Screening Comps from a Starting Universe

You have identified that your target company is a mid-cap software-as-a-service (SaaS) firm. You go to a financial database and search for all SaaS companies. You get a list of 200. Now what? You cannot build a peer set of 200 companies. You need to screen that universe down to 5–10 true peers.

Screening is the mechanical process of filtering a large universe of potential peers using quantitative and qualitative rules, moving from thousands of candidates to dozens to a final set of maybe ten. It requires discipline, transparency, and clear documentation. Done well, screening creates a defensible peer set. Done sloppily, it introduces bias and produces garbage.

This article walks through the screening process step by step, showing you how to build filters, apply them systematically, and document your work so others can follow and critique your logic.

Quick definition

Screening is the process of applying objective filters (size, growth, profitability, leverage, business model) to a starting universe of companies in order to identify a subset of true peers. Screening moves from broad to narrow: universe → candidate list → shortlist → final peer set.

Key takeaways

• Start with a large universe (all companies in an industry or sector) and apply filters sequentially
• Use both quantitative filters (revenue size, growth rate, margins) and qualitative filters (business model, geographic focus, customer base)
• Transparency is critical: document each filter, the rationale, and how many companies pass each stage
• A funnel approach (each filter narrows further) ensures you do not miss candidates at early stages
• Exclude companies that fail a critical criterion, even if they pass all others
• Review the final peer set for obvious gaps (outliers, companies that slipped through despite being misaligned) and manually adjust if needed
• Be prepared to revisit your filters if the final set does not feel right

The screening funnel: from universe to peers

The screening process is best visualized as a funnel, getting narrower at each stage:

Starting Universe: 200+ companies in software/SaaS sector
                    ↓
After Filter 1 (Business Model - SaaS only): 120 companies
                    ↓
After Filter 2 (Scale - $100M to $10B revenue): 85 companies
                    ↓
After Filter 3 (Growth - 10% to 40% revenue growth): 40 companies
                    ↓
After Filter 4 (Profitability - 15% to 40% EBITDA margin): 22 companies
                    ↓
After Filter 5 (Leverage - Debt-to-EBITDA under 2x): 15 companies
                    ↓
Manual Review & Adjustment: 8 final peers

Each filter removes a portion of the universe. By the end, you have a narrowed set that you can review, debate, and finalize.

Filter 1: Business model and product type

The first cut is usually business model. Are you looking for pure SaaS companies, or are you including product-plus-services hybrids? Are you including only cloud-based software, or on-premise as well? Are you including consulting services or only software licenses?

Example: SaaS screening

You are valuing Company X, a pure SaaS firm with a cloud-based platform for HR management. Your starting universe is "Software & Technology" (200+ companies). Apply Filter 1: Include only companies that are 75%+ SaaS or subscription revenue.

This removes:

• System integrators (consulting-heavy, low recurring revenue)
• Hardware + software vendors (mixed business model)
• On-premise software makers (older model, different economics)

You are left with 120 companies (60% of starting universe).

Practical rule: If you are unsure whether to include a company at this stage, include it. You can exclude it later if other filters reveal it is a poor fit.

Filter 2: Scale and market cap

Apply a size filter to narrow the universe based on revenue, EBITDA, or market cap. This removes companies that are too large (different economics, no longer "comparable") or too small (too different operationally).

Example: SaaS size screening

Your target has $500 million in revenue. You want peers within 0.5x to 2x scale, or roughly $250 million to $1 billion revenue.

Apply Filter 2: Revenue between $200M and $1.2B.

This removes:

• Mega-cap SaaS companies like Salesforce ($32 billion revenue) and ServiceNow ($9 billion revenue) — too large
• Micro-cap SaaS companies like smaller niche vertical SaaS makers ($50M revenue) — too small

You are left with 85 companies (71% of previous stage).

Why revenue scale matters: Larger companies benefit from scale economics (lower customer acquisition costs, higher margins) and market liquidity (lower equity risk premiums). A $30 billion company trades at a different multiple than a $500 million company, all else equal.

Filter 3: Growth rate

Apply a growth filter to ensure peers are growing at a similar pace to your target.

Example: SaaS growth screening

Your target is growing at 18% annually (revenue growth YoY). You want peers growing at 12% to 25% (a reasonable band).

