# How to Write Ads Report Commentary Clients Read

> Learn how to write ads report commentary clients read: a four sentence structure, honest cause and effect, and phrases to cut from every Meta Ads report.

Author: The DashOps Team  |  Published: 2026-09-25  |  Updated: 2026-10-01  |  Topic: Agency & Client Work

You have the report built. The charts are right, the numbers reconcile, the file is sitting in your downloads folder. Then you reach the empty comment box under the spend chart and stall, because writing "spend was up this month" is useless and writing anything more feels like guessing. So you leave it blank, or you paste a description of a chart the client can already see. Commentary is the part of a client report that decides whether the client trusts you, and it is the part most people write last, fastest and worst.

This post is about the writing itself. If you are still deciding what belongs in the report, start with the [Facebook Ads client report template](/resources/facebook-ads-client-report-template) and come back here for the words.

![1790349085787-32a80af67abf-httpsdashopsioresourceshow-to-write-ads-report-com](https://pub-ae7c0247bc134935855f21b665422e27.r2.dev/posts/1790860876043-793b442606cf.jpg)

## Commentary answers the question a chart cannot

A chart shows what happened. It cannot show why, and it cannot show what you intend to do about it. Those two things are why the client pays you rather than logging into Ads Manager themselves. If your commentary only restates the chart, you have written a caption, and captions make clients wonder what they are paying for.

Every comment worth keeping does at least one of these:

- Explains the cause of a movement the client can already see
- Flags a movement the client cannot see because it is buried a level down, in one ad set or one creative
- Connects an ad metric to something the client cares about: booked calls, foot traffic, stock they need to clear
- States a decision you have already made and acted on
- Asks for a decision only the client can make: more budget, a new offer, fresh creative
- Sets an expectation for the next period so the next report holds no surprises

If a sentence does none of those, delete it. Nothing is lost.

## The four sentences that make a comment complete

The fastest reliable structure is four sentences, in this order.

**What happened.** One sentence, direction first, no hedging. "Cost per lead rose this week."

**Why it happened.** One sentence naming the place in the account where the change lives. Not "market conditions". Something you can point at: an ad set, a creative, a date you paused something, a landing page that went down, a holiday weekend.

**What it means.** One sentence translating the movement into the client's language. Is this a problem, a normal fluctuation, or the deliberate result of something you chose to do?

**What happens next.** One sentence with an action and a timeframe, or a question for the client.

Four sentences will not fit every situation, and you should break the pattern when it deserves breaking. But when you are staring at a blank box, this gets you unstuck, and it reads as though a person wrote it rather than a dashboard.

## Lead with the decision, not the data

Most people write commentary in the order they discovered things: opened the account, noticed this, checked that, found the cause, changed something. That is the order of your investigation, and it is the wrong order for the reader.

Clients read the first line, then decide whether to read the rest. Put the consequence first and the evidence behind it. "We have shifted budget out of the broad ad set and into retargeting" is a first line. "Broad audience cost per acquisition drifted upward across the month" is a fourth line.

This is the principle behind the [executive summary at the top of the report](/resources/ads-report-executive-summary), applied to the individual paragraph.

## Write the why honestly, including when you are not sure

The reason people avoid real commentary is fear of being wrong about the cause. Attribution is messy, seasonality is real, and Meta's delivery system makes decisions you did not make.

The answer is not vagueness. It is calibrated language that tells the client how confident you are. There is a large difference between these three sentences, and clients understand all three:

- "Cost per lead rose because we paused the prospecting campaign for the bank holiday." That is a known cause.
- "The most likely cause is creative fatigue in the ad set running longest, since frequency there has climbed since the last refresh." That is a strong inference, and you have said why you believe it.
- "We do not yet know the cause. Nothing changed in the account, so we are checking the landing page and the pixel before we touch budget." That is honest, and far better than a confident guess that turns out to be wrong.

The third is the sentence people most fear writing and the one that builds the most trust over a year. What clients cannot forgive is a confident explanation that later collapses.

## Rules for quoting numbers inside a sentence

Commentary sits next to the data, so it does not need to repeat it. A comment stuffed with figures is harder to read than the table it describes.

- Quote a number only when the sentence stops working without it.
- Round to whatever precision matters to the decision. Nobody makes a budget call on the second decimal place.
- Always attach a comparison. A number alone means nothing. Compare against the previous period, the target you agreed, or the same period last year.
- Name the unit and the scope. "Cost per lead across the whole account" and "cost per lead on the retargeting ad set" are different claims, and clients assume the first unless you say otherwise.
- Never invent a benchmark. With no source for an industry average, keep it out of the report. Compare the account against its own history, which is more relevant anyway.

