You're Paying for a Campaign That Runs Itself

A roofing company in Phoenix was paying their agency $4,500 a month to manage their Google Ads. Total monthly ad spend was $12,000. When we audited the account, we found something that made our founder genuinely angry. Every single active campaign was Performance Max. Automated bidding. Auto-optimized creative. The whole thing was running almost entirely on Google's machine learning. And the agency had made exactly four changes to the account in the previous 90 days. Four. In three months.

The agency was billing $13,500 over that period for work that amounted to two hours, maybe three. This isn't a horror story. It's Tuesday in the Google Ads agency world.

Here's the uncomfortable truth that no agency will put in their pitch deck: Google and Meta have automated most of what agencies used to charge a premium to do. Performance Max, Smart Campaigns, automated bidding, responsive ads, dynamic creative. The platforms have eaten the manual labor that justified big monthly retainers. And most agencies haven't adjusted their fees or their workflows to reflect that reality.

So what exactly is your agency doing all month to justify what they're charging? We've managed $550 million in ad spend across 400+ clients and we can answer that question directly. You might not like the answer.

What Google Ads Automation Actually Does Now

To understand why this is a problem, you first have to understand how radically the platforms changed in the last five years. Five years ago, running Google Ads well was genuinely labor-intensive. You had to manually set bids on individual keywords, build out ad group structures with tightly controlled themes, write multiple ad variations and rotate them yourself, and build separate campaigns for search, display, shopping, and remarketing. It took real hours every week to keep an account running efficiently.

Now look at what Performance Max does automatically. It runs across Search, Display, YouTube, Gmail, Discover, and Google Maps simultaneously. It tests your creative assets against each other and shifts budget toward what's working. It identifies audiences likely to convert without you telling it who to target. It adjusts bids in real time based on thousands of signals including device, location, time of day, search history, and browsing behavior. It can even generate headlines and descriptions if you let it.

Smart Bidding takes it further. Target CPA tells Google to find conversions at your target cost and the algorithm handles every bid adjustment automatically. Target ROAS does the same for revenue efficiency. Maximize Conversions just pushes volume within your budget. These aren't amateur features. When set up correctly with solid conversion data feeding them, Google's own data shows Smart Bidding outperforms manual bidding in the majority of accounts.

The platform automated the work. The fees didn't follow.

The Agency Math Doesn't Add Up Anymore

Let's get specific about time. A well-run Google Ads account spending $15,000 a month with a handful of Performance Max and search campaigns needs roughly 10 to 15 hours of active management per month. That's real work done thoughtfully. Not 10 minutes of checking dashboards. Actual strategic work.

At a very generous $150 per hour, that's $1,500 to $2,250 in labor cost. Agencies typically charge 10 to 20 percent of ad spend as their management fee. On $15,000, that's $1,500 to $3,000 a month. Fine. Defensible if they're doing real work.

But plenty of agencies charge $15,000 to manage a $50,000 account. That's 30 percent. At $150 an hour, they'd need to spend 100 hours on your account every month. Nobody is doing 100 hours a month on a single mid-size account. Nobody. The math has never been honest and automation made it worse.

We've seen agencies with account managers carrying 30 to 40 clients simultaneously. Do that math. If each manager works 40 hours a week, 160 hours a month, and has 35 clients, you're getting about 4.5 hours of attention per month. At best. That's your $5,000 retainer: 4.5 hours, much of which is generating the report that makes it look like more work happened.

What Your Agency Is Actually Doing All Month

We're going to say something most agencies will never admit. Here's how a typical month actually looks at a lot of Google Ads agencies with automated campaigns.

Week one. The account is on autopilot. Performance Max is running. Smart Bidding is running. Nobody is touching anything because Google tells you not to make frequent changes during the machine learning phase. This is real advice. Constant interference resets the learning period. So week one, the agency is technically correct to leave it alone.

Week two. Someone checks the performance dashboard. Numbers look okay. No major red flags. Nothing gets changed.

Week three. A junior account manager notices you're slightly under budget pace. They bump the daily budget cap by 15 percent. This takes about six minutes.

Week four. Someone pulls a report. A nice PDF with lots of charts gets assembled in whatever reporting tool they use. This takes maybe two hours including the email they write to you explaining that impressions are up and the algorithm is "optimizing toward your target CPA." The report is emailed.