Apply Filter 3: 3-year revenue CAGR between 12% and 25%.

This removes:

• Mature, slow-growth SaaS (2–8% growth) — different market position, likely different margins
• Hypergrowth SaaS (40%+ growth) — different risk profile, may not be profitable yet

You are left with 40 companies (47% of previous stage).

Why growth matters: A slow-growth SaaS company might trade at 30x revenues; a fast-growth one might trade at 80x revenues. If you mix them, your multiple is nonsensical. Narrow the band to companies growing at a similar rate.

Filter 4: Profitability and margins

Apply a profitability filter to ensure peers have similar earnings quality and operational efficiency.

Example: SaaS profitability screening

Your target has 25% EBITDA margin (or operating margin). You want peers with 18% to 35% margins.

Apply Filter 4: EBITDA margin between 18% and 35%.

This removes:

• Unprofitable SaaS companies or those with very low margins (5–15%) — either burning cash or have a very different model
• Ultra-high-margin SaaS companies (40%+) — likely very different (niche, winner-take-most markets, or lower-touch models)

You are left with 22 companies (55% of previous stage).

Why margins matter: Margins reflect operational leverage, pricing power, and customer economics. Two companies can have the same revenue and growth but vastly different EBITDA margins, signaling different quality and risk profiles. Peers should be in the same margin band.

Filter 5: Leverage and capital structure

Apply a leverage filter to ensure peers have similar debt levels and financial risk.

Example: SaaS leverage screening

Your target has minimal debt (not a capital-intensive business). You want peers with similar low leverage, say debt-to-EBITDA under 2x (or net debt-to-EBITDA under 1x).

Apply Filter 5: Net debt-to-EBITDA under 1.5x.

This removes:

• Highly leveraged SaaS companies that have taken on significant debt (via LBOs or acquisitions) — different financial risk, different cost of capital
• Companies in financial distress — may not be traded fairly

You are left with 15 companies (68% of previous stage).

Additional filters (as needed)

Depending on your industry and target, you might add more filters:

Filter 6: Customer concentration

Some SaaS companies are highly dependent on a few large customers; others have diversified customer bases. If your target has a diversified base, exclude companies with >20% revenue from a single customer.

Filter 7: Geographic exposure

Does your target operate primarily in North America, or globally? If globally, exclude purely domestic-focused companies. If North America-focused, exclude primarily international ones.

Filter 8: Acquisition activity

If your target is "pure-play" (no acquisitions), include only peers with minimal M&A activity. If your target is a serial acquirer, include peers with active acquisition strategies.

Filter 9: Product category or vertical

If your target is a vertical SaaS firm (e.g., SaaS for healthcare), try to include peers in the same vertical. If no vertical-specific peers exist, expand to broader SaaS.

A complete screening example: valuing a mid-market CRM software company

Let's walk through a full screening for a company called SalesHub Inc., a $300 million revenue CRM (customer relationship management) software company, growing at 15% annually, with 22% EBITDA margins and minimal debt.

Starting universe: All software companies (sector: software & IT services) on a major stock exchange: ~250 companies.

Filter 1: Business model

• Include: ≥70% recurring/subscription revenue
• Exclude: System integrators, consulting-heavy, on-premise legacy software
• Result: 140 companies (56% of universe)

Filter 2: Scale

• Include: $150M to $1B revenue (±0.5x to 3x target's scale)
• Exclude: Smaller and larger companies
• Result: 75 companies (54% of previous)

Filter 3: Growth

• Include: 10% to 22% revenue growth (±5% around target's 15%)
• Exclude: Slower or faster growers
• Result: 38 companies (51% of previous)

Filter 4: Profitability

• Include: 15% to 30% EBITDA margin
• Exclude: Unprofitable or ultra-high-margin outliers
• Result: 22 companies (58% of previous)

Filter 5: Leverage

• Include: Net debt-to-EBITDA under 1.5x
• Exclude: Highly leveraged or distressed companies
• Result: 15 companies (68% of previous)

Filter 6: Product category

• Include: CRM, sales automation, customer engagement (same vertical)
• Exclude: Horizontal platforms not focused on sales
• Result: 7 companies (47% of previous)

