LinkedIn
By December 31, 2028, fewer than one in four managers running Google Ads accounts under $100k a month will price primarily as a percentage of spend. Check the proposals when we get there. I expect the old fee to split into a flat price for execution and a separate retainer for strategy. Here are five dated predictions, including what would prove each one wrong.
I managed accounts the old way for years, by hand. The 10 to 20% fee many agencies quote, with around 15% as a benchmark, made sense when work roughly scaled with budget. More spend meant more keywords, more bids to touch, more search terms to mine at 1am. Autonomous execution breaks much of that link. Pricing has not caught up. I used to tell clients the percentage kept us aligned. I was wrong. It kept us paid.
The cost behind the percentage: $3,000 a month, plus what is left out
Say you spend $20k a month on ads. At 15%, management costs $3,000 a month. Published agency minimums run $1,500 to $5,000 monthly, plus $1,000 to $10,000 in setup. That puts the management and setup range at $2,500 to $15,000 in month one, before media spend or any separately billed landing page. The ranges are wide because scope varies. A percentage makes those differences remarkably easy to hide.
A $3,000 retainer buys roughly 15 to 20 hours from a manager juggling 8 to 12 accounts. Here is how I would break down the work behind that fee. The hours in the table are an illustrative allocation, not a vendor’s published timesheet; the linked sources provide the pricing and workload benchmarks.
| Work on a $20k/month account | Hours/month | Cost type | Source for benchmark or listed price |
|---|---|---|---|
| Bids, negatives, budget pacing, search-term mining | 6 to 8 | Recurring; largely mechanical | Swydo on minimum account hours |
| Reporting, calls, change logs, QA | 4 to 5 | Recurring; largely mechanical | Velocity PPC on what a retainer buys |
| Offer, targeting strategy, landing-page direction | 4 to 6 | Recurring; judgment | Velocity PPC on separately billed page work |
| Setup, tracking audit, feed fixes | 10 to 20 | One-off; setup prices vary | PPC.io on setup ranges |
| Landing-page builds | Varies with the pages built | Hidden from the management fee; often billed extra at $1,000 to $5,000 per page | Velocity PPC pricing guide |
Read that like a timesheet. The first two rows contain work that used to expand with spend: more queries to sift, more bids to check, more numbers to carry into a report. Smart Bidding handles the per-auction math. There is still work to do, but a larger budget does not automatically mean proportionally more human bid checks. The third row is different. Someone still has to decide which offer deserves budget, whether the landing page matches the search, and whether a conversion is worth buying again. That is the work a strategist must earn a fee for.
The cheapest mistake is skipping the tracking audit to save on setup. Bad conversion data makes every later decision harder. The most expensive is paying a percentage for mechanical work indefinitely, especially when the fee rises just because the budget does. At the other end, 39% of advertisers spend under $2,500 a month on ads. At $2,500, a 15% fee is $375, while a small account can still need 8 to 12 hours a month. The agency either sets a minimum or leaves work undone. Neither problem is solved by printing “15%” on the proposal.
Prediction 1: flat fees lead sub-$100k proposals
By December 31, 2027, most managers serving accounts spending $10k to $50k a month will lead with a flat fee, not a percentage. In that band, automated execution weakens the connection between spend and management hours. An automatic fee increase when a client raises its budget starts to look less like payment for work and more like a tax on the budget.
The alternative is already on rate cards. Agencies that quote 12 to 20% also publish flat retainers of $1,500 to $10,000 and hourly rates of $100 to $150. Flat-first shops such as SaaSHero start at $1,250 a month for up to $10k in spend, with 30-day exits. Those offers do not prove that flat fees have won. They prove agencies can already sell management without a percentage attached. I covered the competing structures in my comparison of five agency pricing models.
What would prove me wrong: late-2027 agency RFP roundups still show more than half of sub-$100k proposals led by a straight percentage, with no flat alternative. Until then, do not sign a fee that rises automatically with spend unless the scope rises with it.
Prediction 2: minimums rise while small accounts leave the percentage behind
By mid-2027, the typical agency minimum for full management will reach $2,500 a month or require $10k in spend, up from $1,500 today. This is arithmetic, not a sudden outbreak of greed. A $375 percentage fee does not comfortably buy 8 to 12 hours of account work. Agencies need a floor even if the proposal leads with a percentage.
Rate cards show the split: freelancers at $500 to $3,000 versus agencies at $1,500 to $5,000, with setup at $500 to $3,000 and 3- to 12-month lock-ins. Beneath some retainers, white-label fulfillment costs $200 to $1,000 a month. That gap does not mean every agency is overcharging. It means a buyer should ask which parts of the fee buy execution, which buy senior judgment, and which buy a contract they cannot easily leave.
What would prove me wrong: 2027 rate surveys show entry minimums falling toward $1,000 with no spend floor. If you spend under $10k, ask what the minimum buys and what happens when you cancel. The percentage alone will not answer either question.
