The Google Ads Scorecard
This scorecard tracks the numbers that decide whether a Google Ads account is making money or leaking it across Search, Performance Max, and Shopping. It is run by the media buyer and reviewed weekly with the marketing lead and the owner, with live data pulled straight from the account so nobody is reading last month's export.
Why this scorecard exists
Why a generic scorecard fails in Google Ads
Most Google Ads reporting shows spend, clicks, conversions, and cost per conversion, then calls it a day. Those four numbers can all look fine while the account is quietly buying your own brand name at a premium, counting form spam as leads, and letting Performance Max spend half its budget on queries Search was already winning. A generic scorecard cannot see any of that. It measures the output of the auction without measuring the health of the machine that bids in it.
Google Ads is different from other paid channels because so much of the result is decided by structure, not creative. Whether brand and non-brand are split, whether conversion actions are set to primary or secondary, whether the product feed is approved, whether campaign names follow a pattern anyone can filter on. These are operational choices, and they move cost per qualified lead more than a new headline ever will. So the scorecard has to carry rows for tracking health, search term waste, pacing, and naming compliance right next to the performance numbers, with a named owner on each.
When this scorecard is live, the weekly review stops being a debate about whether the numbers are real and becomes a decision about what to change. The media buyer knows which row triggered a search term review, the marketing lead knows which conversion action is being audited, and the owner knows whether the account is on pace for the month before it is over. Every red flag has a rule, so action does not depend on who happened to be looking at the dashboard.
The scorecard
11 numbers. Every one has an owner.
| Metric | What it means | How to set the target | Owner | Cadence | Red flag |
|---|---|---|---|---|---|
| Brand search impression share Brand traffic is the cheapest, highest-intent traffic in the account, and losing it to a competitor or a reseller is the most expensive silent leak there is. | The share of searches for your brand name where your ad showed, measured on brand campaigns only. | Pull the last 8 weeks of brand impression share and set the floor at the median. If the floor is well below where you would expect for your own name, raise it 5 points a quarter until the lost share is only the queries you choose not to buy. | Media buyer | Weekly | Any week below the floor triggers a same-day check of budget caps, ad approvals, and competitor bidding on brand terms. |
| Non-brand impression share lost to budget This row tells you whether the growth ceiling is money or quality, and those two problems have completely different fixes. | The share of eligible non-brand searches your ad missed because the campaign ran out of daily budget. | Look at the trailing 8 weeks per non-brand campaign and note where lost-to-budget sits alongside cost per qualified conversion. Campaigns that are both under target on cost and losing share to budget are the ones that earn more spend, so the target is to drive lost-to-budget down on those specific campaigns while holding cost steady. | Media buyer | Weekly | Lost-to-budget rising two consecutive weeks on a campaign that is beating its cost target triggers a budget reallocation proposal for the owner to approve. |
| Search term waste rate Broad match and Performance Max expand reach by design, and without a weekly check they will spend on queries that have nothing to do with what you sell. | The share of Search and Shopping spend that went to search terms you would not bid on if you saw them one at a time. | Tag every search term in the last 4 weeks as keep, block, or unsure, then calculate the share of spend on block. That is your starting waste rate. Set the target at half of it and rerun the exercise every quarter, because match types drift and new junk terms show up constantly. | Media buyer | Weekly | Two consecutive weeks above target starts a full search term review with new negatives added to the shared negative list before the next review. |
| Conversion tracking health Every automated bid strategy in the account is optimizing toward whatever the conversion actions say, so a broken or bloated tag quietly retrains the machine to chase the wrong thing. | Whether every primary conversion action fired within the last 24 hours at a volume consistent with its trailing average, and whether no unintended action has been promoted to primary. | The target is a clean daily pass: every primary action reports, none reports at more than double or less than half its 28-day daily average, and the primary and secondary settings match the documented list. Any deviation is a fail, not a warning. | Marketing ops lead | Daily | A single failed day pauses any budget increase and starts a tag audit in Tag Manager and the conversion settings page before bids are touched. |
