The Advertising Waste Diagnostic

Four kinds of wasted spend, different enough that one word for all of them is the main reason audits go wrong. Every dollar you question ends up in exactly one of these, and the label decides what you are allowed to do about it. Throughout, a slice is any cut of the spend you can name and put a dollar figure on: a campaign, a search term, a city, an hour, a placement, a network.

The output is not a percentage. It is a short list of findings, each carrying the rule that condemned it, plus one honest number for the spend nobody can see, plus one test that would settle the largest remaining question.

Name the decision before you start, because it changes what counts as a finding. Deciding whether to keep spending at all is a different audit from deciding which of two channels gets next quarter’s budget, which is different again from deciding whether the account is worth handing to someone else. An audit with no decision attached is an inspection, and inspections get filed.

Recent enough that the customers are still identifiable, old enough that the slow ones have closed. For most businesses that is the month before last.

Run every step below on that single month. One period keeps the arithmetic honest: the same denominator under every finding, and no quiet mixing of a good month’s spend with a bad month’s outcomes. Annualizing anything you find is a separate claim, made afterwards, with the assumption that the month was typical written next to it.

Pull all of it before you read any of it. Reading as you export is how the first interesting row becomes the conclusion.

The step nobody writes down, and the one that stops most people before they start.

First and non-negotiable. Until the platform’s count is compared against your own records, every other number in the account inherits an unknown error.

Match lead by lead and order by order, then walk each platform conversion down the ladder and write the number at every rung. One block per platform, because they cannot be added.

Watch which rung the numbers fall off, because one of them is not an advertising problem at all. A heavy drop between found in your records and a reachable person usually means inquiries are arriving and nobody is getting to them in time. No bid change fixes that, and no amount of better targeting survives it. When the reconciliation lands there, the work moves to how the business answers its phone, and the Operations Diagnostic is the better tool for it.

The definition of qualified belongs to whoever lives with the consequences of getting it wrong, which is rarely the person running the ads. Get it written down before you argue about lead quality, and date it. A definition that quietly moves month to month lets anyone improve lead quality by changing the label, and nobody can tell afterwards whether the advertising got better.

Every finding takes this shape. The fields are there because each one is a way audits go wrong when it is missing.

Work down the cuts in this order. The first two decide whether the rest can be trusted, which is why tuning bids and audiences comes last on every experienced practitioner’s list.

Every dollar of the month on exactly one line: the four kinds of waste, plus a fifth line for the spend you matched to customers. Each line carries its criterion. This is the page the whole audit is for.

For the largest thing the evidence cannot settle. One at a time, designed before anything switches off.

A quarter later, close the loop: the leakage you suspected, the change you made, the outcome you expected, the change you observed, and whether the saving was still there. Savings that move rather than persist are the most common way an audit gets remembered as a success it was not.

Written so that the person who built the account, an accountant, or another analyst could check every line and disagree with it specifically.

• The reconciliation, one block per platform, with attribution rules labeled.
• The findings, each in the format above, ordered by size within their kind.
• The opaque number, stated plainly, with what produced it.
• The exclusions, meaning what you looked at and could not conclude anything about, and what would have been needed.
• One test, with its prediction written before it runs.
• The claims you are refusing to make, which is usually a total, an annualized figure, and any claim that the advertising created customers, which no holdout has earned yet.

That last item does more work than it looks like it should. An audit that names its own limits is the one a business owner can act on without wondering what was oversold.

Paste this into any AI assistant along with your exports. It runs the whole method in order and will not hand you a percentage.

I want you to help me audit my advertising spend. Your job is not
to optimize my account or recommend tactics. Your job is to work
out which dollars are not buying customers, and to be exact about
how confident we can be for each one.

I will give you exports: platform spend and conversions, various
performance cuts, and records from my own system of record (CRM,
orders, bookings, or calls) for the same month.

Work through these stages, in order, and do not skip ahead:

1. INVENTORY THE EVIDENCE
List what I gave you and what is missing. For anything missing,
say whether it exists and I did not send it, or whether the
platform does not publish it at all. Do not proceed as though a
missing report is a zero.

2. RECONCILE
For each platform, match reported conversions against my own
records and walk them down this ladder: reported, found in my
records, a reachable person, qualified, became a customer, revenue.
Label every number with the rule that produced it. Never add
figures from two platforms together. Say which matches are
certain, which are probable, and which you cannot make.

3. CLASSIFY
Put every questionable dollar in exactly one bucket:
  VISIBLE - directly observed in an export I gave you.
  MIS-MEASURED - counted as a result I cannot find in my records.
  OPAQUE - the platform does not disclose enough to classify it.
  CANNIBALISTIC - credit for demand that likely existed anyway.
  Always mark this one as unsettled.
Where one finding straddles two buckets, split its dollars and
cite the same finding on both lines. Cite the specific rows that
put each dollar where you put it.

4. SIZE AND STATE THE CRITERION
For each finding: the signal, the evidence, the criterion, the
kind, the size in dollars for this month, and the move. Write the
criterion as a sentence the person who built this account could
argue with. If you cannot write that sentence, it is not a finding
yet.

5. CHALLENGE YOURSELF
For each finding: what evidence contradicts it, what else could
explain the pattern, and what I would need to pull to confirm it.
Flag every slice where the conversion count is too low to support
a conclusion and leave those unclassified.

6. DESIGN ONE TEST
For the largest thing the evidence cannot settle, propose a single
holdout: what stops, where, what stays on as a comparable control,
which outcome from my own records we measure, the prediction if
the spend is incremental (creating customers who would otherwise
not have come), how long, and what result would mean we were
wrong. Work out whether my conversion volume can support it, using
roughly 16 divided by the square of the fractional difference we
want to detect as the outcomes needed per group, and say so if it
cannot.

7. WRITE THE BRIEF
The reconciliation, the findings ordered by size within their kind,
the opaque number, what you could not conclude anything about, the
test, and the claims you are refusing to make.

Rules throughout:
Never give me a single percentage of spend wasted.
Do not call a cost per lead high or low unless I have told you
what a customer is worth after the cost of serving them.
Do not call a lead qualified unless I have given you the
business's own written definition of the word.
If the evidence shows inquiries arriving and not being reached,
say plainly that the binding problem is outside the ad account.
Never quote an industry waste statistic. If a number comes from
outside my data, state its universe, its denominator, its failure
definition, whether it was observed or modeled, and whether
whoever published it sells the remedy.
Never invent a number that is not in my exports.
Do not treat a platform's conversion count as a customer.
Do not treat absence of evidence as evidence of good spend, or as
evidence of waste.
Optimization advice comes last, and only for findings that survived
stage 5.
It is fine, and often correct, to conclude that the evidence cannot
settle something.