Google Ads Change History: How to Find Out What Actually Broke Your Account

You open the account on a Monday and conversions are down 40% from two weeks ago. Nothing in the news, no holiday, no obvious market shift. The first instinct is to blame the algorithm, or the audience, or “the platform changing things again.” Sometimes that’s true. More often, someone changed something in the account three weeks ago, the effect took a while to show up in the data, and nobody wrote it down anywhere except in one place: Google Ads’ own change history report.

Change history is one of the least-used tools in Google Ads, mostly because it’s buried under the Campaigns menu and doesn’t do anything flashy. It doesn’t have a score, a benchmark, or a headline number. But it’s also one of the few places in the platform that gives you a factual, timestamped record of what actually happened in an account, rather than a guess dressed up as an insight. If you manage more than one account, or you’ve ever inherited an account from someone else, learning to read it properly will save you hours of speculation — and stop you from “fixing” something that was never actually broken.

This post covers what the report actually shows, exactly how to find and filter it, a step-by-step workflow for investigating a real performance problem, where the tool falls short, and how the same underlying idea — treating “what changed and when” as a first-class question — matters even more once more than one person, or one automated system, has write access to an account. It’s a less glamorous topic than bidding strategy or ad copy, but it’s the kind of thing that saves real time the first time you actually need it.

What the change history report actually shows

Change history is a chronological log of edits made to an account. According to Google’s own documentation, it covers roughly the past two years of activity, which is generally long enough to catch anything relevant unless you’re investigating something genuinely ancient. Data older than that simply isn’t retained.

The range of what gets logged is broader than most people assume. It’s not just budgets and bids. It covers keyword additions, pauses, and removals; negative keyword list changes at both the campaign and shared-list level; ad copy edits, including headline and description swaps; asset (extension) changes like sitelinks, callouts, and structured snippets being added or removed; audience list additions and removals; geographic and device targeting adjustments; ad rotation settings; bid strategy switches, including moves between manual and automated strategies and changes to a Target CPA or Target ROAS goal; campaign and ad group status changes (paused, enabled, removed); conversion action edits, including a conversion being marked primary or secondary, or its count setting changed; and account-level items like new user access being granted.

Each entry records four things: what was changed, the old value and the new value, the timestamp, and — if the edit was made through the Google Ads interface by someone logged in under their own account — who made it. That last part matters more than it sounds like it should. In an account with one person logged in, “who” is rarely in question. In an account with an agency, a client’s in-house marketer, and a couple of automated processes all touching the same campaigns, it’s often the entire question.

The feature most people skip past, though, is that change history doesn’t just list edits in isolation — it plots them against your account’s performance timeline. Impressions, clicks, conversions, click-through rate, and cost are all graphed alongside the change events, so you can visually line up a dip or spike in performance with whatever was edited around the same time. That’s the whole value proposition of the tool: it turns “something feels off” into “here is exactly what happened right before it went off, and by whom.”

Where to find it, and how to actually read it

Change history sits under the Campaigns menu in the left-hand navigation of Google Ads, usually labelled simply “Change history.” Once you’re in it, there are a handful of controls worth knowing well, because using them badly is the main reason people give up on the tool after one look:

• Date range. Set this to match the window you’re actually investigating. If performance started slipping sixteen days ago, don’t leave the default range at seven days — you’ll see nothing and conclude, wrongly, that nothing changed.
• Change type filter. You can narrow the log to specific categories: budgets, bids, keywords, ad group or campaign status, targeting, ads and assets, and more. If you already suspect the issue is budget-related, filtering to just budget changes removes a lot of noise fast.
• Campaign or ad group scope. You can view changes across the whole account or drill into a single campaign or ad group, which is useful once you’ve narrowed down where the problem lives from your performance reporting.
• User column. If people are logged in with their own Google accounts rather than sharing one login, the report shows whose account made each change. Google also added a “Go to…” shortcut here — you can jump straight from a logged change to the actual campaign or ad group it affected, instead of copying a campaign name and searching for it manually afterward.

The layout combines a timeline graph of performance metrics at the top with a list of individual, expandable change events underneath. Click into any entry and you’ll see the specific old value and new value side by side — not just “budget changed,” but “$50/day changed to $20/day.” That combination of graph-on-top and discrete log-entries-underneath is what makes the tool useful for actual investigation rather than passive record-keeping.