| Cost per qualified conversion Cost per conversion as Google reports it counts spam, duplicates, and tire kickers, and a scorecard built on that number rewards whoever produces the most noise. | Total Google Ads spend divided by conversions that passed your qualification rule, such as a real phone call over a set duration, a form from a real company, or an order over a minimum value. | Take the trailing 8-week median cost per qualified conversion, split by Search brand, Search non-brand, Performance Max, and Shopping, and set each target at its own median. Tighten each one by 10 percent a quarter, and never let a blended number hide a segment that is off. | Media buyer | Weekly | Any segment more than 20 percent over its target for two consecutive weeks moves budget out of that segment until the buyer presents a fix. |
| Qualified conversion rate This row catches lead quality collapsing while volume looks great, which is exactly how a broad match or Performance Max expansion goes wrong. | The share of conversions Google counted that also passed your qualification rule. | Calculate the trailing 8-week rate from your CRM or call log, not from the ads platform. Set the floor at the median and treat a drop below it as a targeting problem first and a landing page problem second. | Marketing lead | Weekly | A drop below the floor in any single week triggers a review of the 20 highest-spend search terms and placements that produced the unqualified leads. |
| Budget pacing variance Google will happily overspend a daily budget by design, and a campaign that runs out on the 22nd or leaves a third of the month unspent is a planning failure no other row will catch. | The difference between where month-to-date spend sits today and where it should sit if the month were spent evenly, expressed as a percentage. | Set the acceptable band from your own history. If the last 6 months show you typically finish within a few percent of plan, use that spread as the band. If you have never finished on plan, start with a wide band and tighten it each month as the buyer gets better at daily adjustments. | Media buyer | Daily | Variance outside the band for three consecutive days triggers a daily budget adjustment and a note to the owner on whether the monthly plan still holds. |
| Performance Max brand overlap Performance Max reports beautiful returns when it is quietly buying your brand traffic, and the only way to see it is to check the search categories and insights reports on purpose. | The share of Performance Max spend that landed on brand queries or on the same products and searches your Search campaigns already win. | Pull the Performance Max search categories report for the last 4 weeks and total the spend on brand categories. That is your starting overlap. Add brand exclusions, then set the target at a fraction of the starting number and hold it there, because the overlap creeps back every time the campaign expands. | Media buyer | Weekly | Overlap above target for two consecutive weeks triggers a brand exclusion update and a side-by-side comparison of Performance Max and Search cost per qualified conversion with brand removed. |
| Shopping feed health A Shopping or Performance Max campaign cannot sell a product Google refuses to show, and feed problems tend to appear on a Friday and sit unnoticed all weekend. | The share of products in Merchant Center that are disapproved, pending, or limited by policy, plus the count of products missing a required attribute. | The target is zero disapprovals on products you actually want to sell. Pull a 30-day baseline of disapprovals and set an interim ceiling at half the baseline while the ecommerce lead works down the backlog, then hold at zero. | Ecommerce lead | Daily | Any disapproval on a top-selling product, or feed-wide issues touching more than a small share of the catalog, gets fixed the same day before any other Shopping work. |
| Offline conversion import freshness If bid strategies only see raw form fills and never learn which ones became customers, the account will keep optimizing toward volume no matter what the scorecard says. | How many hours have passed since the last successful import of CRM outcomes, such as qualified leads or closed deals, back into Google Ads. | Set the target at whatever your import schedule promises, whether that is daily or weekly, and measure the gap against the promise rather than against a wish. If imports are manual today, the first target is to make them scheduled. | Marketing ops lead | Weekly | An import more than one full cycle late triggers a check of the click ID capture on the forms and the CRM export before the next bid strategy change. |
| Naming compliance Every other row on this scorecard is a filter on names, and one campaign called "Test 2" breaks the brand versus non-brand split for the whole account. | The share of campaigns, ad groups, and conversion actions whose names follow your documented pattern, such as channel, campaign type, brand or non-brand, and market. | The target is every active object compliant. Audit the account once, rename or archive the outliers, and then hold at zero exceptions. New campaigns do not launch until the name matches the pattern. | Marketing ops lead | Monthly | Any non-compliant active object found in the monthly audit gets renamed within the week and the person who created it walks through the naming doc with the ops lead. |
Targets are set from your own history, never from someone else's benchmarks. The builder does the arithmetic once you plug in your numbers.
