In this instructional video, Igor Ivitskiy walks advertisers through three quick screens in Google Ads that reveal whether an agency is actively managing their account. He demonstrates how to use the Change History (looking at the number of changes in the past 90 days and whether they were made by a person or auto‑applied), the campaign ad‑scheduling view (identifying how many campaigns have a non‑all‑day schedule), and the spend‑by‑campaign status view (showing the share of spend sitting in paused or removed campaigns). Ivitskiy explains why a quiet account can be healthy if the agency has a documented waiting condition, and why silent spend is a leak. He then provides three concrete questions advertisers should ask their agency—change count and source, spend in paused/removed campaigns, and the stopping condition for the next test—followed by actionable next steps and a non‑accusatory communication approach. The video ends with a call to subscribe to the “Profit Forensics” series.
Chapters
3 changes in 90 days. That is what one screen in your own Google Ads account showed, while the monthly report for the same period ran to 14 pages. In the next few minutes, I'll show you 3 screens that tell you whether anyone is actually working on your account. It takes just 3 minutes. You need no tools, and nobody has to know you looked. So let's open your account. Here is that screen. In Google Ads, you open Tools and Settings, then Change History, and set the range to the last 90 days. What you're looking at is a reconstruction of how that screen looks on a lot of managed accounts. A budget moved from $400 to $420. One ad group paused. And one line that says auto-applied recommendation, which means Google made that change, not a person. 3 entries, 90 days. Now, before you pick up the phone, I want to argue the other side properly because I've seen this go wrong in both directions. Silence in an account is not automatically bad. Constant tinkering is worse than doing nothing. Every time someone changes a bid or a budget, the automated system starts relearning. A manager who edits every day is mostly paying to reset their own data. So a quiet account can be a well-run account. There is exactly one condition. Somebody wrote down why it's quiet. If your agency can tell you in one sentence what they are waiting for and which number will make them act, that silence is a decision. If nobody can name it, that silence is not a strategy. It's an absence. Here's the second screen. And this one is data, not opinion. Across the accounts I analyzed, I counted the campaigns with a real advertising schedule. Anything other than run all day. 104 out of 15,168 campaign rows. That's 0.7%. 3 accounts out of 28. Now, ad scheduling is not always the right lever. That's fine. But think about what that number implies. If nobody ever set the hours, nobody was ever reading the hourly data either. The lever your agency describes in the proposal was in almost every account never touched. Third screen, and this is where the money is. Look at where the spend ends up by campaign time. Of everything spent on Demand Gen campaigns, 25% now sits in campaigns that are paused or removed. For Performance Max, it's about 15%. For Search, it's under 9%. Some of that is healthy. You're supposed to test things and kill the ones that fail. Here's the difference between an experiment and a leak. An experiment is money you agreed to write off before you spend it. A leak is Money you write off afterwards, quietly. In a campaign, nobody mentions again. Across the accounts I have audited, the ones that leak are almost never the accounts that tried something bold. They are the accounts where nobody wrote the stopping conditions. So here are your 3 questions. 1. How many changes in the last 90 days and who made them? A person or the system? What share of your total spend sits in campaigns that are paused or removed? You get that in the campaign view. Filter the status to paused and removed. And the cost column. Set the range to all time. And this is one that matters most. Which number ends the next test? And who wrote it down? Now, what do you do with the answers? If the change history is thin and your agency can explain what they are waiting for, you have a fine relationship. Ask them to put that sentence in the next report and move on. If the change history is thin and nobody can answer question 3, the problem is not the agency yet. It is that nobody even set the terms. That is one meeting, not a search for a replacement. Ask for 2 things: the objective in your margin, not the return metric, and the number that ends the current test. Whether those 2 answers exist in the thing you're actually testing. And if half your spend is sitting in a graveyard nobody declared, that's your first conversation, and it's worth real money. One thing I'd ask you not to do. Don't walk in accusing anyone. Most of the people running your account are not lazy. They are managing 40 accounts, and yours is quiet because it isn't on fire. What you're doing here is making your account loud enough to get attention. With numbers, not with complaints. There is a longer piece behind this on why performance decays when optimization stops on LinkedIn, link below. If this saved your money, subscribe. New Profit Forensics every week.