The workflow for investigating a performance drop

Most people either never open change history, or open it and scroll aimlessly hoping something jumps out. A more structured approach looks like this:

Step one: find the exact date the metric moved, before you open change history at all. Start in your campaign performance view, broken down by day, and look at conversions, cost, and impressions over the affected date range. You’re looking for the day the line actually bends, not the day you happened to notice the problem. Those two dates are often two to three weeks apart, especially with conversion lag, a longer sales cycle, or a Smart Bidding strategy that took time to fully react to whatever changed.

If the issue looks like it’s specifically about visibility — fewer impressions or a lower average position, rather than a drop in conversion rate on similar traffic — it’s worth checking impression share alongside change history, since a meaningful chunk of that can come down to whether you’re losing share to budget or to rank, which points toward completely different fixes than a change-history entry would.

Step two: widen the change history date range to start several days before the bend. Changes don’t always show effects immediately. A bid strategy switch, a large budget cut, or a change to a Smart Bidding target can take days to fully propagate as the algorithm relearns and exits its learning phase. Give yourself a buffer of five to seven days before the date you identified in step one, so you’re not missing the actual cause because it happened just outside your window.

Step three: filter by change type if you already have a hypothesis, otherwise scan everything in the widened window. If cost dropped and conversions dropped roughly proportionally, start with budget and bid changes — that pattern usually means less money went into the auction, not that the account got worse at converting. If conversions specifically fell off a cliff while cost and clicks stayed flat or even rose, that points more toward tracking or a conversion action being edited, paused, or demoted — which is exactly the kind of thing that turns up when you work through an audit of your conversion tracking setup. If it’s specific keywords or a single ad group that fell off while the rest of the account looks normal, filter to keyword and status changes scoped to just that ad group.

Step four: cross-reference against the “who.” If a change was made by a specific, identifiable team member, that’s your fastest route to context — ask them what they were trying to accomplish, because there’s often a reasonable explanation that just wasn’t communicated. If the change was made by the system through auto-applied recommendations, that’s a different conversation entirely, usually with yourself, about whether that recommendation category should still be set to auto-apply at all.

Step five: don’t stop investigating at the first plausible change you find. It’s tempting to spot one edit near the right date and declare the mystery solved. Sometimes there are two or three changes clustered in the same week — a budget increase and a bid strategy switch made within days of each other is a common combination — and only one of them is actually responsible for the shift you’re seeing. Compare the magnitude and exact timing of the performance change against each candidate before you commit to an explanation, especially if you’re about to reverse something a colleague or client deliberately did for a reason you haven’t heard yet.

As a concrete illustration of how this plays out: say a search campaign’s conversions drop by roughly a third starting on a Tuesday, but nobody notices until reporting is pulled the following Monday. Opening change history for the two weeks prior surfaces three things — a shared budget was split into two separate campaign budgets nine days earlier, a negative keyword list was updated eleven days earlier, and one ad group’s bid strategy moved from Maximize Conversions to Target CPA six days earlier. Cost per day barely moved, which points away from the budget split as the main driver. The negative keyword list change happened five days before the actual drop, which is too far in advance to line up cleanly. The bid strategy switch to a Target CPA goal, however, lands almost exactly where the performance line bends — and a fresh Target CPA campaign typically goes through a learning period during which volume can dip before it stabilizes. That’s the kind of reasoning change history is built to support: not a single obvious smoking gun, but a short list of candidates you can rule in or out by timing and magnitude.

What change history won’t tell you

This is the part that trips people up: change history logs edits made to the account. It does not log everything that can affect the account’s performance, and treating it as a complete record leads to false confidence.

A few specific gaps are worth knowing before you rely on this tool as your only diagnostic step:

• Smart Bidding’s internal bid decisions aren’t logged as discrete changes. Automated bidding strategies adjust actual bids at auction time continuously, using signals about the user, device, time of day, and context. None of that shows up as a line item in change history — only when you change the strategy itself, its target, or its budget does an entry appear.
• Quality Score movement isn’t tracked as a change event. The 1–10 score you see in the interface is a historical, aggregated figure, while the actual ad quality calculation happens fresh at each individual auction. You won’t find a change history entry that reads “Quality Score dropped from 7 to 5 on this date” — you have to infer that from CTR, ad relevance, and landing page experience trends instead.
• Competitor behavior and auction dynamics are invisible here. If a new, well-funded competitor enters your auction, or an existing one raises bids sharply, that shows up in your auction insights report and your impression share numbers, not in your change history, because nothing in your own account actually changed.
• External factors aren’t logged at all. Seasonality, a landing page that broke after a site update, a pricing change on your website, or a tracking pixel that stopped firing after a migration — none of that lives inside Google Ads, so change history has nothing to say about any of it.
• Tools outside Google Ads that still affect your account aren’t captured. If someone edits a tag in Google Tag Manager and breaks conversion tracking, or a developer changes a redirect that affects your final URLs, Google Ads change history won’t show any of it, even though the downstream effect on your account is very real.

