This post describes a pattern I've observed across accounts over many years of managing Google Ads campaigns. I want to be upfront about the framing: what follows is my working theory based on what I see in account data, not a verified technical fact about how Google's algorithms operate. Google doesn't publish the internal logic of its bidding systems. But the pattern is consistent enough, and the incentive structure clear enough, that I think it's worth naming and watching for.
I've tracked quarterly CPC trends across a set of well-managed accounts over the last two years. In accounts using fully automated bidding with no manual constraints, average CPCs have drifted upward in most cases even where Quality Scores have held steady or improved. In accounts where I've introduced Target CPA constraints or manual bid caps, the drift has been much less pronounced. That's not conclusive, but it's the pattern that led me to the theory below.
The concept is Minimum Acceptable Return, or MAR.
What MAR is and why it matters
MAR refers to the minimum return on ad spend that a business needs to keep advertising. Every advertiser, consciously or not, has an implicit threshold: below a certain cost-per-conversion or ROAS, the campaign stops making sense and they pause or stop spending. For a retail business, that might be a ROAS of 2.5x. For a B2B lead generation business, it might be a maximum CPA of £100 per qualified lead.
Google's interests are aligned with keeping advertisers spending, but specifically, keeping them spending as much as possible while staying above the point where they'd pull out entirely. If an advertiser is achieving a ROAS of 5x and their MAR is 2.5x, there's what I'd call "headroom": the advertiser could absorb higher costs and still remain active. From Google's perspective, that headroom is potential additional revenue.
What MAR optimisation looks like in practice
This is where the theory becomes observable. I'm not suggesting Google has a deliberate MAR-targeting algorithm. I genuinely don't know. What I am observing is a consistent pattern in account behaviour over time that is consistent with what you'd expect if such a dynamic were operating.
The first pattern is gradual CPC increases without Quality Score changes. Accounts that are performing well, with high Quality Scores and strong conversion rates, should be getting CPCs at or below market rate. When I see CPCs drifting upward quarter-over-quarter in accounts where Quality Score and competition haven't materially changed, that's unusual. It's consistent with the auction dynamics shifting in a way that extracts more per click from high-performing advertisers.
The second is ROAS declining slowly over 6 to 12 months. A campaign that achieved a ROAS of 4x in its first year settling toward 2.5x by year two, without obvious changes to the account or the market. Some of this is normal: campaigns mature, competition increases, the easy early wins get captured. But when the trajectory is smooth and gradual rather than correlated with any identifiable external change, it warrants scrutiny.
The third pattern is budget recommendations that consistently push spend upward. Google's automated recommendations almost always suggest budget increases. The recommendation to "increase budget by 20% to capture missed opportunities" is usually framed as being in your interest. Sometimes it is. But the structural incentive behind that recommendation is Google's, not yours.
The fourth is Smart Bidding strategies pushing toward higher spend. Maximise Conversions with no CPA cap will spend the entire daily budget regardless of the cost of each conversion. If conversion quality isn't monitored closely, this can result in the budget being consumed at a higher CPA than you'd accept if you were reviewing each conversion individually.
Patterns worth monitoring in your account
If you're concerned about whether MAR dynamics might be affecting your performance, these are the metrics worth tracking systematically over time.
Track quarterly CPC trend for your core keywords. Even accounting for seasonal variation, are CPCs in your main ad groups drifting upward? Compare the same quarter year-over-year rather than month-to-month to control for seasonality.
Track ROAS by campaign monthly over at least 12 months. Plot it. Is it stable, improving, or slowly declining? A slow decline that doesn't correspond to any changes you've made or any market shift you're aware of is worth investigating.
Watch conversion value relative to spend, not just conversion volume. An account that's generating more conversions at higher cost may be hitting volume but reducing profitability. Watch the ratio, not just the absolute numbers.
If you're fully automated with Target ROAS and you haven't tested whether a more controlled bidding approach would perform similarly at lower cost, you don't have a baseline for comparison. Experiments are useful here.
What to do if you see these patterns
The first step is to verify that external factors don't explain the change: new competitors entering the market, seasonal volume changes, economic factors affecting conversion rates. Eliminate the obvious explanations before attributing anything to algorithmic dynamics.
If external factors don't account for the pattern, the practical responses are:
- Introduce more controlled bidding, such as Target CPA or Target ROAS with a specific constraint, if you've been running on Maximise Conversions or uncapped Smart Bidding
- Restructure campaigns to reduce Google's ability to optimise toward cheap, easy conversions at the expense of high-value ones (audience exclusions, tighter geographic targeting, explicit keyword exclusions)
- Audit your auto-apply recommendations settings and turn off any that allow Google to make changes to campaigns without your review
- Test segments of budget on manual bidding as a control to compare performance against automated alternatives
None of this requires assuming bad faith on Google's part. Regardless of the mechanism, the practical advice is the same: monitor your account metrics at a level of granularity that lets you detect gradual changes, and don't cede so much control to automation that you can't identify when something is drifting. For smaller businesses in North Wales where ad budgets are tighter and margins less forgiving, catching a gradual ROAS decline early makes a real difference.
Google Ads management in North Wales: if you'd like a review of your campaign's performance trends and whether your bidding strategy is working in your interest, get in touch.
