The PPC Audit We Actually Run: The Production Checklist With Findings From $18.1M Under Our Roof

A working PPC audit checklist runs in a strict order: conversion data quality first, budget control second, waste elimination third, bidding signal fourth, and creative and assets last. The order matters because each phase depends on the one before it. Negative keywords don't matter if the data deciding what "converts" is noise, and tuning a bid strategy is pointless when it's optimizing toward scroll events.

This is the checklist our audit system runs in production. We run 64+ active Google Ads accounts for 50+ clients (counted October 16, 2026). In all, $18.1 million in Google Ads spend has run through accounts we manage or have managed since 2015 (measured through September 30, 2026). In June 2026 we ran this audit across the whole portfolio at once. It covered a scan of 60 active accounts, conversion-health snapshots of 68 accounts, and a search-term collector that has swept our accounts weekly since mid-March 2026.

Where these numbers come from

"X of 60" figures come from the portfolio-wide scan we ran on June 2, 2026. "X of 68" figures come from our per-account conversion-health and asset snapshots of June 2026, which include a few house and inactive accounts. The landing-page and account-access checks come from companion audit collectors that had covered 64 and 66 accounts by June 2026. Search-term waste figures come from our collector's flags between March 19 and May 25, 2026, counted once per account and term. Every example comes from one of those accounts, and we don't name any clients.

Our portfolio skews nonprofit: 35+ of the accounts we run use the Google Ad Grant (counted October 16, 2026). We marked the two Grant checks below "if you run Ad Grants," and everything else applies to any account spending real money.

The 17-point PPC audit checklist

#	Check	How often we find a problem
1	Conversion actions measure real outcomes, not engagement	Conversion-action clutter in 35% of accounts; most damaging finding when present
2	No overlapping actions double-counting one lead	Common in nearly every deep conversion audit
3	Conversion tracking exists, and a test lead proves it	Occasional, catastrophic when found
4	Enhanced Conversions turned on	Missing in 79% of accounts
5	No enabled campaigns recording zero conversions	43% of accounts
6	Every landing page tagged, reachable and matched to its ad	All 64 accounts our landing-page auditor had covered
7	Paid accounts: impression share lost to rank vs budget, read before any budget change	7% of scanned accounts were budget-capped winners
8	Ad Grants: full use of the $10,000 monthly credit	Roughly half of Grant accounts underspending
9	Ad Grants: no campaign budgeted over the $329 daily cap	Occasional; a quick fix when present
10	No spend on zero-conversion search terms	87% of accounts
11	No disapproved-ad clusters choking delivery	23% of accounts
12	No dark or zombie campaigns	23% of accounts
13	Bid strategy matches the goal	Rare; mostly accounts unmanaged for years
14	Bid targets reviewed, not set and forgotten	Not scored in the June 2026 scan
15	Assets, extensions and PMax components complete	Asset gaps in 53% of accounts
16	Ad copy says something specific	Not scored in the June 2026 scan
17	Account access hygiene	75% of accounts have at least one access finding

Phase 1: Conversion data quality (checks 1 to 6)

Everything in a Google Ads account is downstream of its conversion data. Smart Bidding spends your money toward whatever the account calls a conversion. We always start here, because if that definition is wrong, the machine works hard at the wrong job.

1. List every conversion action and ask what it actually counts

Pull the full list of conversion actions and check which ones are set as Primary. The failure pattern on a paid account is engagement events set as Primary, such as 30-second page views, 90% scroll depth and driving-direction clicks. The account "converts" constantly while the bidding algorithm optimizes toward noise.

In one account, we found 1,352 conversions logged on roughly 392 clicks, spread across 38 separate conversion actions. Our June 2026 snapshots found 10 or more conversion actions set up in 24 of 68 accounts (35%). On a paid account the fix costs nothing. Demote engagement events to Secondary, and keep form submissions, calls and purchases as the Primary actions.

A Google Ad Grant account works differently. Micro-conversions, like a visit to a key page, give a Grant's bidding a signal to bid toward, and that helps it spend its credit. We keep a balance of micro and macro conversions on a Grant, and we report calls and forms on their own line.

2. Hunt for overlapping actions counting one lead several times

One user behavior should fire one Primary conversion. In practice, accounts pile up trackers over the years and nobody retires the old ones. One deep audit found five call-shaped conversion actions on at once, so a single tap-to-call could land in several of them. The same account valued an appointment form at a quarter of a plain tel-link tap, and its sister account valued the identical form at five times as much. Every report and every bid in the account rested on data that contradicted itself.

3. Confirm conversion tracking exists at all, then test it

It sounds too basic to check. Check it anyway. Our June 2026 snapshots found 10 of 68 accounts with zero active conversion actions, which means bidding had nothing to optimize toward. One was a home-services account that spent over $5,000 in the prior 30 days. This finding is catastrophic on an account spending real money, and it's always the first fix.

