Most paid media reports are built to be read once and never again. That is why nobody can tell you whether the last six months actually went anywhere.
A recurring paid media report should be organized around the four things your team can actually change, which are landing pages, creative, channels and tracking. Report those four in the same order every month, keep the money numbers in a short header above them, and the document becomes comparable across a year instead of readable for a week. This is written for B2B marketing leaders who send or receive a monthly paid readout and still cannot say what progress looks like.
A 40-metric deck is a reporting failure, not a data failure
The standard monthly report is a screenshot of every platform, in whatever order the tabs happened to open. Impressions, clicks, CTR, cost per click, conversions, cost per conversion. Then the same set again for the second channel. Then a LinkedIn tab nobody scrolls to.
Every number in it is accurate. None of it answers the question the reader is actually asking, which is whether the program is in better shape than it was in June.
The reality is that a month of paid media contains very few real decisions. Maybe three. The report should be shaped around those decisions and the work behind them, and the performance figures should sit at the top as context rather than filling the body of the document.
So the header is four or five lines. Spend, qualified conversions, cost per qualified conversion, pipeline created, and the comparison to last month. Everything after that is the work.
Ready for paid ads that pay off?
Book your free auditThe four pillars, and why it is these four
Each pillar is something your team has hands on. If you cannot change it, it does not get a section.
Landing pages. Which pages paid traffic hit, what changed on them, and what happened to the page conversion rate. Not the site-wide number. The specific pages the money went to.
Creative. Which concept is running, how long it has been live, and what is queued behind it. A month where creative did not change is a month where creative did not change, and that belongs in the report.
Channels. Where the budget sat, what moved, and why it moved. Geography, campaign type and platform all live here as reporting dimensions, which only works if your campaign and UTM naming is consistent enough to group on.
Tracking. Which conversion actions exist, which ones fired this month, which single one bidding is pointed at, and anything that broke. This is the section people cut first and the one that costs the most when it is wrong, which is why monitoring conversion tracking deserves a standing slot rather than an emergency.
Four is not a magic number. It is roughly what a mid-market paid program can hold in its head at once. Add a fifth pillar if you genuinely own a fifth lever, such as a sales-development handoff. Do not add one because a stakeholder asked a question once in March.
Write "no change" instead of deleting the section
This is the part people argue with, so I will be direct about it. Every pillar appears in every report, every month, even when there is nothing to say. You write "no change" and you move on.
Deleting an empty section feels like editing. It is actually hiding the most useful signal the report produces. If tracking reads "no change" three months running, that is not a tidy report. That is a team that has not looked at its conversion actions since spring, and the document just said so out loud.
It works the other way too. If landing pages has an entry every month for five months while creative sat blank the whole time, you have found where your attention goes and where it does not. No performance chart tells you that. Section presence does.
A plain report you can stack against five older versions and read down a column is worth more than a beautiful one that only makes sense in the month it was written.
Your numbers are provisional, so stop freezing them
Google Ads does not report a conversion on the day it happened. It reports it against the day of the click. Google's own conversion lag documentation puts it plainly, noting that "because Google Ads reports conversions by query date, you may not always see the most updated conversion numbers" and that this can make CPA look inflated and ROAS look deflated.
Say you report 42 conversions at $310 apiece on the 3rd of the month. Three weeks later that same closed month shows 51 conversions at $255. Nothing improved. The month just finished arriving.
Google Analytics has its own version of this. Its data freshness page states that "data processing can take 24-48 hours" and that during that window the data in your reports may change.
So, three rules.
1.) Put a data-as-of timestamp on the front page, in the same spot every month.
2.) Restate last month's headline numbers at the top of this month's report. One line. Reported X, settled at Y.
3.) Publish your own conversion lag once and stop arguing about it. Segment the campaigns page by Conversions and then Days to conversion and you have the actual distribution for your account. Google notes that lag affects recent data far more than older data, so a freshly closed month is the least trustworthy number in the whole document.
Nobody minds that numbers move. People mind when numbers move and nobody warned them.
