When Google Ads performance changes suddenly, the fastest explanation is often not in another report. It is in the account’s change history.
A budget may have been reduced. A bid strategy may have changed. An ad, keyword, or campaign may have been paused. Someone may have edited targetin – or an automated system may have acted.
The useful question is not “What happened in this account?”
It is: “What changed near the moment performance moved?”
This guide shows how to answer that question with HireOtto using a two-pass workflow: scan for affected resources, then drill into the exact edits. The workflow is read-only; it helps you investigate, not roll changes back.
What Google Ads change history can tell you
Google Ads’ web interface keeps campaign change history for the past two years. It lets you filter by campaign, ad group, change type, user, tool, item changed, and experiment. Google also maps changes against performance metrics so you can investigate whether an edit may have contributed to a shift.
HireOtto makes that investigation conversational. Its current change-history workflow has two levels:
Change status: a broad scan of resources changed in the selected period. It is useful for finding which campaigns, ad groups, ads, or criteria were touched. Google’s Change Status data can cover a window within the past 90 days, but it returns only the latest change for a resource inside that window.
Change events: a detailed timeline of recent edits, including the timestamp, operation, changed fields, before/after resource snapshots, user email, and client type when Google exposes them. Detailed events must be queried within the past 30 days.
Start broad when you do not know where to look. Go straight to detailed events when you already know the affected campaign and time window.
The two-pass change-history workflow
1. Define the symptom and the comparison window
Do not begin with “Show me all changes.” Begin with the business symptom:
Spend jumped or fell.
Conversions dropped after a specific day.
A campaign stopped serving.
CPA or ROAS changed abruptly.
A client or teammate noticed an unfamiliar setting.
Choose a narrow window around the first visible movement. If conversions fell on a Monday, start with the preceding few days through the following day. A tighter window reduces noise and makes the sequence of events easier to interpret.
Prompt:
For account [customer ID], scan change status from [start date] to [end date].
Show the most recently changed resources, grouped by campaign and resource type.
Highlight campaign, ad group, ad, keyword/criterion, budget, bidding, and status-related changes.2. Use the broad scan to find where to investigate
The broad scan is an index, not a forensic log. Because it keeps the latest change for each resource in the selected period, use it to answer:
Which campaigns and ad groups were touched?
Which resource types changed most?
Did changes cluster around the performance shift?
Is the account unusually active for the period?
Shortlist the resources that are both relevant to the symptom and close to the date it began. Ignore unrelated housekeeping unless it shares a dependency – for example, a portfolio bid strategy or shared budget.
3. Pull the detailed timeline
For the shortlisted campaign, request change events in chronological order. The detailed view is where you can inspect the operation, changed fields, and old/new values. It can also show the user email and the client used to make the change when Google provides them.
Prompt:
Get detailed change events for campaign [campaign ID] from [start date] to [end date].
Sort oldest to newest. For each event, show timestamp, operation, resource type, changed fields, old value, new value, user, and client type.
Export the full result to CSV.Use CSV for busy accounts or longer result sets. Google requires a limit and caps a Change Event query at 10,000 rows; narrowing the dates or campaign is safer than asking for the entire account at once.
4. Translate technical fields into business impact
A field change matters only in context. Convert the event log into an operational explanation:
Budget: Could the new amount constrain or expand delivery?
Bidding: Did the strategy, target CPA, target ROAS, or bid limit change?
Status: Was a campaign, ad group, ad, or keyword enabled, paused, or removed?
Ads and assets: Did messaging, final URLs, or eligibility change?
Keywords and targeting: Did match types, negatives, locations, audiences, or other criteria change?
Conversion setup: Did a goal or conversion-related setting change near the performance movement?
Prompt:
Summarize these change events for a PPC manager.
Rank them by plausible impact on spend, traffic, conversions, CPA, and ROAS.
Explain the mechanism for each ranking, and mark any conclusion that requires performance data to verify.5. Correlate changes with performance
Change history supplies a timeline, not proof of causation. Pair it with daily campaign performance for the same period. Look for the sequence: baseline → edit → delivery or efficiency shift.
A useful comparison includes spend, impressions, clicks, conversions, conversion value, CPA, and ROAS where applicable. Allow for conversion lag, learning periods, seasonality, auction changes, and tracking issues before attributing the result to one edit.
Prompt:
Compare this chronological change timeline with daily campaign performance for the same dates.
Identify changes that occurred before a material shift.
Separate confirmed facts, plausible relationships, and unsupported speculation.
Recommend the smallest reversible test or verification step for each plausible relationship.6. Decide what to do next
A good investigation ends with a controlled action—not a blame statement. For each high-impact change:
Verify that the old and new values are interpreted correctly.
Confirm whether the change was intentional and approved.
Check dependencies such as shared budgets, portfolio strategies, experiments, or conversion goals.
Choose whether to keep, adjust, or reverse it in Google Ads.
Document the decision and the metric you will monitor.
HireOtto’s change-history workflow does not undo edits. Make any reversal in Google Ads after reviewing the affected scope and current account state.
