Google Ads management fees: when 15% of spend costs more than a flat retainer

Every guide to Google Ads management pricing tells you the same three things: agencies charge 10% to 20% of ad spend, or a flat retainer of roughly $1,500 to $5,000 a month, or some hybrid of the two. All of that is true and none of it answers the question you actually have, which is whether the fee in front of you is worth paying at your spend level. That is an arithmetic question, and it has an answer.

What Google Ads management actually costs in 2026

Published agency rate cards cluster tightly, and the pattern is consistent enough to treat as a market rate rather than a guess. Three structures dominate.

A fourth model, performance-based pricing tied to qualified leads or ROAS, exists but is rarer and harder to compare, mostly because the definition of a qualified lead is negotiable and the agency that writes the definition has an interest in it. It is worth asking for, and worth reading carefully.

The crossover table: where a percentage stops being cheaper

Because a percentage fee rises with spend and a retainer does not, every pairing has exactly one spend level at which the two cost the same. Below it the percentage wins, above it the retainer wins. Here is that number for the retainers you are most likely to be quoted.

Read a row like this. If you are quoted a $2,500 retainer and a 15% alternative, the retainer is the cheaper option from $16,700 of monthly spend upwards, and the percentage is cheaper below it. If your spend is planned to double over the next year, the crossover point is where you should be looking, not today's number.

The same arithmetic disposes of the hybrid quote quickly. A $1,000 base plus 5% of spend costs exactly the same as a straight 15% fee at $10,000 of monthly spend. Under that figure the hybrid is the more expensive of the two; over it, the cheaper. Ask for both quotes and you can work out in one line which side of the line you are on.

The test the pricing guides skip: does the fee earn itself back?

Choosing the cheaper fee model is the small question. The larger one is whether paying any fee beats running the account yourself, and that has a clean answer.

Your real cost per conversion is not what Google Ads reports. It is your ad spend plus the management fee, divided by conversions. So the fee breaks even at the point where the managed account's all-in cost per conversion matches what you were achieving alone. Write that out and the fee cancels into a single ratio:

Expressed as a cost-per-conversion reduction rather than a volume increase, the same requirement looks slightly gentler, because the fee is spread over more conversions. Both columns below describe the identical break-even point.

The bottom two rows are the ones worth sitting with. At $3,000 of monthly spend against a $1,500 minimum fee, a manager has to cut your cost per conversion by a third before you have gained a cent. That is not impossible on an account that has never been touched, and it is close to impossible on an account that is already competently run. This is the honest reason minimum fees and small budgets sit badly together, and it has nothing to do with the agency's quality.

The same table also explains why a good agency on a large account is easy to justify. At $50,000 of spend a 12% fee needs a 12% conversion uplift, which a single well-executed structural change can deliver, and the account is complex enough that such changes exist to be found.

The floor underneath all of this: does your spend generate enough data?

There is a second constraint that decides whether management can help at all, and it is set by Google's own bidding systems rather than by the agency. Smart Bidding strategies need conversion volume to learn. Google's own Target CPA and Target ROAS documentation requires sufficient recent conversion data without committing to one universal figure; the working minimums cited consistently across practitioner guidance are around 30 conversions in the trailing 30 days before Target CPA is worth using, and around 50 for Target ROAS, which has to learn value as well as count.

Turn that into a spend floor and it becomes concrete. Multiply your current cost per conversion by 30, or by 50 if you want value-based bidding.

Now put the two constraints together, because they bite from opposite directions. A business with a $150 cost per conversion needs roughly $4,500 a month for Target CPA to have data to work with. A $1,500 minimum fee on that spend is 33% of it, so the manager must cut cost per conversion by about a quarter to break even. Both things are true at once: the account is barely large enough to bid automatically, and the fee is large enough relative to spend to be hard to justify.

That combination is a signal, not a dead end. Below the crossover, the money is usually better spent on fixed-scope work than on a monthly fee: a one-off build and tracking implementation, a conversion-tracking audit, a landing page that actually converts the traffic you are already buying. Ongoing management earns its keep once there is enough spend and enough conversion data for optimisation decisions to be based on evidence rather than on hunches. If your constraint is lead quality rather than lead volume, the more useful model to work through first is cost per qualified lead, because a cheaper form fill that never closes is not a saving.

What the fee does and does not include

Two quotes at the same number are rarely the same purchase. The gap is almost always scope, and these are the line items that most often sit outside a management fee and get billed separately or not done at all.

1. Conversion tracking setup and repair. Including server-side tagging, consent-mode configuration, and importing offline conversions from your CRM. If this is excluded, everything downstream is being optimised against numbers nobody has verified.
2. Landing pages. Whether the fee covers building or only advising on them. Traffic sent to a page nobody is allowed to change is the most common reason a competent account still underperforms.
3. Creative. Image, video and Demand Gen assets. Increasingly the binding constraint on paid social and Demand Gen performance, and frequently excluded.
4. Shopping feed management. For retail accounts this is a large share of the actual work, and a feed nobody owns will quietly degrade.
5. Additional channels. Whether the fee covers Google only, or Microsoft Ads, Meta and the rest. Ask whether the percentage applies per channel or across total spend.
6. Measurement beyond the platform dashboard. Incrementality tests, geo holdouts and blended reporting. Platform-reported conversions are the agency's own scorecard, which is a reason to ask who else is marking the work.
7. Named hours and escalation. Who actually works on the account, how many hours, and who you reach when spend spikes on a Saturday.

Ask for the exclusions list explicitly rather than the inclusions list. Inclusions lists are marketing documents; exclusions lists are where the second invoice lives. The same asymmetry applies to website work, which we covered in retainer versus project pricing for web agencies.

