Use the account change log as the spine of the client update. For each meaningful change, state what changed, why the buyer made the decision, the relevant before and after number, and whether the result helped, hurt or remains unclear.
That is the cleanest way to show clients the work PPC requires without burying them in platform activity. The client gets a record of the decisions behind the account, while the buyer gets a practical way to explain work that might otherwise be invisible.
Turn the change log into a decision log
A platform change log records activity. It tells you that a budget moved, an ad changed or a campaign paused. It rarely explains the full decision.
The client update should add that missing context. Treat each meaningful change as a short decision record with five fields:
- What changed?
- Why did it change?
- What number triggered the decision?
- What happened after the change?
- What happens next?
This turns a technical record into something a client can follow. It also keeps the update grounded in account activity rather than general commentary.
Start with one working list
Pull the change history for the reporting period from every platform you manage. Google Ads calls this Change history. Meta Ads Manager provides Activity history.
Do not paste both histories directly into the client report. Put the entries into one working list first. Normalize the dates, account names, campaign names and change types so the buyer can review them in one place.
The working list might use these columns:
| Date | Platform | Campaign | Change | Reason | Before | After | Status |
|---|---|---|---|---|---|---|---|
| Example: Monday | Nonbrand Search | Added negative keyword | Query was unrelated to the offer | Irrelevant clicks present | Query excluded | Monitoring | |
| Example: Tuesday | Meta | Retargeting | Replaced creative | Cost was rising as frequency increased | Previous asset active | New asset active | Too early to call |
| Example: Thursday | Brand Search | Restored campaign | Billing issue was resolved | Campaign paused | Campaign delivering | Confirmed live |
These are illustrative entries, not client results. Their purpose is to show the level of detail needed.
Separate activity from judgment
Not every recorded action deserves its own line in the update. A corrected label or naming cleanup may matter operationally without changing delivery. Several small maintenance actions can usually become one sentence.
The changes that belong in the main update are the ones that affected money, reach, targeting, message, delivery or the way results should be interpreted. That is where the buyer exercised judgment, and that is what the client needs to understand.
A useful comparison is:
| Raw change log | Client decision log |
|---|---|
| Budget changed | Budget moved because delivery was behind plan |
| Keyword added | Search term showed relevant intent and was added for direct control |
| Ad paused | The ad was paused after performance weakened |
| Creative uploaded | A staged asset replaced one showing fatigue signals |
| Campaign enabled | Delivery resumed after the issue blocking it was resolved |
The platform supplies the event. The buyer supplies the reason, the relevant number and the next decision.
Choose the changes that belong in the update
A client update should be selective, but it should not be vague. The test is simple: would this change help the client understand where money went, what the buyer learned or why the account now behaves differently?
Six kinds of change usually pass that test.
Budget changes
Record meaningful increases, decreases and reallocations. Include the reason and any limit that shaped the move.
Example: say a campaign has a $50 target CPA. Its budget was reduced after CPA moved above that target and stayed there long enough to justify action. The update should identify the target, the observed direction and the size of the budget change.
Avoid writing only, “Adjusted budgets.” That proves activity but explains no judgment.
Negative keywords and targeting exclusions
List exclusions that materially changed who could see or trigger the ads. Group related negatives when listing each one would make the update harder to read.
A useful entry says what kind of traffic was excluded and why it did not match the offer. The client does not need a long export unless the detailed list helps with a business decision.
Copy tests
State what the test changed. It might be the offer, headline, proof point or call to action.
Then explain what stayed fixed. A test is easier to trust when the client can see that the buyer changed one meaningful variable and gave both versions a fair reading. For a deeper testing workflow, see how to run a clean Google Ads test on ad copy.
Creative swaps
Explain which asset was replaced and what prompted the swap. Useful signals can include frequency rising, CPM or CPC increasing, or CPA getting worse.
Do not call a creative winner or loser before the evidence supports it. If the replacement has only just started running, mark it as monitoring.
Pauses and un-pauses
A pause is a decision about where not to spend. An un-pause is a decision that the account is ready to resume delivery.
Include both. State what triggered the action, whether it was planned, and what was checked before delivery resumed. This is especially useful when an expired card, rejected ad or other account issue interrupted delivery.
Changes that did not help
These belong in the update too. A complete record includes changes that were reversed, tests that were inconclusive and ideas that did not improve the target metric.
Leaving them out creates a distorted story. Including them shows the actual management process.
