and duller: do these numbers describe something real, and do they describe the same thing consistently?

For roughly a third of the documented spend, the answer was no.

The exact technical cause of every issue was not established from the exports alone, and none is asserted here.

What I checked

Eight questions, asked of every account before any of its numbers were used to justify a decision.

  1. 01

    Conversion recording

    Whether the account records conversions at all, and in which column.

  2. 02

    Conversion plausibility vs clicks

    Whether reported conversion counts and rates are arithmetically possible given recorded click volume.

  3. 03

    Conversion diagnostics

    The platform’s own status flags on each conversion action.

  4. 04

    Placeholder values

    Conversion values that are defaults or stand-ins rather than real business values.

  5. 05

    Agreement between reports

    Whether campaign-level and account-level figures reconcile with each other.

  6. 06

    Measurement parity across campaign types

    Whether different campaign types in the same account are measured on comparable terms.

  7. 07

    Conversion-action status

    Whether each conversion action is verified, active and currently recording.

  8. 08

    Platform events vs business outcomes

    Whether a recorded platform event can be tied to a defined business outcome at all.

The scale of the issue

  1. ₹1,11,000+

    Documented spend reviewed

  2. ₹38,898.65

    Spend where measurement did not hold up

  3. 34.9%

    Share of documented spend

The middle figure is the sum of the three affected areas set out below. It is spend where the reporting could not be trusted — not spend that was necessarily wasted, which is a different and unproven claim.

Four failure modes

01

Inflated conversion reporting

Reported conversion rates that cannot be reconciled with recorded click volume.

One Google account reported campaign-level conversion rates far above 100%, alongside an account-level figure of 256%.

A conversion rate above 100% is not automatically an error — duplicate counting, multi-conversion actions and attribution settings can all produce one legitimately. What it is, always, is a signal to stop and check before the number is used to justify a budget change.

₹10,638.93

spend in the affected account

Documented

256%

account-level reported conversion rate

Reported
Campaign-level reported conversion rates in the affected account
CampaignReported conversion rate
Campaign 1345%
Campaign 2476%
Campaign 3160%
Campaign 4133%
Account level256%
Campaigns are numbered rather than named here because the point is the pattern across the account, not any individual campaign.

The exports show mathematically implausible conversion reporting; the precise technical cause was not established from the exports alone.

02

Placeholder values

Conversion values recorded as ₹1 rather than a real business value.

Three campaigns contained ₹1 conversion values.

A ₹1 value is a placeholder, not a price. Any downstream figure built on it — conversion value, value per conversion, ROAS — inherits the placeholder and reports it as though it were revenue.

3

campaigns with ₹1 conversion values

Documented

Value-based reporting was therefore not meaningful for those campaigns, and no value-based metric from them appears anywhere on this site.

03

Unverified or inactive conversion actions

Conversion actions present in the account but not verified and not currently recording.

In one preschool account, the platform’s own diagnostics flagged a set of conversion actions as unverified or dormant, while campaigns in that account continued to spend.

The Performance Max campaign in the same account is the clearest illustration: substantial click volume at a low cost per click, a lead-funnel figure reported in one place, and nothing at all in the conversions column.

₹10,979.48

spend in the affected account

Documented

2

conversion actions unverified

Documented

9

with no recent conversions

Documented

0

actively recording

Documented

A campaign recording 0 conversions in the platform’s conversion column is a statement about measurement, not about whether the campaign produced enquiries. This review does not establish that it produced none.

04

Incomplete or zero-result tracking

Spend running against campaigns with no usable recorded outcome.

Across two further preschool accounts and a set of Meta website lead campaigns, spend accumulated against tracking that was incomplete or recorded no results.

This is the least dramatic failure mode and the most common one. Nothing looks broken on the surface; the account simply cannot answer the question it is being asked.

₹17,280.24

combined spend

Documented

3

sources: two preschool accounts and Meta website lead campaigns

Documented

This describes the state of the recorded tracking. It does not establish that these campaigns generated no enquiries, and it does not identify the technical cause.

The three spend figures above — ₹10,638.93, ₹10,979.48 and ₹17,280.24 — sum to the ₹38,898.65 quoted at the top of this page.

Two campaign types, two measurement states

Two campaign types in the same account, presented side by side because they were measured on completely different terms — not because one of them won.

Performance Max

Spend
₹8,637.27Documented
Clicks
8,537Documented
CPC
₹1.01Calculated
Recorded conversions
0Verified

Search

Spend
₹928.39Documented
Clicks
7Documented
CPC
₹132.63Calculated
Recorded conversions
4Verified

This is not evidence that Search outperformed PMax. The sample and measurement conditions are not comparable enough to make that conclusion.

The central lesson

What this review does not establish

Limitation

This review was carried out on exported reporting. It does not establish:

  • Revenue
  • Qualified leads
  • Sales
  • Bookings
  • ROAS
  • Reliable cost per acquisition
  • The precise technical cause of any issue
  • Actual business outcomes
  • Lead quality

This was a review of exported reporting, not a technical tracking audit of the live accounts. A live audit — tag firing, event parameters, conversion action configuration, deduplication — is a separate piece of work and would be needed to establish causes.

Related

Service this evidences

Related case study

Scaling Meta Lead Generation in Patna

The Meta lead campaigns whose spend this audit examined.

Next case study

Diagnosing Preschool Google AdsRead case study

Not sure whether your conversion data is trustworthy?

That is exactly what a free ad audit is for.

Get a Free Ad AuditChat on WhatsApp