Three questions that separate the field faster than the fee does

Once you have worked out which model is cheaper at your spend, the fee stops being the interesting variable. These do more work.

1. "Under your percentage model, what happens to your fee if you cut my wasted spend by 20%?" The honest answer is that it falls. An agency that acknowledges that plainly and explains how they handle it is telling you something useful; percentage-of-spend structurally rewards larger budgets and penalises efficiency work, and the good ones know it.
2. "What will you report that Google Ads does not already tell me?" If the answer is a reformatted platform dashboard, you are paying for reporting rather than for analysis. The useful answer involves your revenue data, not just the platform's conversion column.
3. "What is the first thing you would change, and why that first?" This tests whether they have looked at your account or at your industry. It also tells you whether the first month is going to be setup or optimisation, which changes when the fee starts earning back.

If the account is currently built around one campaign type and nobody can explain why, the answer to the third question often lives in the Performance Max versus Search allocation decision, and it is worth asking them to show their reasoning there.

How to use all of this in one sitting

Take your planned monthly ad spend for the next twelve months, not last month's. Run it through three steps.

1. Find your crossover. Divide each flat retainer you have been quoted by each percentage. That gives you the spend at which they are equal, and tells you which quote is cheaper at your number.
2. Check the break-even bar. Express the fee as a percentage of your spend. That percentage is the conversion uplift the fee has to deliver before you gain anything. Decide whether you believe it is achievable in your account, given how well it is run today.
3. Check the data floor. Multiply your cost per conversion by 30. If your spend is well below that, buy fixed-scope work such as tracking, landing pages and account structure instead of ongoing management, and revisit when spend clears the floor.

None of those three steps needs a call with an agency to complete, which is the point. Walking into the conversation already knowing your crossover point and your break-even bar changes what you are negotiating about, from a monthly number to a scope and a target. For the wider question of what your budget should be before you worry about who manages it, the scaling spend without wrecking CAC and industry ROAS and CAC benchmarks pieces are the better starting points, and our performance marketing page sets out how we structure the work itself.

Frequently asked questions

How much does Google Ads management cost in 2026?

Published 2026 agency pricing clusters into three models. Percentage of ad spend usually runs 10% to 25%, tiered so that smaller accounts pay a higher percentage: roughly 20 to 25% under $5,000 of monthly spend, 15 to 20% between $5,000 and $20,000, and 10 to 15% above $20,000, with minimum fees commonly starting around $500. Flat retainers run about $500 to $2,500 a month for an experienced freelancer, $1,500 to $5,000 for a small or mid-size agency, and $5,000 to $10,000 or more for large or complex accounts. Hybrid models combine a base fee with a smaller percentage, commonly structured as around $1,000 plus 5% of spend.

Is a percentage of ad spend or a flat fee cheaper?

It depends on one number. A percentage fee and a flat retainer cost the same at exactly one spend level, found by dividing the retainer by the percentage. A 15% fee equals a $1,500 retainer at $10,000 of monthly spend, equals a $2,500 retainer at $16,700, and equals a $4,000 retainer at $26,700. Below the crossover the percentage is cheaper; above it the flat fee is. Because the crossover moves with your spend plan, compare at your projected annual spend rather than last month's.

How do I know if a Google Ads management fee is worth paying?

Express the fee as a percentage of your ad spend. That percentage is the conversion uplift the fee has to deliver, at the same spend, just to break even. A fee worth 15% of spend needs 15% more conversions, which is the same thing as a 13% lower cost per conversion. A $1,500 minimum fee on $3,000 of spend is 50% of spend and needs a 33% cut in cost per conversion before you are level. Judge the fee against how well the account is already run: an untouched account may have that much slack, a competently run one usually does not.

What is the minimum ad spend that justifies hiring an agency?

There is no single figure, but there is a calculable floor. Smart Bidding needs conversion volume. Google's Target CPA and Target ROAS documentation requires sufficient recent conversion data without naming one universal figure, and the working minimums cited consistently across practitioner guidance are around 30 conversions in the trailing 30 days for Target CPA and around 50 for Target ROAS. Multiply your cost per conversion by 30 to get the spend that produces that data: $1,500 a month at a $50 cost per conversion, $4,500 at $150, $9,000 at $300. Well below that floor, fixed-scope work such as conversion tracking, account structure and landing pages usually returns more than a monthly management fee.

Does percentage-of-spend pricing create a conflict of interest?

Structurally, yes. Under a percentage model the fee rises when your budget rises, whether or not the extra spend was profitable, and it falls when the agency cuts wasted spend. A campaign spending $50,000 at 2x return produces the same fee as one spending $50,000 at 4x. That does not make percentage pricing wrong, and it is often the cheaper option for smaller accounts, but it means budget-increase recommendations come from an interested party. A flat retainer removes that specific conflict; performance-based pricing shifts it to how the performance metric is defined.

What is usually excluded from a Google Ads management fee?

The most common exclusions are conversion tracking setup and repair including server-side tagging and CRM offline conversion imports, landing page design and build, image and video creative, Shopping feed management, channels other than Google, and measurement beyond the platform dashboard such as incrementality or geo holdout tests. Ask for the exclusions list rather than the inclusions list, because two quotes at the same monthly number frequently describe very different amounts of work.

The takeaway

The three fee models are not competing philosophies, they are three lines on a graph that cross at a spend level you can calculate in one division. Work out your crossover point, then apply the harder test: a fee worth 15% of your ad spend has to produce 15% more conversions at the same spend before it has paid for itself, and a minimum fee on a small budget can require a third off your cost per conversion just to reach level. If your spend sits below the volume Smart Bidding needs to learn from, buy fixed-scope tracking, landing page and structure work instead, and revisit ongoing management when the numbers support it.