How to see who made a Google Ads change
In the Google Ads interface, changes made through the interface can show the user’s email. Automated changes may appear under a system name, while API-originated changes may appear as “Google Ads API” or sometimes a specific tool name.
In HireOtto’s detailed timeline, use the user email together with client type. Treat both as attribution signals, not guaranteed identity:
The user email may be blank or unavailable for some changes.
The client type may identify the interface category rather than the exact third-party application.
Google or automated-system activity can look different from a named human edit.
Very recent changes can take up to about three minutes to appear in Google’s API data.
When attribution is ambiguous, verify access history, approvals, automation rules, and the Google Ads interface before assigning ownership.
Agency takeover checklist
For a new account, review change history before recommending changes. It reveals how actively the account has been managed and helps explain why the current configuration exists.
Scan the past 30–90 days for changed campaigns and resource types.
Pull detailed events for the most important campaigns within the past 30 days.
Flag recent budget, bid strategy, status, ad, keyword, targeting, and conversion-setting edits.
Identify repeated edits, reversals, and changes made by automation or API clients.
Compare high-impact edits with daily performance.
Record questions for the outgoing team or client before changing anything.
This should sit alongside a broader Google Ads account audit and reporting analysis. Change history explains the sequence of edits; an audit evaluates the current state, and reporting measures outcomes.
Ready-to-use prompts
Weekly change digest
Review change status for the last 7 days. Group changed resources by campaign and resource type. Flag unusual volume, repeated edits, and changes involving budgets, bidding, status, ads, keywords, targeting, or conversions. Then list the campaigns that need a detailed timeline.Performance drop investigation
Conversions dropped starting [date]. Scan the surrounding change window, shortlist relevant campaigns, and pull detailed events for those campaigns. Compare the timeline with daily performance. Separate facts from hypotheses and rank the most plausible causes.Account takeover review
Create an account-takeover change review. Use a broad scan for the past 90 days, then detailed events for priority campaigns within the past 30 days. Summarize who or what made high-impact edits, what changed, the likely business effect, and questions to resolve before we change the account.AI and automation verification
Show detailed changes made around [date/time] for campaign [ID]. Include user email and client type when available. Identify edits that may have come through an API or automated system, but do not infer a specific tool unless the data names it.Common mistakes
Reading the log without a symptom. A full account timeline creates noise. Anchor the investigation to a metric movement or operational question.
Treating the broad scan as a complete event log. It reports the latest change per resource in the chosen window, not every edit.
Assuming the editor is always identifiable. User and client fields depend on what Google exposes.
Confusing sequence with cause. A change that precedes a drop may be unrelated; compare performance and account context.
Requesting too much data at once. Use campaign and date filters, then export detailed results when needed.
Reversing a change without checking dependencies. Shared settings and learning periods can make a “simple” rollback consequential.
Where HireOtto helps
HireOtto is a remote Google Ads MCP server that lets a compatible AI assistant inspect and manage Google Ads through conversation. For change history, it supports the two-pass workflow in this guide: scan recently changed resources, pull detailed field-level events for shortlisted campaigns, and export busy timelines to CSV.
The useful part is not asking AI to assign blame. It is bringing the change log and performance reports into the same conversation, so you can move from “something changed” to a reviewable hypothesis while keeping the operator in the approval loop. The change-history workflow itself is read-only and does not roll changes back.
Already using HireOtto? Start with the performance-drop or account-takeover prompt above. New to Google Ads MCP? Read the complete guide.
Get started here: https://docs.hireotto.com/setup/connect-ai-tool
Frequently asked questions
How far back can I review Google Ads change history?
The Google Ads web interface provides campaign change history for the past two years. HireOtto’s broad Change Status scan uses Google Ads API data from a window within the past 90 days. Detailed Change Events are limited to the past 30 days.
Can HireOtto show who changed a setting?
Detailed events can include user email and client type when Google exposes them. Automated, system, or API-originated edits may be labeled differently, and some attribution fields may be blank.
Can HireOtto undo a change?
No. The change-history workflow is for investigation and reporting. Review the evidence, then make any reversal in Google Ads with the appropriate approval.
Why does the broad scan show only one change for a resource?
Change Status returns the most recent change for that resource within the selected period. Use Change Events when you need every available field-level edit in a recent timeline.
Why is a very recent change missing?
Google notes that changes can take up to about three minutes to appear in Change Status or Change Event results. Retry after a short delay before assuming the edit was not recorded.
Can change history prove that an edit caused a performance change?
No. It can establish timing and show what changed. You still need performance data and account context to evaluate causation.
Use change history as evidence, not a verdict
The most useful change-history review is narrow, chronological, and tied to a decision. Start with the symptom. Find the resources that changed. Pull the detailed timeline. Translate fields into business impact. Then compare the sequence with performance before acting.
That workflow turns an account log into something operational: a defensible explanation of what changed, who or what may have changed it, and what to verify next.
About Me
I’m Suyash – badminton junkie, ex‑GroupM ad‑ops grunt, first marketer at a B2B SaaS startup, and creator of Hiretto: Google Ads MCP Server.
My mission: less clicking, more thinking.
Let’s build leverage together.