Impression share at 92% tells you the campaign is winning almost every auction it enters. It says nothing about whether the clicks it bought were worth buying. That gap is the whole problem with default search campaign reporting metrics: the Google Ads interface puts account-health signals in the first columns you see, so they end up in the first tiles the client sees, and the client reads a page of green numbers that never answers the question they actually have.
The question is usually some version of “should I keep spending this much on this?” Impression share, Quality Score and CTR can support an answer to that. They cannot be the answer. And when they sit at the top of the report, they crowd out the numbers that can.
The split between the two groups is cleaner than it looks, and it has nothing to do with how sophisticated the metric is.
Two sets of search campaign reporting metrics, and who each set is for
A health metric answers a question about the campaign’s mechanics. Is the budget capping delivery? Are the ads relevant enough to the queries to win cheap auctions? Is the ad copy pulling clicks at a competitive rate? These are the questions a PPC specialist asks on a Tuesday afternoon with the search terms report open. Impression share, search lost IS (budget), search lost IS (rank), Quality Score and its components, CTR by ad group: all health.
A client metric answers a question about the money. What did we spend, what did we get, what did each thing we got cost, and is that direction improving or deteriorating? Spend, conversions, cost per conversion, conversion rate, revenue and ROAS where ecommerce tracking exists, qualified leads where the client can send that back.
The honest reason health metrics end up on page one is that they are always available and always complete. Nobody has to chase the client for them. Cost per qualified lead requires the client’s CRM to come back to you, usually late, often incomplete. Quality Score is sitting right there in the column picker. The report is due today. So the available metric wins over the relevant one, which is a reporting-process failure dressed up as a measurement choice.
One more thing about Quality Score specifically. Putting it in a client report is a choice you almost always regret. It is a 1 to 10 diagnostic scored relative to other advertisers on the same keywords, it moves for reasons the client cannot influence, and once it is on the report the client starts treating it as a grade. An average of 8 out of 10 reads like a B+ to someone who has never seen the metric before. Then it drops to 6 after a landing page redesign the client’s own team shipped, and you spend fifteen minutes of the review call explaining a number that was never evidence of anything they bought.
What the report is actually being asked to prove
Three things, in roughly this order.
That the money went where the plan said it would. Spend against budget, by campaign, with pacing visible. If the client approved €12,000 across branded, non-branded and competitor campaigns, the report should show that split without the client doing arithmetic.
That the spend produced outcomes at an acceptable cost. Conversions and cost per conversion, trended over enough periods to show direction rather than a single month’s noise. One month is a data point. Four months is an argument.
That you know why the numbers moved. This is the part that justifies the retainer, and it is where health metrics come back in as supporting evidence rather than headline figures.
Anything on the report that does not serve one of those three has to earn its space. That is a stricter test than it sounds, and it is the same test worth applying to any client dashboard before you add the fourteenth tile.
When CPC falls and cost per lead rises
Here is the scenario that makes the case better than any principle.
Suppose average CPC drops from €2.40 to €1.95 over a month. Clicks rise from 5,000 to 6,100. Spend is flat at roughly €12,000. On a report built from platform defaults, that is three pieces of good news: cheaper clicks, more traffic, steady budget. CTR probably improved too, since the campaign is winning more clicks per impression.
Then cost per lead goes from €96 to €134.
The metric that explains it is conversion rate, and specifically conversion rate segmented by match type or by campaign. Cheaper clicks are frequently worse clicks. Broader queries, looser intent, more informational searches that were never going to convert. The campaign got more efficient at buying traffic and less efficient at buying outcomes, which is exactly the trade-off a client needs to understand before they read “CPC down 19%” as a win and ask you to do more of it.
A search report that shows CPC without conversion rate next to it is structurally incapable of telling that story. So pair them. Every efficiency metric on the report should sit beside the outcome metric it affects: CPC next to conversion rate, clicks next to conversions, spend next to cost per conversion. The client does not need to be told which direction is good. They need the two numbers close enough together to see the relationship themselves.
Where platform metrics still earn their place
Demoting health metrics is not the same as deleting them. They belong in a diagnostics section, below the outcome section, where they do a specific job: explaining movement.
Search lost impression share to budget is the single most useful one for a client to see, because it is a direct spending recommendation with a number attached. If the campaign is losing 34% of available impressions to budget on a campaign that converts at an acceptable cost, that is the clearest possible case for an increase, and it is the one health metric worth promoting when it is relevant.
Search lost impression share to rank earns its place when you are explaining why performance dipped without a budget change. Competitive pressure is real, and it is better to show it than to let the client assume the campaign got worse on its own.
CTR earns its place when you changed the ad copy and want to show the effect, or when it diverges sharply between branded and non-branded and you are using it to make a point about intent. CTR as a standing top-line tile every month, with no comparison and no decision attached, is filler.
Quality Score earns its place roughly never in a client-facing report. Keep it in your own account review. If a client has specifically asked for a quality score client report because a previous agency trained them to watch it, give it to them in the diagnostics panel with a one-line definition beside it, and spend the review call moving their attention to cost per conversion.
Branded and non-branded need separate lines
If there is one change that improves a Google Ads client report more than any metric swap, it is splitting branded from non-branded.
Blended search numbers hide the thing the client most wants to know. Branded search converts at a high rate and a low cost because the person already decided. Non-branded is where the agency is actually generating demand, and it costs more per outcome by nature. Report them together and you get a flattering blended cost per lead that falls whenever brand awareness rises for reasons unrelated to your work, and that spikes whenever you push into new non-branded territory, which is the thing the client hired you to do.
Split them and the conversation changes. Branded becomes a demand-capture line with a cost-of-defence story attached. Non-branded becomes a growth line with its own cost per conversion, judged against its own trend. The client can see that a rising blended cost per lead is the result of a deliberate shift in mix, not deteriorating performance. That distinction has saved accounts.
Competitor campaigns, if you run them, get a third line for the same reason.
A structure that survives “so is this working”
The order I use for search, top to bottom:
- Spend and pacing. Total spend, budget, percentage through the period, split by campaign group (branded, non-branded, competitor).
- Outcomes and cost per outcome. Conversions, cost per conversion, conversion rate, trended across at least three periods. Split by the same campaign groups.
- Traffic efficiency, paired. Clicks, CPC, CTR, each sitting beside the outcome metric it explains.
- Diagnostics. Lost impression share to budget and rank, top query themes, notable Quality Score movement if it explains something. Short, and clearly labelled as explanation rather than result.
- What changed and what is next. Written, specific, tied to the numbers above.
Section order matters more than tile design here, which is a point worth more than one article’s worth of attention: where a section sits determines whether it gets read. A diagnostics panel at the bottom gets read by the one person on the client side who cares. The same panel at the top gets read by everyone and anchors the whole conversation on account mechanics.
The reason PPC client reporting keeps drifting back to platform defaults is that the layout gets re-decided every month in a spreadsheet or a deck, and under deadline the path of least resistance is whatever the export put in column A. Fixing that is a template problem, not a discipline problem. Decide the search layout once, with cost per conversion and the outcome trend on the primary tiles and the diagnostics mapped to a secondary panel, then reuse it for every search client every month. In Screebl that means building the template in the drag-and-drop builder once, picking the group-by and metrics per row, and generating each month’s report from the same upload motion instead of rebuilding the hierarchy under pressure. Different channels need different shapes, which is why one dashboard rarely survives contact with a multi-channel account, and search deserves its own.
If you want to see how the template and sharing side works before committing your search clients to it, the plans and what each includes are the fastest way to check whether the shape fits your account mix.