Prediction 3: spend-tiered software adds flat plans or loses accounts
By the end of 2027, the best-known optimization tools that price by spend tier will either publish flat plans or keep losing mid-market accounts to flat execution. Software changes the hours required to manage an account. If its bill still climbs every time the budget crosses a threshold, the buyer gets the old pricing problem through a new dashboard.
Optmyzr’s listed tiers climb from $208 a month up to $25k in spend, to $416 up to $100k, to $832 up to $500k. Users report 8 to 12 hours a week saved, alongside a 4- to 6-week learning curve and 15 to 25% extra for seats and support. The saved hours matter. So does the time someone still spends learning and operating the tool. I used to recommend that trade without putting enough weight on what happened at the next spend tier.
The alternative is flat execution without the spend ratchet. groas runs Search Ads for $1,499 a month regardless of spend, with no cut of ad spend and no setup fee. The fee crossover math at $5k, $20k, $50k and $100k shows why that distinction matters more as budgets grow. A tool can still be useful; usefulness does not explain why its fee must rise with media spend.
What would prove me wrong: spend-tiered tools hold their share through 2027 without adding flat, unlimited-spend tiers. Watch the pricing page, not the automation pitch.
Prediction 4: revenue-share pricing stays niche
Through 2028, fewer than 15% of Google Ads engagements will price primarily on a share of revenue or profit. I want this one to be wrong. Pay for results sounds cleaner than pay for hours. Then someone has to define the result when a sales cycle runs 90 days, offline conversions arrive late, or tracking breaks. The argument over attribution can consume the alignment the contract was supposed to create.
UK practitioners describe performance pricing as the least-used model because setting a bonus that motivates the agency while preserving client ROI is difficult. A hybrid can carry a similar problem: a base fee plus around 8% on top still rises with the metric that triggers that percentage. Changing the label does not remove the incentive question.
What would prove me wrong: holdout-tested, pay-on-pipeline deals become a standard offer-sheet item by late 2027 rather than a custom enterprise contract. If you want a bonus in the meantime, tie it to closed-won revenue both sides can see in the CRM, not merely to ad-platform conversions. Do not make the bonus a percentage of spend the manager also controls.
Prediction 5: the named strategist becomes a separate line item
By December 31, 2028, the standard proposal will show two lines: a flat platform fee for execution and a separate fee for a named strategist. The first pays for continuous pacing, mining and account changes. The second pays for decisions the machinery cannot settle on its own: which offer to push, which page to build, and when to stop funding a campaign that converts but never closes. Separate the lines and the buyer can judge each on its merits.
The pressure to do so is visible in today’s bundled prices. SaaSHero’s worked example puts $50k in spend at a 15% fee at $7,500, against a $4,500 flat fee: a $3,000 monthly gap. It also shows how a two-month exit penalty at the percentage price reaches $15,000. That is a lot to pay without knowing exactly what belongs to execution, what belongs to strategy, and who is accountable for either. Unbundle the work and the strategist has to earn the retainer every month.
What would prove me wrong: 2028 proposals still bundle everything into one percentage, with no named owner and no separate strategy scope. Ask who owns your account by name and how many other accounts they own. A strategist with 10-plus years on accounts like yours and a weekly action log may be worth a retainer. A percentage with no name on it is worth a renegotiation.
The scoreboard I will use
I have been wrong before. This time I am putting the checks in one place so I cannot quietly swap definitions in 2028. Count proposals, rate cards and pricing pages, not opinions. The predictions above do not all mature on the same date:
- December 2027: Do most $10k-to-$50k account proposals lead with a flat fee? Do more than half of sub-$100k proposals still lead with a straight percentage and no flat alternative?
- Mid-2027: Does the typical full-management minimum reach $2,500 a month or a $10k spend floor, or fall toward $1,000 without one?
- End of 2027: Have spend-tiered optimization tools added flat, unlimited-spend plans, or held share without them?
- Through 2028: Are revenue- or profit-share engagements still below 15%, or have pay-on-pipeline deals become standard?
- December 2028: Do standard proposals name a strategist and price strategy separately from execution?
If three of the five go the other way, I misread the break. I will write the retraction myself.
What to do with your contract this quarter
If you are comparing Google Ads management prices now, act like the split has already happened even if your shortlist has not. Refuse a 12-month percentage that rises automatically with spend. Ask for month-to-month terms, a flat cap on execution, a named strategist, and a weekly log of changes with the reason beside each one. You should be able to tell what you are buying without reverse-engineering a timesheet.
At $20k a month in spend, a 15% management deal costs $3,000 whether the manager spends ten minutes or ten hours checking bids. A flat $1,499 fee with no cut of spend and no setup charge leaves roughly $1,500 a month compared with that percentage deal. That money could go toward pages and creative instead of an automatic fee increase. It is the trade I wish someone had shown me when I was sending the 1am negative lists.
Sign the shortest term that still gets you senior attention, and put any bonus on closed-won revenue visible in your CRM. If I am right, you will have moved before the proposals do. If I am wrong, you will still have kept the exit open. Check me in December 2028. I will be here with the scoreboard.