Tracking that exists can still break, so test it every month. Submit a test lead, check that the conversion fires in Google Ads, and confirm the count matches what your CRM or inbox shows.

4. Check Enhanced Conversions

Enhanced Conversions sends hashed first-party data alongside the conversion, which improves match rates and the signal Smart Bidding learns from. Our June 2026 snapshots found it off in 54 of 68 accounts (79%). It's usually a one-time setup task that pays back on every conversion afterward.

5. Find enabled campaigns recording zero conversions

In the same snapshots, 29 of 68 accounts (43%) had at least one enabled campaign with zero conversions over 30 days. One church Grant had 8 enabled campaigns at zero, including all five of its city-targeted local campaigns. The reflex is to call those campaigns failures and pause them. Usually the cause is a goal-mapping or tracking gap, where the campaign produces outcomes the account never records. Diagnose before you pause.

6. Walk the landing pages

Every ad's destination needs four things: it loads, it carries the tag plan, it matches the ad, and it gives the visitor something to do. All 64 accounts our landing-page auditor had covered by June 2026 had at least one tracking or quality finding. One multi-page microsite had Google Tag Manager on none of its five audited pages. Another account had four unreachable landing pages, its appointment-request page among them, so its ads were buying clicks to pages that didn't load.

Matching the ad means a search ad for one service lands on that service's page, not your homepage. The visitor finds the thing they searched for without hunting.

Phase 2: Budget control (checks 7 to 9)

7. On a paid account, read impression share lost to rank vs lost to budget before touching any budget

On a paid account this is the most useful check in the audit, because the two readings call for opposite fixes. A campaign losing most of its impressions to rank won't gain from a bigger budget. It needs Quality Score work, tighter ad groups and better ads. A campaign losing impressions to budget while it converts at a cost you like can grow with more budget.

Those budget-capped winners are rarer. In our June 2, 2026 scan, 4 of 60 accounts (7%) had campaigns with good costs per acquisition held back by budget. One paid account was losing about 76% of its impression share to budget on campaigns that converted fine, and others were losing 29% to 37%. The fix there is a budget line, not a rebuild.

Don't use the rank figure on a Google Ad Grant account. The Grant's bid limits cap how much rank an account can buy, so the figure sits at Google's top reading on almost every Grant. It describes the program, not your account. Budget-lost impression share still counts on a Grant, because it maps to credit you aren't spending.

8. If you run Ad Grants: check how much of the $10,000 credit gets used

Google's Ad Grant gives a qualifying nonprofit up to $10,000 a month in search ads, and Google pays for the ads. The audit checks how much of it the account uses. In our June 2, 2026 scan, 17 of 35 Grant accounts were spending under $8,000 of the credit, and 7 were under $5,000. The worst case was a Grant account using less than 1% of its $10,000.

Underspending is almost never a budget-settings problem, so raising campaign budgets rarely fixes it. Look for the cause instead: keyword coverage, Quality Score, limited serving, ads waiting on approval, or throttling after two months under the 5% click-through rate floor. That's how we run Google Ad Grant management. In September 2026, on average 80.9% of the $10,000 monthly credit was spent across the 36 Grant accounts we ran all month.

9. If you run Ad Grants: check every campaign against the $329 daily cap

Google spends at most $329 of Grant credit a day. A campaign budgeted above that can't use the rest. We found a PMax campaign and one other campaign budgeted over the cap, and both sat in otherwise healthy accounts that were converting. It doesn't happen often, and the fix takes a minute. Bring that campaign down to the cap and move the rest of its money to other campaigns, so the account's total doesn't drop.

Phase 3: Waste elimination (checks 10 to 12)

10. Pull search terms with spend and zero conversions

Our search-term collector flagged 8,706 unique search terms with spend and zero conversions between March 19 and May 25, 2026, across 55 of our 63 accounts (87%). Those flags cover ten weeks, not a year, but the pattern holds account by account: nearly every audit finds money riding on queries that have never converted.

One church Grant kept spending credit on "sermon audio" and "what is the sinner's prayer," with zero recorded conversions. One pregnancy center paid for clicks on searches for no-cost baby products, and another kept paying for "gynecologist" searches that never converted. Broad match leakage lives in this bucket too, because irrelevant query families riding broad keywords show up here as zero-conversion terms. The fix is negative keywords, added every week, not once at launch. Our search terms study shows what 1.3 million rows of this data look like.

11. Check for disapproved-ad clusters

Single disapprovals are routine, but clusters quietly choke delivery. In our June 2, 2026 scan, 14 of 60 accounts (23%) had disapproved ads. In the worst case, Google disapproved 10 of 14 enabled ads (71%) in an account's core campaigns, so its best campaigns could barely serve. Another account had half its ads disapproved. Nobody in those accounts had gotten an alert, because Google's warnings about disapprovals are easy to miss.