Report what you made worse on purpose
Pointing a campaign at a qualified milestone instead of raw form fills makes the conversion count fall. That is the mechanism working the way it should, not a performance problem.
If the report does not say so before the reader reaches the chart, the reader decides for themselves, and they decide it is failure. Then somebody senior asks why leads dropped 40% and the real answer arrives a week late, in a defensive email.
So the intentional-damage note goes above the numbers it explains, not in an appendix. Something like this. We changed the conversion goal on the 8th, the raw conversion count will fall through October and cost per conversion will roughly double, and the qualified number is the one to watch instead.
Give it a date to be judged on as well. Google documents that it can take "up to around 50 conversion events or 3 conversion cycles" for a bid strategy to calibrate after the objective changes, so a reader demanding a verdict in week two is asking for noise and will get it.
A dead dashboard tile does not look dead
The quietest way to lose a channel is to lose its data connection. When a dashboard loses access to an ad platform, the channel usually does not throw an error. It shows nothing. Rows go absent, the chart still renders, and the team stops discussing a channel that is still spending money every day.
The mechanism is boring. Google's Data Studio, the product recently renamed from Looker Studio, lets a data source run on Owner's Credentials, which means everyone's view of a report is authorized by one person's account. The documentation notes that such sources "continue to retrieve data even if you can no longer edit the data source", and that revoking those credentials stops the source pulling data as you. When that one person changes roles, the tile goes quiet rather than red.
Two fixes, both cheap.
→ Add a liveness line per source on the front page. Rows returned and last data date, per channel. A zero is visible in a way that a missing chart never is.
→ Name a human owner for every connection, and make it someone who is not the person most likely to change jobs this year.
While you are in there, pull change history. Google keeps two years of it, shows the email address of whoever made each change, and labels anything done through the API with a tool name, which is how you discover that an automated system moved a budget at 4am. Google also admits that not every account-level change appears there, so treat it as strong evidence rather than a complete log.
A sudden drop in a search program traces back to a settings or targeting change far more often than to the market. Change history is the first artifact to produce when somebody asks what happened, not the fourth.
Cut the status call, keep the document
Most recurring paid media calls are people reading a report out loud to people who could have read it themselves. That is an expensive way to distribute a PDF.
My view is that a monthly written readout plus a live dashboard replaces the weekly status call outright, and the time that frees becomes a working session every other week where decisions actually get made. Accounts with layered internal approvals feel this hardest, because the constant re-explaining eats hours that never reach a campaign.
Keep the written update short. What changed, what we learned, what we are doing next, what we need from you. If it takes longer than fifteen minutes to write, the month had too many half-finished things in it.
Build it in one afternoon
1.) Write the header. Spend, qualified conversions, cost per qualified conversion, pipeline created, and the restatement of last month.
2.) Create the four pillar headings and leave them empty.
3.) Add the data-as-of stamp and a liveness line for every channel.
4.) Fill the pillars with this month only. No history, no padding, and "no change" wherever that is the truth.
5.) Put the intentional-damage note above the chart it explains, with a date for the verdict.
6.) Send it as a document rather than a deck, and keep every version in one folder so the sixth can be read against the first.
7.) At the quarter, read each pillar down the column and find the one that has been blank the whole time. That is your next project.
Where this is the wrong report
Two cases, and both are common enough to name.
If you are three weeks into a first campaign, a pillar report is overkill. Report spend, conversions and what you learned, in four lines, weekly, until there is enough history for a comparison to mean anything.
If your reader is a board rather than an operator, they want channel contribution to pipeline and very little else. Reporting contribution by channel on a fixed cadence is what makes it politically possible to cut a program later, and that job deserves its own single page rather than a section inside the operating readout.
The one change worth making this month
Take last month's report, add the four pillar headings, and fill them in from memory. Where you end up writing "no change," leave it there and let it sit in the record.
Then open the version from six months ago and try to answer one question. What is better now than it was then? If the answer is not in either document, the reporting was the problem all along, not the account.
That is usually the first thing we rebuild when we take over a B2B Google Ads program, before touching a single bid.
