Imperfect real data over perfect claims.
Case studies are how we prove — or fail to prove — that distribution works. We run them on our own product portfolio first, and we publish the numbers even when they aren't flattering. Here's exactly what we measure and the rules we hold ourselves to.
What we track per product
Distribution
- Starting position + existing listings
- Directories attempted → approved → indexed
- Time required + human interventions
- Languages used
Outcomes
- Referral visits, signups, qualified leads
- Revenue attributed (conservatively)
- Brand-search change + impressions
- AI-search citations, where measurable
Efficiency
- Cost per approved listing
- Cost per acquired customer
- Results at 7 / 30 / 60 / 90 days
The rigor rules
Correlation is not causation
A ranking or traffic change that happens after a distribution run is not proof the run caused it. We say what we can attribute, what we can't, and what other factors were in play over the same window.
No blanket attribution
We do not credit all traffic, signups, or revenue growth to StartupAmplify. Attributed numbers are the conservatively-measured slice we can actually tie to referrals or listings — not the whole trend.
Real, dated, imperfect
Imperfect real data is more citable than perfect claims. A published case study shows actual numbers with dates and screenshots, and states its limitations — including a small sample or an inconclusive result.
The bar to publish
A case study ships only with real, dated numbers, evidence (screenshots), and an explicit limitations section. Testimonials and customer results follow our customer-evidence policy — nothing gets published that we can't verify. Aggregate structural findings feed the benchmark.