The weekly review
Twenty minutes, same order, every week.
// 01
Cost per qualified conversion by segment
The media buyer opens with the four segments: Search brand, Search non-brand, Performance Max, and Shopping, each against its own target. Blended numbers are not allowed on screen. This is the number that decides whether the account earned its budget this week.
// 02
Tracking, feed, and pacing status
The marketing ops lead and ecommerce lead report the daily rows as pass or fail for the week: conversion tracking health, feed health, and pacing variance. If any of them failed, that becomes the first decision of the meeting, because nothing else on the scorecard can be trusted until they are green.
// 03
Waste, overlap, and impression share
The media buyer walks the search term waste rate, Performance Max brand overlap, and both impression share rows, and reads out which red-flag rules tripped. The marketing lead speaks to qualified conversion rate and what the sales team is saying about lead quality. Nobody explains the numbers. They state them and name the rule.
// 04
What changes this week
The owner and marketing lead decide on budget moves, approve or reject any reallocation proposal, and assign every tripped red flag to a named person with a due date. The list of decisions gets logged next to the scorecard so next week's review starts by checking whether they happened.
Where scorecards die
Mistakes we see constantly.
Reading blended cost per conversion
One number across brand, non-brand, Performance Max, and Shopping hides the fact that brand is cheap and everything else is expensive. Split it on the scorecard or the account will drift toward buying its own name.
Trusting the platform's conversion count
Google counts every form fill, duplicate, and spam submission as a conversion and then bids toward more of them. The scorecard has to be built on qualified conversions from your CRM or call log, or every row underneath it is fiction.
Letting Performance Max grade its own homework
Performance Max looks like the best campaign in the account right up until you check how much of its spend was brand queries. Add the brand overlap row and the exclusions, or you will keep funding a campaign that is mostly recycling Search wins.
Checking pacing at month end
A pacing problem found on the 28th cannot be fixed. Pacing variance is a daily row with a daily owner, and the only useful time to see it is when there is still budget to move.
Skipping naming and structure
Teams treat naming compliance as housekeeping and drop it from the scorecard first. Then a campaign gets named wrong, the brand split breaks, and the scorecard silently starts lying. Structure is what makes every other row filterable.
Other scorecards
Same discipline, different numbers.
By platform
By business type
Want it set up live in your dashboards?
We install scorecards on real data, wire the owners and alerts, and coach the weekly review. Scorecards and accountability is the service.
Questions we get
Straight answers.
Do we need separate scorecards for Search, Performance Max, and Shopping?
No. Run one Google Ads scorecard and segment the rows that need it, mainly cost per qualified conversion and impression share. Splitting into three documents just means three meetings and no view of where budget should move between them.
How do we define a qualified conversion?
Write the rule down before you touch the scorecard. For lead generation it is usually a call over a set duration or a form from a real prospect confirmed in the CRM. For ecommerce it is an order over a minimum value with no refund inside a set window. The rule matters less than the fact that everyone uses the same one.
Who should own the scorecard if we do not have a full-time media buyer?
The rows still need owners, and the owner can be the person who has access, even part time. What does not work is leaving the ops rows like tracking health and naming compliance unowned because nobody thinks of them as marketing. Those are the rows that break first.
What if our Google Ads account is managed by an outside partner?
Then the scorecard is how you manage the partner. They report the rows, you own the targets and the red-flag rules, and the account, the data, and the conversion setup stay in your name. If a partner cannot produce a search term waste rate or a brand overlap number, that tells you something.
How long before the scorecard changes results?
The tracking, waste, and overlap rows usually surface problems in the first review because most accounts have never looked. Fixing them changes cost per qualified conversion within a few weeks of budget cycles, since automated bidding needs time to relearn. The bigger change is that the weekly meeting stops arguing about whether the numbers are real.
A scorecard is only as good as the system feeding it.
Start with a diagnostic call. We will tell you which of these numbers you can trust today and what it takes to trust the rest.