12. Find the dark and zombie campaigns

Enabled doesn't mean running. In the same scan, 14 of 60 accounts (23%) had campaigns switched on and delivering nothing, with zero impressions or zero spend over 30 days. We found paused budgets, over-tight targeting, billing issues and leftover $1-a-day budgets behind them. One paid account delivered $0 and zero impressions over 30 days with ads enabled. Another had 4 of its 7 campaigns at zero impressions.

Phase 4: Bidding signal (checks 13 and 14)

13. Confirm the bid strategy matches the goal

Manual CPC on a lead gen account, or Target Impression Share where conversions are the goal, means the account is steering with the wrong tool. It's rare in our data, one or two accounts out of 60, and it shows up in accounts nobody has managed for years. The cost is real, though. One paid account's lone Manual CPC campaign converted at a high cost per acquisition while losing 60% of its impressions to rank. The fix was conversion-based bidding, which is only safe once checks 1 through 6 have made the conversion data worth bidding on. That's why this check is number 13 and not number 1.

14. Check that someone reviews the bid targets

Target CPA and Maximize Conversions need calibration, and a target set once and never revisited drifts away from the account it was set for. Set a target CPA too tight and Google throttles your impressions. Set it too loose and you overpay for each lead. A sound order is Maximize Conversions first to gather data, then Target CPA once the campaign has at least 30 conversions in a 30-day window. Whatever the target, put a date on the calendar to look at it again.

Phase 5: Creative and assets (checks 15 and 16)

15. Audit assets, extensions and PMax completeness

In our June 2026 snapshots, 36 of 68 accounts (53%) had asset gaps: too few images, fewer than four sitelinks per campaign, missing callouts or no structured snippets. Every one of the 29 accounts running PMax had at least one PMax gap. The gaps ran from an asset group missing its logo, business name and final URL to campaigns running with no audience signals. These are the cheapest fixes on the list. Sitelinks, callouts and call assets make your ad bigger at no extra cost per click, and Google counts assets in Ad Rank.

16. Read the ads and look for specifics

An ad that says "We offer great service" competes against ads with specific numbers and outcomes, and it usually loses. Look for concrete details in every ad: response times, years in business, customer counts or prices. Responsive Search Ads give you 15 headline slots, so use them to test specific offers instead of repeating vague benefits.

The housekeeping check most audits skip

17. Account access hygiene

It's the lowest-severity item and still worth two minutes. By June 2026, 50 of the 66 accounts in our access audit (75%) had at least one access finding. One account had zero users with direct access, so no person could manage it. Another had 5 access invitations pending for over 30 days. Several had exactly one admin, which is one departure away from an orphaned account.

How to run this on your own account

You can work this list yourself in an afternoon. Go in order. Phase 1 is the part most owners skip because it's the least visible, and it's where the expensive problems live. If you only do one phase, do that one. Then spend real time in your search terms report, for check 10's waste and for the words your buyers use, which belong in check 16's ads.

If you'd rather hand the fixes to us, our pricing page lists every plan and its price. You can also book a call to talk it through.

Questions people ask

What should a PPC audit checklist include?

It should cover five layers in this order: conversion data quality, budget control, waste elimination, bidding signal, then creative and assets. The fifth layer includes ad copy. Finish with account access hygiene. The order matters because every later check depends on clean conversion data.

What is the most common problem found in a PPC audit?

The most widespread findings in our June 2026 data were landing-page tracking or quality gaps and missing Enhanced Conversions. The landing-page gaps showed up in all 64 accounts our landing-page auditor had covered, and 79% of accounts were missing Enhanced Conversions. The most damaging is engagement events counted as Primary conversions on a paid account. Smart Bidding then spends toward that noise.

How much waste does a PPC audit find?

Our search-term collector flagged 8,706 unique search terms with spend and zero conversions between March 19 and May 25, 2026, across 55 of its 63 accounts (87%). The count is specific to our portfolio, but the pattern holds: nearly every account has money riding on queries that have never converted.

How often should you audit a Google Ads account?

Run the full audit when you take over an account, then about once a year. Run the lighter checks on a schedule: search terms weekly for the first 90 days and a test lead every month. Check disapproved ads and budget caps daily if you can automate it. Most of the costly findings in our data sat unnoticed for months because nobody had scheduled the check.

What does a Google Ad Grant audit check that a regular PPC audit does not?

It checks three things: how much of the $10,000 monthly credit the account uses, the $329 daily cap, and the 5% minimum click-through rate rule. In our June 2, 2026 scan, 17 of 35 Grant accounts were spending under $8,000 of the credit. That's usually a keyword coverage and Quality Score problem, so the fix is tighter keywords and better ads, not bigger campaign budgets.

See every Google Ads plan and price