The practical implication is that change history is a great first stop and a poor last one. If you go through the account edit log carefully and find nothing that lines up with the timing of the drop, that’s genuinely useful information too — it tells you the cause is probably external to the account itself, and you should widen the investigation to tracking, the website, or the broader market rather than continuing to dig through Google Ads settings that never actually moved.

The “who changed this” problem in agencies and shared accounts

The value of the user-attribution column scales directly with how many hands are in an account. A freelancer running their own single account rarely needs it, because there’s usually only one plausible answer to “who did this.” An agency managing accounts across multiple clients and team members needs it constantly, because the number of people who could plausibly have made a change — and the number of accounts to check across in the first place — both go up at the same time.

The common failure mode is a client who still has their own login and makes a “quick fix” — pausing a keyword they didn’t like the look of, raising a daily budget without telling anyone, editing an ad’s headline — without mentioning it to the agency managing the account day to day. Weeks later, performance looks strange, and the agency spends an afternoon investigating an issue that a two-minute look at change history, filtered to that client’s user account, would have resolved in the first search.

The same problem shows up in reverse: a junior team member making a change without full context on why a campaign was structured a certain way, or a former employee’s login still having live access to an account long after they’ve left the agency or the client company. Change history won’t stop any of these things from happening on its own, but it turns “I have no idea why this happened” into “here’s exactly what happened, when, and who did it,” which is a much faster and less political starting point for whatever conversation follows. It’s also worth a periodic look at who currently has account access at all — stale logins from people who no longer need it are a common, boring source of exactly this kind of surprise edit.

Auto-applied recommendations are a change too — and they show up here

Google’s optimization score generates a running list of recommendations, and for certain categories you can choose to let them apply automatically instead of reviewing each one yourself. It’s worth being clear-eyed about which recommendation types are worth auto-applying and which ones deserve a manual look first, because once auto-apply is switched on for a category, it keeps making that kind of change indefinitely without asking again — it doesn’t check in with you a year later to confirm you still want it running.

The reasonably good news is that these auto-applied changes are recorded in change history the same way manual edits are, just attributed to the recommendation system rather than a named person. If you go to the Recommendations page, open Auto-apply settings, and check the History tab, you can see exactly what’s been auto-applied and when. You can also find the same information by filtering change history for the date auto-apply was first turned on for a given category — the record shows which specific user enabled it in the first place, which is useful when nobody currently on the team remembers doing it.

This matters because auto-applied changes are exactly the kind of thing that quietly erodes performance over months without anyone noticing, precisely because nobody manually clicked a button at the time it happened and nobody remembers flipping the setting on eighteen months earlier. A recurring habit of scanning change history for “Google Ads” or “Recommendations” as the change author, not just human team members, is one of the few ways to catch this category before it compounds.

Exporting and documenting changes as a habit, not just a rescue tool

Change history is also worth using proactively rather than only reactively. If you’re handing an account off to a new team member, taking over an account from a previous agency, or documenting why a campaign is structured the way it is for a client who asks, the change log is a factual record you can point to instead of relying on institutional memory that walks out the door when someone leaves. You can export the underlying data for a given period, which is useful for building a simple audit trail alongside major account milestones — a Performance Max migration, a bidding strategy overhaul, a big seasonal budget shift — so that six months later, when someone asks “why did we do that,” there’s an actual answer instead of a shrug.

The habit worth building isn’t complicated: whenever you make a deliberate, meaningful change to an account, note briefly why you made it somewhere durable — a shared doc, a project management tool, whatever your team already uses. Change history will tell a future investigator what changed and when with total accuracy. It won’t tell them why, and “why” is usually the part that actually matters when you’re deciding whether to reverse something or leave it alone. A one-line note next to a budget change — “raised to $80/day ahead of the product launch on the 12th” — takes ten seconds to write and can save a future colleague an entire afternoon of guessing.

