Problem #1
Your ROAS may look fine while the account is still training Google on the wrong signals.
In e-commerce, weak Google Ads performance can hide behind acceptable ROAS. The account can quietly shift spend toward low-margin SKUs, branded demand, weak feed segments, returning customers, or products already limited by stock, price, or landing-page friction.
Margin-blind scaling
The account may spend more on products that convert while leaving too little gross profit after ad cost, discounts, refunds, and fulfillment.
PMax hides the real mix
Performance Max can blend branded, non-branded, remarketing, Shopping, Display, and YouTube signals into one good number - making it harder to see what actually created growth.
Feed and stock issues distort decisions
Missing attributes, weak titles, poor product grouping, out-of-stock items, price changes, and promo spikes can make campaign results look better or worse than the actual opportunity.
Tracking quality changes the fee conversation
Before performance-based pricing makes sense, revenue source, attribution rules, exclusions, and conversion signal quality need to be clear.
The Solution: baseline first, scaling second.
We start by checking account structure, tracking, feed quality, product economics, search terms, landing-page readiness, and current ROAS/MER logic. Only after the baseline is clear do we agree which ROAS band, attribution source, and exclusions should guide the management fee.
Separate branded-only uplift from real acquisition growth.
Check whether spend is moving toward profitable product groups.
Review tracking and feed readiness before scaling.
Tie the fee model to agreed rules, not vague platform screenshots.