What's the right way to handle a POC where the customer keeps asking for more features mid-trial?

POC Scope Creep: Setting Guardrails

Feature requests during trials happen. The question is whether you're proving value or building custom. Lock your scope day one — document what success looks like, what's in-bounds, and what gets queued for post-pilot.

Core Moves

1. Reset expectations upfront — Frame the POC as a 30-60 day test of core workflows, not a custom build. Show the customer the scope matrix: what they're evaluating vs. What ships after commercial.
2. Track requests separately — Use a shared board (Monday.com or Salesforce Chatter) to log all asks. Mark them "In Scope" or "Post-POC." Transparency kills scope creep.
3. Schedule weekly reviews — Bring your SE, CSM, and customer sponsor weekly. Review requests together, not via Slack. Rank them against the original success criteria.
4. Say "not yet" clearly — "That's great feedback and we're capturing it for Q3 roadmap" is sharper than silence. Pair it with what *is* working.
5. Hold the line on timeline — POCs end on the date promised, with or without polish. Shipping on time builds credibility more than shipping perfection.

POC Success Metrics (What You Actually Prove)

• Rep workflow — Does the platform cut call prep time by 20%+?
• Lead quality — Are discovered accounts net-new or redundant?
• Adoption — Did 3+ users touch it without handholding?
• Expandable — Are there clear upsell vectors post-pilot?

A POC that proves a small truth beats a POC that half-proves everything.

Reference Models

Tools That Help

Pavilion (POC orchestration) and OpenView's POC playbook template establish what "done" means before day one. Bridge Group research shows 80% of failed POCs had vague success criteria — tighten yours immediately.

Key principle: POCs aren't cheaper than sales. They're faster. Speed is your only leverage — use it.

Anchor Citations

• CB Insights State of Venture / Sales Tech: https://www.cbinsights.com/research/
• Bessemer Cloud Index + State of the Cloud: https://www.bvp.com/atlas/state-of-the-cloud
• Crunchbase News (funding + M&A): https://news.crunchbase.com/
• SaaS Capital industry survey + valuation: https://www.saas-capital.com/research/
• PitchBook venture + private markets: https://pitchbook.com/news
• a16z Marketplace / SaaS frameworks: https://a16z.com/category/saas/

Operator Benchmarks (2025 Data)

The Bear Case (Operational Concentration)

Three concentration risks:

1. Customer concentration — any single >20% of revenue is asymmetric.
2. Channel concentration — 60%+ from one channel is existential.
3. Geographic concentration — NA-centric exposed to NA macro/regulatory.

Mitigation: customer top-1 < 20%, channel top-1 < 40%, geography top-region < 70%.

See Also (related library entries)

Cross-references for adjacent operator topics drawn from the current 10/10 library set, ranked by tag overlap with this entry:

• q1782 — How does Outreach onboarding compare to Salesloft?
• q1723 — What does Datadog churn math look like under AI pressure?
• q1597 — What does Snowflake churn math look like under AI pressure?
• q1253 — How'd you fix Faraday Future's revenue issues in 2026?
• q247 — How do you scale a customer reference program past 10-15 active references without burning out your champions?
• q238 — How do you measure SE (sales engineer) ROI without making them feel like commodities?

Follow the q-ID links to read each in full.