Attach the reason and the right numbers
When you show clients the work PPC buyers perform, the reason matters as much as the action. “Changed budget” is a task. “Reduced budget because CPA moved above the campaign target” is a decision.
The numbers should support that decision without turning the section into a second performance report.
Use the smallest useful number set
For each change, include:
- The target or constraint that mattered
- The number before the change
- The number after the change, when enough data exists
- The evaluation window
- The current status
Do not attach every available metric. Choose the metric that caused the action and the one that determines whether it worked.
For a CPA decision, the useful fields might be target CPA, CPA before the change and CPA after it. For a creative swap, they might be frequency, CPM or CPC, and CPA. For a delivery issue, they might be planned spend, actual spend and the time delivery resumed.
Keep before and after comparisons honest
A before and after number is useful only when the periods are reasonably comparable. Note any condition that changes the interpretation, such as a new offer, a holiday, a large budget shift or a landing page change.
Use one of four status labels:
| Status | Meaning |
|---|---|
| Confirmed | The change landed and the account reflects it |
| Monitoring | The change is live, but the result is not ready |
| Inconclusive | There is not enough evidence to make a call |
| Reversed | The change did not help or created another problem, so it was rolled back |
“Inconclusive” is a valid result. It tells the client that the buyer is waiting for enough information rather than forcing a clean story onto weak evidence.
Use a compact writing pattern
Each line can follow this structure:
Changed [setting] from [before] to [after] because [reason tied to target or constraint]. After [evaluation period], [result or current status]. Next, [decision].
Example:
Increased the campaign budget from $100 to $120 because CPA was below the illustrative $50 target and delivery had room to grow. After the first review period, the change is still being monitored. Next, we will keep the current cap until the result is clear.
The numbers are deliberately round and illustrative. A real update should use the account’s actual numbers.
Report the misses and make the process repeatable
The strongest update is not the one in which every change worked. It is the one in which every meaningful change has an honest status.
If a change made the target metric worse, say so. Record whether it was reversed, modified or left running for a stated reason. If the evidence is mixed, explain what is missing.
A practical entry might read:
Example: We tested new copy against the existing version. The new version did not improve the target metric during the evaluation period, so we ended the test and kept the original. We will retain the result when planning the next copy test.
That sentence gives the client the action, outcome and next decision. It does not pretend that activity equals improvement.
Use the same review sequence each time
A repeatable process keeps the update from becoming a scramble:
- Pull platform change histories for the reporting period.
- Remove housekeeping that did not affect delivery or interpretation.
- Group related actions into one decision.
- Add the reason from the buyer who made the change.
- Attach the target, before number and available after number.
- Confirm that the change actually landed in the account.
- Mark the result as confirmed, monitoring, inconclusive or reversed.
- Add the next decision and its timing.
- Check that campaign names and numbers match the account.
- Edit the section for client readability.
The finished section can feed the broader PPC report format a client actually reads. It should complement the performance summary, not repeat it.
This workflow also creates a useful internal record. When someone asks why a campaign changed or why a test ended, the answer is attached to the action.
Frequently asked questions
How do you prove to a client that you worked on their ad account?
Use the account change history to document each meaningful action, who made it and when it happened. Add the reason for the change and the relevant before and after number so the client can see both the work and its effect.
Which account changes belong in a client update?
Include changes that affected spend, targeting, delivery, messaging or the interpretation of results. Budget moves, negative keywords, copy tests, creative swaps, pauses and un-pauses usually belong, while routine housekeeping can be summarized.
Should you tell a client when a change did not work?
Yes. State what changed, what happened and whether you reversed it, kept it running or need more data before deciding.
How do you pull a change history from Google Ads and Meta?
Use Change history in Google Ads and Activity history in Meta Ads Manager, then filter to the reporting period and the people who made changes. Move the meaningful entries into one working list before adding reasons and results.
How detailed should the 'what we did' section of a report be?
Give enough detail for the client to understand the decision, its scale and what happened next. One or two lines per meaningful change is usually more useful than a raw platform export.
What Ads Arsenal does for this today
In Ads Arsenal, every approved change is written to the ad account and checked a few minutes later to confirm that it landed. The Actions tab keeps the list with its verification state, while the recommendation includes the reasoning and numbers behind the proposed change.
After the change runs, Ads Arsenal tracks the result and grades the recommendation at 3, 7 and 14 days. It keeps what works for that account, marks results honestly when the data is inconclusive and leaves every account change under human approval.
Ads Arsenal is in a free design-partner beta for agencies and in-house teams running real spend on Google, Meta and TikTok. Become a design partner.