That's the month. Four to six hours of actual work. You paid $5,000.

Is this every agency? No. There are genuinely good ones that do meaningful strategy work. But after auditing hundreds of accounts, we can tell you this is common enough that you should be skeptical until proven otherwise.

What Real Google Ads Management Looks Like Now

The job changed. Good agencies adapted. Bad ones just kept cashing the checks. Here's what legitimate management of an automated Google Ads account actually involves.

Search Term Analysis and Negative Keyword Expansion

Performance Max doesn't eliminate the need to audit what searches are actually triggering your ads. It just makes that data harder to find. Google limited search term reporting inside PMax, which means an experienced manager has to dig through insights reports, asset group performance, and auction data to identify wasted spend. Adding negative keywords at the account level remains one of the highest-leverage actions in any campaign, automated or not. It has to be done consistently.

Audience Signal Refinement

Performance Max accepts audience signals as guidance, not hard targeting. You're essentially telling Google where to look first. Weak signals mean Google casts a wide net and wastes budget finding its footing. Strong signals built from first-party customer data, website visitor segments, and CRM uploads dramatically shorten the learning period and improve efficiency. Building and refining those signals is real work that pays off measurably. Most agencies skip it or do it once at setup and never revisit.

Conversion Tracking Integrity

Smart Bidding is only as smart as the conversion data feeding it. If your tracking is misconfigured and counting the same conversion twice, your Target CPA algorithm thinks it's performing better than it is and scales spend. If a form submission stopped firing and you're not catching it, the algorithm thinks nothing is converting and starts restricting delivery. Conversion tracking audits should happen monthly, not at setup and never again. We see broken tracking in at least a third of accounts we inherit. It's one of the most common and most expensive problems in Google Ads.

Asset Performance and Creative Strategy

Automated campaigns still need good creative inputs. Performance Max tests your headlines, descriptions, images, and videos against each other, but it can only test what you give it. Low-asset accounts with three headlines and two images are leaving significant performance on the table. Strong creative strategy means regularly adding new assets, retiring underperformers, testing angle variations, and aligning ad creative with landing page messaging. None of that happens automatically.

Budget Architecture Across Campaigns

Most advertisers don't run a single campaign. They have brand campaigns competing with their own Performance Max campaigns for budget, shopping campaigns overlapping with PMax, and remarketing running alongside prospecting. How those campaigns are structured and how budget flows between them dramatically affects efficiency. A good manager actively engineers the relationship between campaign types to prevent cannibalization and ensure budget is working its hardest.

Landing Page Alignment and CRO Input

The best bidding algorithm in the world can't fix a landing page that's bleeding conversions. Google Ads management that stops at the ad click is half a job. The agency should be looking at post-click behavior, identifying high bounce pages, flagging poor mobile experiences, and pushing for landing page testing. Most agencies treat this as out of scope. The smart ones know that improving conversion rate has a multiplier effect on every dollar of ad spend.

Competitive Intelligence

Google's Auction Insights report tells you exactly which competitors are appearing alongside your ads and how their impression share compares to yours. When a competitor starts aggressively outbidding you or a new player enters the auction, a manager paying attention can identify this quickly and recommend a response. Nobody is doing this on your account if they're spread across 35 clients checking dashboards once a week.

The Performance Max Problem Specifically

Performance Max deserves its own section because it's the campaign type most responsible for agency laziness right now. PMax is genuinely powerful. We're not anti-PMax. We run it for clients regularly and it delivers. But it's also the easiest campaign type to set up poorly and ignore.

Here's what we see constantly. An agency launches a single Performance Max campaign with all products or services shoved into one asset group. No audience signals added. Default creative: a couple of headlines, one image from the website, and no video. Then they let it run for months without structural improvement.

What should actually happen is that asset groups get structured around product lines, service categories, or audience types so performance can be isolated and optimized at a granular level. Brand exclusions need to be considered so PMax isn't cannibalizing your existing brand search campaigns. Video assets, even simple ones, dramatically expand reach into YouTube inventory. Customer match lists built from your CRM data should seed the algorithm with quality signals from day one.

The set-it-and-forget-it version of PMax is everywhere. The well-structured version that actually gets managed and improved monthly is rare. Guess which one most agencies are running for you.