## Phrases to delete on sight

These appear in almost every report and carry no information:

- "We will continue to monitor." Monitoring is not an action. Say what you will change, or say you are leaving it alone and why.
- "Performance was mixed." Every period is mixed. Name the two things pulling in opposite directions.
- "Due to algorithm changes." Unless you can point to something Meta announced, this reads as an excuse.
- "As you can see from the chart above." If they can see it, you do not need to write it.
- "Optimisations were made." Which ones, where, and what happened afterwards.
- "Great results this month." Let the numbers say that. Your job is to explain them.

## Match the depth to who is reading

A founder checking the report on a phone needs different commentary from a marketing manager who will forward it to their own boss. Write for whoever actually opens the file, and put deeper analysis further down. It also helps to take pressure off the document: a read-only client login in DashOps, or a share link to a dashboard snapshot, lets the client check live numbers whenever they like, so the commentary can carry judgement rather than every figure. There is more on this in [how much detail belongs in a client ads report](/resources/how-much-detail-in-a-client-ads-report).

## A worked example

Here is a weak comment for a fictional client, Northwind Solar:

"Spend increased this week and cost per lead was higher than last week. We will continue to monitor and optimise."

Here is the same week, written properly:

"Cost per lead rose this week, and almost all of the increase sits in retargeting rather than prospecting. Frequency there has been climbing since we paused the prospecting campaign over the holiday, so the same people kept seeing the ads. Prospecting is still bringing in leads at the cost we agreed at the start of the quarter, so the account level figure looks worse than the underlying performance. Prospecting is back on as of Monday, and from next week we will show the two cost per lead figures separately so you can watch them move independently."

Same data, same week, same work in the account. Only the second version gives the client something to act on.

## Where the time comes from

The obstacle is not skill, it is that assembling the report eats the hour you needed for thinking. Pulling numbers by hand, rebuilding the same tables and reconciling figures across ad accounts leaves nothing for the part that carries the value.

This is where tooling earns its place. DashOps builds the Meta Ads reports and the recurring email digests for you, so the numbers land on schedule, and drilling from campaign to ad set to ad takes seconds when you are hunting the cause of a movement. The point is not the export. It is that the hour you get back goes into the four sentences the client will actually read. There are more ways to claw that time back in [reducing time spent on Facebook Ads reporting](/resources/reduce-time-spent-on-facebook-ads-reporting).

## What to do next

Open your most recent client report and read only the commentary, ignoring every chart. Ask whether someone who never saw them would learn anything from those words alone. If the answer is no, rewrite one section using the four sentence structure: what happened, why, what it means, what happens next. Do that for one client this week, keep the phrasing you like as a reusable pattern, and apply it to the rest next month.

## Frequently asked questions

### How long should ads report commentary be?

Short enough that the client reads all of it. A few sentences per section, and a paragraph for the summary at the front, is usually the ceiling. Length is not the signal of effort your client is looking for. If a comment runs long, it is normally because it is describing the chart rather than explaining it. Cut the description, keep the cause and the decision, and the paragraph shrinks on its own.

### Should I write commentary for a metric that did not move?

Only if the client expected it to move, or if stability is itself the news. A metric that held steady while spend rose is worth a sentence, because holding steady under more pressure is a real result. A metric that sat flat in a quiet week is not. Silence on an unchanged number is fine. Writing something for the sake of filling the box trains the client to skim past everything you write.

### What do I write when the results were bad?

Say it plainly in the first sentence, then explain what caused it and what you have already changed. Clients forgive a bad period far more easily than they forgive finding out about it a month late, or reading commentary that buries it under paragraphs of unrelated wins. State the problem, separate the part you control from the part you do not, and give a specific next step with a date attached so the client knows when to expect an update.

### Can I reuse commentary wording between clients?

You can reuse the structure and the phrasing patterns, and you should, because that is what makes reporting fast. You cannot reuse the substance. Any sentence that would be equally true for two different accounts is a sentence worth deleting. Build a short library of openers and transitions, then fill the cause and the decision fresh from that account's own data every time.

### Who on the team should write the commentary?

Whoever actually made the decisions inside the account that period. Commentary written by someone who was not in the ad account tends to describe charts, because a chart is all they have. If an account manager writes the client facing version, have the person running the media leave short internal notes on what changed and why, then translate those notes into client language rather than inventing an explanation.

---

Full page: https://dashops.io/resources/how-to-write-ads-report-commentary