Most automation platforms charge per operation: every API call, every webhook processed, every message sent. It’s a clean model for the vendor — revenue scales with usage regardless of whether that usage produces any value.
The problem is incentive alignment.
When you pay per operation, your vendor’s job is to process as many operations as possible. When you pay for outcomes, your vendor’s job is to deliver results. These are not the same thing.
The Per-Operation Problem
Imagine a skill that fires on every Shopify order — 10,000 orders per month — and does a cursory inventory check that almost always returns “in stock.” You’re paying for 10,000 operations. The skill deflects maybe 80 tickets. Your cost-per-deflection is enormous.
Nobody in a per-operation model is motivated to fix this. The vendor wants high operation counts. The customer doesn’t have easy visibility into cost-per-outcome until they do the math themselves.
How Clarissi’s Model Works
Clarissi uses a hybrid pricing model:
Base retainer — covers your dedicated operator’s time (~10 hours/month), platform access, and skill infrastructure. This is a flat monthly fee, not tied to volume.
Outcome fee — $2–3 per ticket deflected above your pre-activation baseline. Calculated monthly from your execution log. Billed via invoice.
The outcome fee is the alignment mechanism. If skills fire 10,000 times and deflect 10 tickets, you don’t pay an outcome fee on the other 9,990 executions. You pay for the 10 deflections.
This means Clarissi is motivated to:
- Tune skills to maximize real deflection, not raw execution count
- Minimize misfires (executions that don’t produce deflections)
- Expand to additional skills where deflection opportunity is highest
How the Baseline Works
Before any skill activates, your operator establishes a 90-day baseline: your average monthly ticket volume per category, pulled from your Gorgias or Zendesk history.
“Tickets deflected above baseline” means: if you normally receive 200 backorder tickets per month and you receive 80 after activation, that’s 120 deflections. Those 120 are what you pay the outcome fee on.
The baseline is set once and is not adjusted for seasonal variation in the first year. If your ticket volume naturally drops in Q4 (because fewer orders, fewer problems), you still pay only on deflections above that original baseline — which means we’re conservative about what counts.
What If We’re Not Delivering?
If skill misfires are high, if deflection rates are low, if the numbers don’t justify the fee — your operator should be telling you that in the monthly report, not hoping you don’t notice.
In practice, the outcome fee structure means your operator has a direct stake in the quality of each execution. Tuning is not a nice-to-have. It’s what protects the fee.
If after 90 days the outcome numbers don’t justify continuation, you cancel with 30 days’ notice. No long-term lock-in, no multi-year commitment required to get our attention.
Why We Think This Is Better
For customers: you can calculate exactly what Clarissi is worth in dollar terms. Deflections × your cost-per-ticket gives a value figure. The outcome fee should be a fraction of that figure. If it’s not, the model doesn’t work and you should cancel.
For us: outcome fees force us to build skills that actually deflect tickets, tune them continuously, and expand to high-deflection opportunities. We’re not rewarded for volume — we’re rewarded for results.
That’s the alignment we wanted to build from the start.