The agency operator's compensation trap.

The canonical agency bonus structure for an operator is some version of: “X% of the performance gain this quarter vs baseline.” This looks great on paper. It supposedly aligns the buyer with the client and creates upside for excellent work.

In practice, here's what we've watched happen repeatedly at 7-figure monthly spend:

1. Buyers chase short-term lifts at the expense of account health.A pause of a fatiguing winner, which would reset the account's trajectory long-term, hurts this quarter's bonus. So the winner keeps running past the point of incremental return. The account erodes. The buyer's bonus is fine.
2. Buyers hoard winning accounts and avoid hard accounts. When a new problem client comes in, the senior buyer makes sure it goes to a junior. When a long-time client is struggling, the buyer who has the account spends less and less time on it because the bonus math is against them. The client feels the neglect.
3. Buyers don't share learnings across the team. A tactic the senior buyer found on their account is worth real bonus dollars. Sharing it with the rest of the team dilutes that edge. Information silos form. The agency loses the compounding value of a learning team.
4. Creative velocity decisions get distorted. Testing new creative costs money that hits this quarter's bonus. Sticking with known winners preserves the bonus. Buyers under performance comp run less creative volume than buyers not under performance comp. We've watched this happen.

None of this is because buyers are greedy or unprofessional. It's because they're responding rationally to the incentives the comp structure puts in front of them. Every comp structure produces the behaviors it rewards. If you're getting the wrong behaviors, look at the structure first.

When agencies design performance bonuses, they usually peg them to one of: platform ROAS, platform CPL, or platform-reported conversions. All three are the wrong metric to incentivize on.

Platform metrics are what the platform says happened. As we've written elsewhere, at high spend those numbers diverge from CRM reality. A buyer who optimizes to platform ROAS and hits the bonus while CRM revenue is flat has delivered exactly what the comp structure asked for and nothing the client actually wanted.

The metrics that actually correlate with client value — and therefore with agency retention — are:

These are harder to measure in real time than platform metrics. That's exactly why agencies default to platform metrics for comp. The metric that's easiest to track is the one that gets used, even when it's not the one that should.

The fix isn't to abandon performance comp entirely. The fix is to track the right metrics, which usually means investing in the reconciliation infrastructure we've written about in other posts. A comp plan based on CRM revenue takes a monthly CRM export and a weekly reconciliation job. Most agencies don't have the pipeline. The ones that do can build compensation plans aligned with client value.

Here's a table of rough base-pay numbers for performance-marketing buyers in 2026 US markets, based on offers we've seen or extended over the last two years:

That distribution is rough. The top of each range is usually available only at larger agencies or in-house teams at spenders. Smaller agencies pay closer to the bottom of each band.

Here's the weird part. A senior buyer at $120K is managing on average 4–8 clients in an agency context. If those clients average $500K/month in spend, the senior buyer is directly responsible for $24M–$48M of annual client spend. On a performance basis, they're producing (or failing to produce) somewhere between 30–80% of that back as client revenue, depending on vertical. That's a buyer controlling $7M–$30M of client revenue being paid $120K.

The in-house equivalent role at a $50M revenue ecom company — the VP of Growth or Director of Paid Media — would be paid $180K–$250K plus equity. They're controlling the same spend. The gap is real.

The industry-wide reason for the gap is agency margin compression. Management fees at 10–15% of spend don't leave room to pay in-house-equivalent salaries. The agency absorbs the margin pressure by paying buyers below in-house-equivalent rates and accepting the flight risk that produces.

What the flight risk actually costs

When a senior buyer leaves, the agency bears:

1. Client flight risk.Clients attach to buyers, not to agencies. A client whose senior buyer leaves is meaningfully more likely to churn in the following 90 days. We've seen 20–40% client attrition rates following a senior buyer departure on a specific book.
2. Onboarding cost for the replacement. A new senior buyer needs 60–90 days to get up to full productivity on a complex book. During that window, account performance typically dips 10–20%. That dip is a real cost to the client and a real risk to retention.
3. Institutional knowledge loss.The patterns the buyer learned — what worked on this audience, what the client's offer constraints were, what's been tried — leaves with them if it wasn't written down. Most of it wasn't.
4. Team morale hit. When a respected senior buyer leaves, the next tier of buyers starts updating their LinkedIn. Replacement salary pressure rises.

The fully-loaded cost of losing a senior buyer at a 7-figure- book agency is usually somewhere between $200K and $500K over 18 months when you count client churn and onboarding. That number dwarfs what it would have cost to pay the buyer competitively.

After several years of trying different structures and watching several more at peer agencies, here's the three-layer model that consistently produces durable account health, buyer retention, and client trust:

Layer 1 — Competitive base, scaled to book size.

Base pay is indexed to the monthly spend the buyer manages, not to years of experience alone. A buyer managing a $2M/month book gets paid more than a buyer managing a $500K/ month book at the same seniority level. Reasoning: the complexity, the decision-weight, and the client expectation all scale with book size.

