Designing Great Criteria: Required vs. Nice-to-Have

The criteria you define in FounderScan directly shape how startups are evaluated and ranked. Well-designed criteria lead to better selections; poorly designed ones create noise and missed opportunities.

This guide shares what we've learned from working with dozens of accelerators on their evaluation frameworks.

The Purpose of Criteria

Before diving into specifics, let's clarify what evaluation criteria should accomplish:

1. Encode your thesis: Criteria should reflect what makes startups successful in your program
2. Enable consistency: Different reviewers should reach similar conclusions
3. Save time: Good criteria help you quickly identify strong matches and clear mismatches
4. Support decisions: When you pass on a startup, criteria provide defensible reasoning

Criteria are not meant to perfectly predict startup success (impossible) but to systematically identify startups aligned with your program's focus and capabilities.

Required vs. Nice-to-Have

FounderScan distinguishes between two types of criteria:

Required Criteria

These are non-negotiable. If a startup doesn't meet a required criterion, it's likely not a fit for your program regardless of other strengths.

Examples:

• At least one technical co-founder
• B2B focus (if you're a B2B-only accelerator)
• Pre-seed to seed stage
• Full-time commitment from founders
• Operates in your geographic region

Required criteria should be few and clear. If you have more than 4–5 required criteria, consider whether some are actually preferences.

Nice-to-Have Criteria

These are preferences that strengthen an application but aren't deal-breakers. Startups scoring well on nice-to-haves rise above those that merely meet the requirements.

Examples:

• Prior startup experience
• Existing paying customers
• Domain expertise in target market
• Strong demo or prototype
• Warm introduction or referral

Nice-to-haves let you differentiate between qualified candidates.

Criteria Design Principles

Be Specific, Not Vague

❌ "Strong team"
✓ "At least one founder with 5+ years of relevant industry experience"

❌ "Good traction"
✓ "Has acquired 10+ paying customers or achieved $5K+ MRR"

Vague criteria lead to inconsistent scoring. Specific criteria are evaluable.

Focus on Observables

❌ "Founders are passionate"
✓ "Founders have worked on this problem for 1+ years or have direct personal experience with the pain point"

Passion is hard to assess from an application. Observable behaviors and history are not.

Match Your Value-Add

What makes your accelerator special? Design criteria that identify startups who will benefit most:

• Deep tech focus: Require technical novelty, value prior research experience
• Go-to-market support: Value founders with strong products but limited sales experience
• Industry connections: Look for startups in industries where you have networks

Avoid Criteria Creep

It's tempting to add criteria for every edge case. Resist this. Each additional criterion:

• Adds evaluation time
• Increases false negatives (rejecting good startups on technicalities)
• Dilutes focus on what truly matters

Start with 3–4 required and 4–6 nice-to-have. Add more only when you find consistent gaps.

Example Criteria Sets

Early-Stage B2B SaaS Accelerator

Required:

1. B2B software focus
2. Technical co-founder on team
3. Pre-seed or seed stage
4. Full-time founder commitment

Nice-to-Have:

1. Existing revenue or LOIs
2. Prior startup or scale-up experience
3. Warm introduction to program
4. Clear product demo or prototype
5. Domain expertise in target vertical

Climate Tech Program

Required:

1. Direct climate impact (emissions reduction, adaptation, monitoring)
2. Technical innovation (not purely operational)
3. Founders with relevant scientific or engineering background
4. US-based or willing to relocate

Nice-to-Have:

1. Published research or patents
2. Prior pilot with enterprise customer
3. Hardware or deep tech experience
4. Grant funding secured
5. Academic or lab partnership

Generalist Pre-Seed Fund

Required:

1. Pre-product or early product stage
2. Seeking $500K–$2M raise
3. At least two co-founders
4. Full-time commitment

Nice-to-Have:

1. Technical founder
2. Repeat founders
3. Unique insight or distribution advantage
4. Large market opportunity ($1B+)
5. Early user engagement signals

Weighting and Scoring

FounderScan uses your criteria designations to calculate overall scores:

• Required criteria: Weighted 2x by default
• Nice-to-have criteria: Weighted 1x

This means a startup scoring 8/10 on all required criteria and 6/10 on nice-to-haves will rank higher than one scoring 7/10 across the board.

You can adjust these weights in batch settings if your program has different priorities.

Iterating on Criteria

Your criteria should evolve based on results:

After each cohort, ask:

• Did highly-scored startups actually succeed in the program?
• Did we miss great startups due to criteria technicalities?
• Were any criteria consistently hard to evaluate?
• Are there patterns in successful companies we didn't capture?

Use these insights to refine criteria for the next cycle.

Common Mistakes

Over-indexing on Pedigree

"Founder from FAANG or top-10 startup" sounds good but excludes many exceptional founders. Consider what pedigree is actually a proxy for (execution ability, network, technical depth) and measure those directly.

Ignoring Stage Fit

A Series A company with $3M ARR might score well on "traction" but be a poor fit for a pre-seed accelerator. Make stage requirements explicit.

Too Many Required Criteria

If you require 8 things, you'll reject almost everyone. Required should be 3–5 absolute essentials.

Criteria Without Evidence

"Founders get along well" can't be assessed from a written application. Either remove it or translate it to something observable.

---

Ready to put these principles into practice? Schedule a demo and we'll help you set up your evaluation criteria.

For more on interpreting the resulting scores, see our guide on understanding AI reasoning.