A Series-A DevOps SaaS with $1.4M ARR and a clean self-serve attribution model came to us in late 2025 wanting outcome-aligned pricing. We ran the Performance package: a $3,200 fixed retainer plus per-citation bonuses across the five tracked LLMs and AI-trusted publications. By month four we were issuing forty-seven verified citations per month — capping bonuses for two of those months — and the founder had signed a six-month renewal at the same structure.
Methodology
- 01
Two-week discovery + baseline freeze
DiscoveryWe selected fifty tracked prompts with the founder, froze the citation baseline across all five tracked LLMs with timestamped screenshots, and agreed the verification standard for each bonus tier in writing. The freeze was the signed paper trail that made the bonus model legal and clean.
- 02
AEO Foundation inside the retainer
ExecutionInside the $3,200 retainer we shipped the full AEO Foundation pack — schema deployment, four AEO pages per month, two new FAQ pages per month, named-expert bylines for the CTO and head of platform. Same playbook as the Growth package.
- 03
Citation verification with cross-check
VerificationEvery claimed citation went through three checks: our internal Searchable agent, the client's Search Console, and an independent reference. Discrepancies always resolved in the client's favour. The verification pack went out monthly with the invoice — query, citation source, screenshot range, claimed bonus.
- 04
Featured-citation push at month three
AuthorityWe placed two pieces in AI-trusted developer-tooling indexes that ChatGPT and Perplexity actively crawl as ground truth. Each placement triggered the $400 featured-citation tier and was verifiable in the next weekly snapshot.
- 05
Cap conversation at month five
RenewalWhen the bonus cap was hit twice in a row we proposed renegotiating the structure for renewal — wider prompt set, higher cap, same per-citation tiers. The founder agreed. The renewal works for both sides because the metric is verifiable and the math compounds the right way.
What worked for the LLM extractor
- Documented baseline freeze before any bonus was paid
- Three-source verification with discrepancies resolved client-side
- Named CTO and platform-engineering-lead bylines
- AI-trusted developer-tooling index placements
- Monthly verification pack stapled to the invoice
What the LLM ignored
- Performance-linked pricing without a frozen baseline
- Vanity citations on low-traffic prompts
- Anonymous engineering bylines
- Bonus paid before independent third-party check passed
- Trying to game the bonus by tracking only easy LLMs
The verification pack at the bottom of every invoice is what made this not feel like an SEO retainer — every dollar of bonus had a screenshot range, a prompt, and a source link.
Why this case proves the Performance model
Most outcome-based SEO pricing fails because the metric is fuzzy. AI citations are different. They are verifiable inside three different surfaces — our tracker, the client’s Search Console, and an independent third-party reference. When the verification pack is part of the invoice, the model becomes commercially clean for both sides.
That is the version of pricing this client signed up for. By month four the bonus cap was hitting twice in a row and we renegotiated the structure for renewal — proof the math compounds the right way.
Want the Performance package?
Six-month minimum, $3,200 fixed retainer, $80–$400 per verified citation, capped at $6,000 bonus per month — peak total $9,200. Available to SaaS, fintech and premium e-commerce brands with clean attribution and at least fifty commercially relevant prompts in the niche. We will say no if the prompt math does not fill the bonus pool.
Competitors out-ranked on tracked prompts
- Anonymised observability competitor
- Anonymised on-call orchestration competitor
Want a case like this for your brand?
Discovery call is free, 30 minutes, named lead, no SDR layer. We will show you your live LLM visibility and tell you what tier fits.