The Percentage Model Is the Root of the Problem

None of this is accidental. The incentive structure of percentage-of-spend pricing actively rewards agencies for doing less. When your fee is tied to how much you spend rather than what results you get, there's no financial motivation to push for efficiency. An agency earning 15 percent of your $30,000 budget makes $4,500 a month. If they help you get the same results from $20,000 through better optimization, they just cut their own revenue by $1,500. The math actively discourages the work that helps you most.

Flat fee pricing changes that dynamic completely. When an agency earns the same amount regardless of what you spend, they're finally aligned with your actual goal: getting the best possible return on every dollar, not maximizing how many dollars go through the platform.

Look at what percentage pricing does when automation is involved. An agency sets up a well-structured Performance Max campaign that runs efficiently with minimal intervention. The machine is working. The smart thing from a performance standpoint is to let it run, monitor it weekly, make small refinements, and focus energy on creative and landing page testing. From a billable-hours perspective, that's not a lot of work. Under a percentage model, the agency is tempted to justify their fee by doing things. Changing bids that the algorithm was already optimizing. Restructuring campaigns that were working. Manufacturing activity to make their monthly report look busier. We've seen it.

A flat fee removes that temptation entirely. Do the right thing for the account. The fee doesn't change either way.

How to Tell If Your Agency Is Actually Working

You don't have to take their word for it. Here are the direct questions to ask and what honest answers look like.

Ask for a monthly activity log. Not a performance report. An activity log that lists every change made to the account, when it was made, and why. In Google Ads, this is called the Change History. It's visible to anyone with account access. If you have access to your own account, you can check it yourself right now. Go to Tools, then Change History, and see what your agency has actually touched in the last 30 days. What you find might surprise you.

Ask how many active clients your account manager handles. Industry standard is 15 to 25 clients per manager. Anything above 20 and you're getting the absolute minimum viable attention. Above 30 and you're essentially unmanaged. This one question will tell you more about your service level than any pitch deck stat.

Ask what specific tests are currently running in your account. Real management involves ongoing creative tests, audience experiments, landing page variations. If your agency can't name a specific active test when asked, nothing is being tested. The account is static and your results will be too.

Ask what search terms have been added as negatives in the past 90 days. This is one of the most consistent indicators of whether someone is actively managing the account. If they can't answer or can't pull the list, the answer is probably nothing meaningful.

What You Should Actually Be Paying For

Automation didn't eliminate the need for expert management. It changed what that expertise looks like. The job shifted from executing manual tasks to thinking strategically about account architecture, data quality, creative direction, and performance interpretation.

Good Google Ads management in 2026 requires someone who understands how Performance Max asset groups should be structured to isolate performance by product and audience. It requires someone who can read auction data and identify when a competitive shift is affecting your costs. It requires someone who understands conversion tracking deeply enough to audit it systematically and catch the drift that happens over time. It requires creative thinking about message and offer, not just keyword bidding.

That expertise has real value. But it doesn't take 40 hours a month to apply on a mid-size account. And it definitely shouldn't scale linearly with your ad spend. If you spend $50,000 a month instead of $20,000, your account doesn't need 2.5 times more management hours. It needs the same strategic oversight, maybe with more asset group complexity, but not proportionally more labor.

The reality is, the agency model where fees grow automatically as your spend grows made sense in an era of manual bid management. That era is over. Google did the industry a favor by automating the grunt work. Now you should expect your management fee to reflect what's actually being done, not what used to be done five years ago.

The Bottom Line

Smart Campaigns and Performance Max are legitimately good products when set up correctly and managed strategically. They're not a reason to fire your agency. They are a very good reason to audit what your agency is doing and whether your current fee structure makes any sense given how much the platforms now do automatically.

If you're paying $5,000 a month and getting four changes in 90 days, you're paying for a campaign that runs itself. And you're paying someone else to watch it run.

Real management in the automation era means smarter setup, more strategic thinking, stronger creative inputs, tighter conversion tracking, and better-informed decisions about where to spend your budget. It's not less valuable than the old way. It's just different work. And it shouldn't cost you three times what it's worth because an agency hasn't updated their pricing since 2018.

If you want to know exactly what's happening in your account, get someone to show you the Change History. Ask the questions above. Look at whether the work being done matches the fee being charged. You might be surprised what you find.