Knowledge · Conversion Engineering
Conversion Engineering vs CRO Agency: Why We Don’t Sell Retainers
Most CRO “agencies” are retainer factories. $4K-$8K/month for a monthly deck, a couple of color tests, and a quarterly executive readout. The model exists because agencies need recurring revenue, not because mid-market sites need monthly tests. We sell two things instead: a $500 audit and a $3,500 sprint. Here’s the math on why that’s the honest model.
№ 01The CRO retainer math (why it doesn’t work)
A typical mid-market CRO retainer is $5K/month, $60K/year. To pay back, the retainer needs to lift conversion enough to generate >$60K in additional pipeline. On a site doing 200 leads/month at a 20% close rate and $25K average deal size, that’s $1M/month in pipeline. Lifting conversion 6% on that math produces $60K — just enough to break even.
Most mid-market sites can’t physically run enough properly-powered tests in a year to produce 6% sustained lift. The retainer math assumes 12 tests/year landing 0.5%+ lift each. The actual data: most mid-market retainers produce 2-4 statistically-significant lifts/year, often offsetting each other. The retainer pays the agency; it rarely pays the client.
№ 02The audit-and-sprint model
Our model: Conversion Audit ($500, 5-day delivery, refundable) identifies the 5-10 highest-leverage changes on your site. CRO Sprint ($3,500, 14-day delivery) implements the top 5. Total: $4,000 for shipped changes that move numbers. No monthly commitment, no recurring fee, no scheduled deck.
Why this works: the diminishing-returns curve. The first 5 changes on a typical mid-market site account for 70-80% of the available conversion lift. The next 20 changes account for the remaining 20%. After the sprint, the math on more work usually doesn’t pay back inside 6-12 months. Better to wait, let data accumulate, and run a second sprint later.
№ 03When a retainer DOES make sense (rare)
The CRO retainer math works when: site traffic is 100K+ sessions/month (you can run continuous tests), you have a defined product roadmap that requires CRO coordination (SaaS releases), or the ACV is high enough ($100K+) that a 1-2% lift pays back the retainer. None of those describe most mid-market B2B service sites.
If you fit that profile, we’ll set up a $2,500/mo CRO Partner Retainer. But we’ll tell you out loud whether the math actually supports it; most clients walk away with the audit + sprint and a healthier P&L.
№ 04What we won’t do
The retainer offer we don’t sell, even when asked:
- Monthly heatmap reports. Heatmaps go stale slowly. A monthly review produces 10 hours of agency time and 2 useful observations. We do quarterly heatmap reviews if you’ve added new pages, not monthly.
- Continuous A/B testing on under-30K-session sites. The math says you can’t. We won’t pretend.
- “Strategic CRO consulting” calls. Calls don’t ship changes. We bill for shipped work, not for talking about shipping work.
№ 05The handoff after the sprint
At sprint completion, you get: documented event spec in Google Sheets, GTM container with all tracking wired, GA4 explorations preset for the metrics that matter, a 1-page “when to call us back” sheet (what data thresholds trigger another sprint), and 30 days of monitoring support.
After 30 days, you’re self-sufficient. Most clients return for a second sprint 6-12 months later when traffic has grown enough to justify the next round of testing. That’s the cadence the math supports; we follow the math.
⚠What to avoid
- Buying a CRO retainer for a 5K-session/month site. The traffic math doesn’t support monthly testing. You’re paying for activity, not outcomes.
- Hiring a retainer to “run experiments.” If the retainer can’t commit to specific shipped changes, you’re funding research, not optimization..
- Renewing a retainer that produced 2 wins in 12 months. The break-even bar is higher than 2 wins/year for $60K. Either renegotiate the model or end the engagement.
Related questions
Go deeper
Three Ways to Start · No Sales Pitch
Want this analyzed on your site?
$500 audit. 5-day delivery. Refundable on engagement.