Building a Google Ads Report That Survives Client Scrutiny

What to put in a paid search report so it answers the questions a sharp client asks, and how to defend every number in it.

· QueryCut

The short version

  • Open with conversions, cost per conversion, value, and spend tied to a named goal, not with impressions or CTR.
  • Derive your zero-conversion cut threshold from the client's own numbers: target CPA divided by CPC gives break-even clicks, then wait two to three times that before trusting a cut.
  • Show the terms you cut and the negatives you added, not just the wins, and put the last-click date next to every decision.
  • Call out PMax's limited query data, attribution lag, and budget-capped campaigns before the client does.
  • On an inherited account, audit conversion tracking first because every other number depends on it.

A client-ready Google Ads report answers three questions before they are asked: what did the money buy, what did you change and why, and what happens next. Lead with conversions and cost per conversion tied to the client's actual goal, show the search terms and placements driving (and wasting) spend, and attach a decision rule to every action so a skeptical reader can check your logic instead of taking it on faith. Everything below is built so that when a client pushes back on a line, you already have the answer.

Start with the outcome, not the activity

The fastest way to lose a client's trust is to open with impressions and click-through rate. Those are inputs. The client paid for outcomes. Your first screen should be conversions, cost per conversion, conversion value, and spend, compared against the same period last month and the same period last year if the account is old enough to have that data. Everything else in the report exists to explain movement in those four numbers.

Pick the comparison window on purpose. Month-over-month captures your recent changes but gets distorted by seasonality and by months having different numbers of weekdays. Year-over-year controls for seasonality but mixes in a year of unrelated changes. Show both and name the distortion out loud: if last month had an extra weekend, say so, because a client who notices it before you do will wonder what else you missed.

Tie the conversion number to a specific action the client recognizes. 'Conversions' means nothing if they do not know whether it counts a form fill, a phone call, or a newsletter signup. In the report, write the literal conversion action names you are counting and whether you are reporting 'Conversions' or 'All conversions', because those two columns can differ substantially and the gap is the first thing a numerate client will ask about.

Show the search terms, including the ones you cut

The search terms report is where a report earns its credibility, because it shows the client what their money actually matched to. Pull it from Campaigns > Insights and reports > Search terms. Include the terms that converted, the terms that spent with no conversion, and the negatives you added this period. Showing the cuts matters more than showing the wins: anyone can report a good week, but demonstrating that you actively removed waste is what separates management from autopilot.

Performance Max is the awkward exception. There is no full search terms report for it in the standard UI. You get a limited view under the campaign's Insights tab (search categories and themes, not raw queries), and it is less granular than Search campaign data. Be honest about this in the report. Write that PMax does not expose complete query-level data, show the search category themes you can see, and note that negative keyword control for PMax is more limited than for Search. A client who later reads that PMax is a 'black box' elsewhere will respect that you flagged it first.

If you manage both Search and PMax, separate their numbers rather than blending them. PMax will often claim conversions that Search would have captured anyway, and a blended cost per conversion hides which campaign type is actually pulling its weight.

Attach a decision rule to every cut

The hardest question a client asks is 'why did you turn that off?' If your answer is a feeling, you lose. If your answer is a rule you apply consistently, you win, even when a specific cut turns out to be wrong. So state your rule in the report and apply it the same way every time.

Here is the mechanism behind a defensible zero-conversion cut, and why the thresholds are what they are. A search term with zero conversions is only evidence of a bad term if it has had enough clicks that you would have expected at least one conversion by now. Work it backward from the client's target cost per conversion. If the target CPA is your average cost per click times some number of clicks, then roughly that many clicks is the point at which one conversion 'should' have landed. Say CPC is $2 and target CPA is $100: you expect about one conversion per 50 clicks, so a term at 30 clicks and zero conversions tells you almost nothing, because a single conversion you simply have not gotten yet would have put it right on target. At 150 clicks and zero conversions, the term has had three chances to hit your CPA and missed all of them, which is real evidence.

That is why a flat '100 clicks' rule is a starting point, not gospel. Derive your own threshold: take your target CPA, divide by your average CPC to get your break-even clicks-per-conversion, and wait for roughly two to three times that many clicks before trusting a zero-conversion verdict. The spend version of the same logic: if a term has spent three times your target CPA with no conversion, it has consumed three customers' worth of budget and returned none, which is enough to act. The reasoning to write in the report is simple: below those volumes, one missed conversion swings the read, and the cost of waiting a week is usually smaller than the cost of cutting a term that was about to convert.