Does Google Ads tell you when something changes?

Not by default, and this is worth being explicit about, because a lot of people assume the platform will flag anything unusual on its own. Google Ads doesn’t send a native email or in-app alert every time a budget is edited, a keyword is paused, or a bid strategy is switched. Change history is a passive log you have to go and open — it doesn’t push anything to you.

There are two ways people close that gap, and both require setup rather than coming built in. The first is automated rules: you can configure a rule against specific conditions — pause a campaign if cost exceeds a threshold, for example — and choose to be emailed when the rule runs. That’s useful, but it’s reactive to performance thresholds you defined in advance, not a general “tell me when anything changes” alert. The second is Google Ads Scripts, which can poll the ChangeEvent or ChangeStatus data mentioned above and email a summary on a schedule, or compare recent performance against a historical baseline to flag anomalies. Google even publishes a reference account anomaly detector script that does roughly this for a single account.

Both routes work, but both put the burden of building and maintaining the monitoring layer on you — writing the script, hosting it, deciding what counts as anomalous, and updating it as your account structure changes. For a single account managed by one person, that’s a reasonable weekend project. Across a portfolio of accounts with different budgets, seasonality, and bidding strategies, hand-rolled scripts tend to either miss things or generate so many false positives that people stop reading the emails, which is the same failure mode as not having monitoring at all.

Beyond the interface: change history at API scale

The interface version of change history is built for looking at one account at a time, manually, after you’ve already noticed a problem. If you’re managing several accounts, checking each one’s change log by hand every few days doesn’t scale, and by the time you do get around to checking, the damage from a bad change has often been running for a week or two already.

The Google Ads API exposes the same underlying data through two resources with different tradeoffs. The ChangeEvent resource gives you field-level detail — the specific resource changed, the old value, the new value — but only over a rolling 30-day window and capped at 10,000 rows per query. The ChangeStatus resource covers a longer 90-day window and is lighter weight, telling you which resources changed without the full before-and-after detail, which makes it better suited to a “has anything changed since I last checked?” polling pattern run across a large number of accounts at once.

In practice, this is the real gap between checking an account and monitoring one. Manually opening change history is fine for investigating something you’ve already noticed. Catching a bad change while it’s still cheap to reverse — before it’s been live for two weeks and quietly burned through a meaningful chunk of the month’s budget — requires something watching continuously, rather than a person remembering to go and look.

That’s the specific problem Growera’s continuous account monitoring is built around. Instead of relying on someone to open change history after they’ve already noticed a problem, ERA checks the account daily, flags changes and anomalies as they happen, and surfaces the ones that actually warrant attention — rather than handing you a raw, unfiltered log you still have to interpret from scratch every time.

Building the habit

None of this requires new tooling to start. The cheapest version is simply making change history part of your regular account review, rather than something you only open in a panic after a client asks what happened. If you’re trying to work out how frequently that regular review should happen in the first place, how often you actually need to check a Google Ads account is worth reading alongside this, since the two habits reinforce each other directly — a regular check is what lets you catch a bad change within days rather than a month, and change history is what tells you exactly what that change was once you’ve noticed something’s off.

A simple version of the habit: once a week, open change history for the account, set the range to the last seven days, and scan it — not because you’re expecting to find something every time, but because the one week you skip is statistically the week something worth catching happened. Pay particular attention to budget changes, bid strategy switches, conversion action edits, and anything attributed to auto-apply rather than a named person. None of that takes more than a few minutes once it’s a habit rather than a one-off investigation.

The bottom line

Change history won’t diagnose everything wrong with a Google Ads account — it has real gaps around Quality Score, auction dynamics, Smart Bidding’s internal decisions, and anything happening outside the platform entirely. But for the specific, extremely common question of “something changed and I don’t know what,” it’s the most direct source of truth Google Ads gives you, and it stays underused mostly because it’s not the first place people think to look when something goes wrong.

The workflow is straightforward: find the exact date performance moved using your own reporting, widen the change history window to include the days before that date, filter by whatever you suspect, check who or what made the change, and don’t stop investigating at the first plausible answer you find. Do that consistently — whether by hand on a weekly schedule, through a script you’ve built and maintained yourself, or with something monitoring the account for you continuously — and “no idea why this happened” stops being an acceptable answer for an account you’re responsible for. The account will keep changing whether or not you’re watching. The only real choice is whether you find out about it in an afternoon or in a quarter.