Rough scaling we've used: add 15–25% to base for each doubling of managed monthly spend, capped at the top of the senior band. A buyer managing $250K/month on base of $110K scales to $130K if they're promoted onto a $500K/month book, $150K at $1M/month. The client contract has to support it; see Layer 3 on contracts.

Layer 2 — Quarterly portfolio health bonus.

Bonus is paid quarterly against metrics the client and agency both care about and can both verify:

1. CRM-back CPL improvement quarter-over-quarter. Not platform CPL. CRM-reconciled. Measured against the previous quarter, adjusted for seasonal baseline.
2. Creative runway maintained at 3+ weeks. Measurable. Visible. Drives the volume behavior you want.
3. CAPI quality above 8/10 EMQ on every active account in the book. Another measurable system- health metric that correlates with performance.
4. Client NPS or quarterly QBR rating. Not a personality vote; a specific question about whether the client feels the buyer is driving the account forward. Requires actually asking it.

Bonus is capped. Annualized, it should be 20–30% of base at target. We've seen comp plans that cap at 80% of base and they produced the problematic behaviors we described earlier. Capping bonus keeps the rewarded behaviors balanced against all the other things the buyer needs to do.

Layer 3 — Annual retention bonus.

Every buyer who stays a full year gets an additional bonus paid in Q1 of the following year, equal to 10–15% of annualized base. It's paid for the simple act of staying on the team and doing the work consistently for 12 months.

This sounds like “paying people to not quit,” and in a sense it is. The point is to reward exactly the behavior that's most valuable: buyers who stay on books long enough to accumulate institutional knowledge about specific client audiences, offers, and patterns. That compounding knowledge is what differentiates a 3-year senior buyer from a 6-month senior buyer. The comp structure should acknowledge it.

The economics of the structure

Total compensation at target for a senior buyer under this structure: base $140K + ~$35K quarterly bonuses + ~$20K retention bonus = $195K/year. For a buyer managing a $1M/month book, that's 1.6% of managed annual spend. Against a 12% management fee, it leaves 10.4% of spend for everything else the agency does — creative production, account management, reporting infrastructure, tooling, and profit.

That math works if the client contract pays against spend. It doesn't work if the contract is flat-fee and spend doubles mid-year. Which brings us to the final point.

You cannot build a sustainable comp plan on top of a broken client contract. Every comp structure is downstream of the revenue structure. If your retainer is flat-fee with no performance component, you can't afford to pay performance bonuses. If your retainer doesn't scale with spend, you can't afford to scale base pay with book size.

The client contract structures we've seen that support healthy buyer compensation:

1. Percentage-of-spend management fee, tiered. 12–15% on the first $500K/month, 8–10% on the next $500K/month, 5–7% above that. Rewards the agency for scaling the client; gives the client a declining cost curve as spend grows.
2. Flat management fee plus performance bonus. Agency charges a flat monthly fee that covers the buyer's base-equivalent time, plus earns a bonus tied to specific performance thresholds (CRM-back CPL below X, quarterly revenue growth above Y). The agency's bonus directly funds the buyer's bonus.
3. Retainer plus creative line item. Separates creative production from management fee, removing the perverse incentive to under-invest in creative. The retainer covers the buyer; creative is billed at cost-plus or per-unit.

The structure that kills comp sustainability: flat monthly retainer with unlimited scope, covering everything from media buying to creative to reporting to strategy. Agencies sign these to win logos, then discover six months later that the client is spending 2x the original budget and expecting 2x the service on the same fee. Buyer gets overloaded. Bonus gets squeezed. Retention suffers. Client ultimately churns because the service quality dropped.

If your comp structure feels misaligned and you're running high-spend books:

1. Audit the behaviors your comp plan is producing. Are buyers hoarding? Siloing learnings? Running less creative volume than they should? Skipping hard conversations with clients? If any of those are happening, it's the comp plan, not the people.
2. Check what metrics you're paying on. Platform ROAS and platform CPL drift from CRM truth at scale. If you're bonusing on those, you're rewarding buyers for optimizing to distorted numbers.
3. Benchmark base pay against managed book size.A senior buyer on a $1M/month book paid the same as one on a $200K/month book is a flight risk you haven't noticed yet.
4. Move to a three-layer plan: competitive base, capped quarterly bonus on right metrics, annual retention bonus. Expensive on paper; cheaper than replacing buyers every 18 months.
5. Fix the client contract first.You can't pay a buyer 1.6% of managed spend out of a flat fee that doesn't scale with spend. The contract structure determines which comp plans are possible.

The short version: compensation is strategy at an agency. It's the lever that decides whether you keep the people who keep the clients. At 7-figure monthly books, the cost of getting it wrong is much larger than the cost of paying above the market. Most agencies are running comp structures designed for smaller books and absorbing the cost of that mismatch without recognizing it.