Name the failure modes before the client does

The standard advice (cut zero-converting terms, scale winners) breaks in predictable ways, and a report that survives scrutiny calls them out. The biggest trap is cutting on stale data. A term that spent heavily and converted well six months ago but has had no clicks in 45 days is not a 'winner' to defend, it is a term the auction has already stopped serving. Conversely, a term flagged for a cut whose last click was 40 days ago may just be seasonal. Always show the last-click date next to the decision, and treat anything older than about 30 days as a signal that the term has gone quiet rather than a verdict on its quality.

The second trap is attribution lag. Lead-gen and considered purchases convert days or weeks after the click. If you cut a term at the clicks threshold but your sales cycle is two weeks, you may be killing terms before their conversions have landed. Check the conversion delay in your own data (Tools > Conversions shows time-lag reporting) and do not finalize zero-conversion cuts on terms whose recent clicks fall inside that lag window.

The third trap is the budget-capped campaign. If a campaign is limited by budget, its 'Search impression share lost (budget)' is high, which means Google is throttling your delivery and the performance you see is not the performance the campaign is capable of. Reporting a flat cost per conversion without noting the campaign was budget-limited misleads the client into thinking you have hit a ceiling when you have really hit a wallet. Pull the impression share columns and flag any campaign losing meaningful share to budget.

Auditing an account you just inherited

If this report is your first on a newly inherited account, the audit comes before the report. Start with conversion tracking, because every downstream number depends on it. In Tools > Conversions, confirm each action has a recent conversion, a sane counting setting (one per click for leads, every for ecommerce), and no obvious duplicates. An account that reports great numbers on broken or double-counted conversions is the single most common inherited landmine.

Then review the negatives and the match types. Pull the search terms report over the last 60 to 90 days and look for spend on terms unrelated to the business, a sign negatives have not been maintained. Reviewing hundreds of terms by hand is slow, which is why tools exist for this: QueryCut, for example, scans the search terms report and surfaces candidate negatives automatically, so you spend your time deciding rather than reading. Whatever method you use, apply the same clicks and spend thresholds from above so your first round of cuts is defensible from day one.

Finally, document the account's starting state in the report. A clean 'here is what I found' baseline (tracking status, impression share, wasted spend, structure problems) protects you later: when results improve, you can show the before, and when a legacy problem surfaces, you have already noted it was there when you arrived.

Questions people also ask

How do I audit a Google Ads account I inherited?

Start with conversion tracking in Tools > Conversions: confirm each action fires recently, counts correctly, and is not duplicated, because every performance number depends on it. Then pull the last 60 to 90 days of search terms to check whether negatives have been maintained, and review match types and campaign structure. Document the starting state as a baseline so you can show before-and-after later and point to pre-existing problems you did not create.

Why can't I see a full search terms report for Performance Max?

Google does not expose complete query-level data for Performance Max. You get a limited view under the campaign's Insights tab showing search categories and themes rather than raw search terms, and negative keyword control is more restricted than for Search campaigns. Report this limitation openly and show the category themes that are available rather than implying you have Search-level visibility you do not have.

How many clicks should a search term get before I cut it for zero conversions?

Work it from the client's own economics. Divide the target cost per conversion by your average cost per click to get your break-even clicks per conversion, then wait for roughly two to three times that many clicks before trusting a zero-conversion read. Below that volume, a single conversion you simply have not received yet would swing the verdict, so cutting early risks killing a term that was about to convert.

What's the most common mistake in a client report?

Reporting a flat cost per conversion without noting that a campaign was limited by budget. When 'Search impression share lost (budget)' is high, Google is throttling delivery, so the numbers reflect the budget ceiling, not the campaign's real potential. Pull the impression share columns and flag any budget-limited campaign so the client understands the constraint is the wallet, not the market.

Should I report 'Conversions' or 'All conversions'?

Report 'Conversions' as your primary number, because that is the column Smart Bidding optimizes toward and the one tied to your chosen primary actions. Show 'All conversions' separately if secondary actions matter to the client, and name the gap between the two, since a numerate client will notice it and ask. Always list the literal conversion action names you are counting so there is no ambiguity about what a 'conversion' means.

More from the blog

Landing Page Experience: The Changes That Actually Move ItImpression Share Lost to Rank vs Budget: What Each One MeansHow Many Campaigns Is Too Many for the Budget You Have?Primary vs Secondary Conversion Actions: What Each One Actually Changes