The weekly ad report: what to include and what to cut

Most weekly reports fail for the same reason: they describe what happened instead of saying what it means. The client reads a table, understands nothing new, and the report becomes a ritual that costs you two hours and buys no trust.

The six blocks that make a weekly report work

1. The headline. One sentence. "Cost per lead fell 12% after the creative swap; volume held." If someone reads only this, they should know whether the week was good.

2. The three numbers. Spend, results, cost per result, each with the change against the previous equivalent week. Not fifteen metrics. Three.

3. What changed in the account. Every edit you made, with the date. This is the block people skip and the one that makes the rest interpretable: a cost increase after Tuesday's budget change is a different story from one with no cause.

4. What you are watching. Anything trending wrong but not yet actionable. Naming it early means it is never a surprise later.

5. Next week's plan. Two or three specific actions, not intentions. "Three new creatives for the retargeting set" rather than "continue optimising".

6. Anything you need from them. Approvals, assets, information. Put it at the end, in one list, so it is easy to act on.

That is the whole format. It fits on one screen and takes fifteen minutes once the numbers are in front of you.

Who reads it, and when

Most weekly reports are read on a phone, between other things, in under two minutes. That single fact should shape the format more than any template.

It means the conclusion goes first and has to survive being read alone. It means tables need to be short enough to render on a narrow screen. And it means anything that requires zooming in is effectively invisible.

A report built for a laptop and read on a phone gets skimmed, and skimming a report whose conclusion sits at the bottom produces exactly the outcome you were trying to avoid: the client forms an impression from the numbers without the explanation attached to them.

What to cut

• Impressions and reach as headline figures. They rise by spending more.
• Screenshots of the ads manager. They show effort, not insight.
• Every metric the platform offers. Availability is not a reason to include something.
• Explanations of what CTR means. Define it once, in the first report, then stop.
• Long paragraphs describing the table. If the table needs narrating, the table is wrong.

The test for every line: does this change what anyone does next week? If not, cut it.

Who the report is really for

Two readers, with different needs. The person paying wants to know whether the money is working and whether they need to do anything. The person operating wants a record of decisions to look back on.

One document can serve both if the order is right: conclusion first, evidence second, operational detail last. The client reads the top and stops; you read all of it in three months when you are trying to remember why a campaign was restructured in September.

What breaks this is burying the conclusion under the evidence, which is the default instinct of anyone who did the work and wants it seen. The work shows in the quality of the conclusion, not in the length of the appendix.

Why the "what changed" block matters most

Without it, every metric movement looks like weather. With it, the report tells a causal story: you did something, this happened, here is what follows.

It also protects you. When a cost increase follows a deliberate scale-up, the log makes that visible, and the conversation is about a planned trade-off rather than an unexplained problem. Teams that keep this log well spend far less time defending themselves.

How to write the headline sentence

Three patterns cover almost every week:

• Improvement: what improved, by how much, and why you think so.
• Deterioration: what got worse, the diagnosis, and what you are doing about it.
• Stability: what held steady, and why no action was the right call.

That third one is the hardest to send and the most valuable. A report that says "everything held; no changes were needed and here is why" demonstrates judgement. A report that invents activity to look busy eventually produces changes that hurt the account.

The line that prevents most follow-up questions

Every report generates the same handful of questions: why did spend move, is this good, what are you doing about it. Answering them before they are asked takes three sentences and saves a round of email.

The pattern that works: state the number, attribute it to a cause, say what happens next. "Spend rose 18% because we scaled the top campaign on Tuesday; cost per lead is expected to run higher for a few days while delivery recalibrates, and we will review on Friday."

That sentence does three jobs at once. It reports, it explains, and it sets the expectation that prevents an alarmed message on Wednesday.

Comparing periods without fooling anyone

Seven days against the previous seven, matching weekdays. A week with a public holiday gets flagged in the report rather than silently compared. If the account has strong weekday patterns, say so once so the client stops reading Monday dips as problems.

For accounts with low volume, weekly numbers swing on chance. In those cases it is more honest to report a rolling fourteen-day figure and explain the reason than to present a 40% swing as a finding.

Making it take fifteen minutes instead of two hours

The time cost is almost never the writing. It is the assembly: opening each platform, exporting, pasting, reconciling. Any setup where the numbers are already consolidated turns a two-hour job into a fifteen-minute one, and the difference shows up in quality, because the time goes into the interpretation instead of the spreadsheet.

The second saver is a fixed template. Same six blocks, same order, every week. The client learns where to look, and you stop deciding what to include.

For the monthly business-level version, continue with client reporting for small agencies, and for choosing what belongs in it, marketing KPIs. The pricing page shows how many accounts each plan covers.