Manual review: You examine the 7 companies remaining and note:

• HubSpot: $2 billion revenue (too large, exclude)
• Salesforce: $35 billion revenue (way too large, exclude)
• Veeva Systems: $1.8 billion revenue (too large, borderline)
• Zendesk: $1.4 billion revenue (large, borderline)
• Slack (now Salesforce subsidiary): Acquired, exclude
• Monday.com: $400 million revenue, 20% growth, 12% margin (close fit)
• SolarWinds: Different vertical (IT ops), exclude
• Alteryx: Different category (analytics), exclude
• Twilio: Different category (communications), exclude

After manual review, you are left with only 2–3 companies in your narrow vertical. This is too few. You broaden Filter 6:

Filter 6 (revised): Product category (broader)

• Include: CRM, sales, customer engagement, and adjacent business software (broader net)
• Result: 10 companies

Manual review (revised): The 10 remaining companies are:

1. HubSpot (2B revenue) — too large
2. Zendesk (1.4B) — borderline large
3. Monday.com (400M revenue, 20% growth, 12% margin) — fit
4. Smartsheet (2.5B) — too large
5. Alteryx (900M) — different category (analytics)
6. Datadog (1.5B) — different category (infrastructure monitoring)
7. Okta (600M) — different category (identity management)
8. Twilio (1.2B) — different category (communications)
9. Fortive (8B) — industrial conglomerate, way off
10. RingCentral (700M, 18% growth, 10% margin) — adjacent (communications, not CRM)

Pare this down using judgment and end with:

1. Monday.com — strong fit (400M, 20% growth, 12% margin)
2. Zendesk — borderline large but fits profile (1.4B, 15% growth, 15% margin)
3. Atlassian — adjacent category but strong fit (5B+ rev, too large—exclude)
4. Okta — identity management, adjacent to CRM (600M, 30% growth, 5% margin—too high growth)

After further review, you have limited direct peers. This is common in software. You have two options:

Option A: Narrow your focus

Accept that SalesHub has limited direct public peers and use:

1. Monday.com
2. Zendesk

Supplement with transaction comps (recent M&A of similar CRM companies).

Option B: Broaden your peer set slightly

Include adjacent software categories (customer engagement, sales platforms, communication platforms) and apply adjustments for differences in growth, margin, etc. Expand to 6–8 peers, noting the trade-off between precision and sample size.

Documenting the screening process

Create a screening worksheet that shows:

1. Starting universe: Description and count (e.g., "All SaaS companies, N=250")
2. Filter 1: Criterion, logic, companies removed, remaining count
3. Filter 2–5: Same as above
4. Final shortlist: List of companies passing all filters
5. Manual review notes: Any companies excluded at the end despite passing filters, and why
6. Final peer set: List of peers with key metrics (revenue, growth, margin, leverage)

This transparency allows others to:

• Understand your logic
• Debate whether each filter was too strict or too loose
• Suggest alternative screening criteria
• Identify if you accidentally excluded a good candidate or included a bad one

Common screening mistakes

Mistake 1: Filters that are too strict

If your filters eliminate 95% of the universe and you are left with only 3 companies, your filters are too tight. Peers do not have to be clones of the target; they have to be reasonably similar. Loosen the bands slightly (e.g., ±10% instead of ±5% for growth) to capture more candidates.

Mistake 2: Filters that are too loose

If your 8 final peers have revenue ranging from $50 million to $5 billion and growth rates from 2% to 40%, your filters were too loose. You have included companies that are barely comparable. Tighten the key filters to increase homogeneity.

Mistake 3: Ignoring outliers at the screening stage

As you apply filters, some companies will pass all your tests but will feel "off" when you examine them manually. Trust that feeling. You might see that Company X is in the same industry and has similar metrics, but it operates in a different geography with different regulations, or its revenue comes from a different product line. Exclude it; your job is not to force-fit companies into the peer set.

Mistake 4: Using stale data

If you screen using Q2 financial data but it is now Q4 (six months later), your growth rates and margins are outdated. Use the most recent full-year or trailing-twelve-month data available. For a screening, this is usually not a dealbreaker, but for a final valuation, use current data.

Mistake 5: Arbitrary filter thresholds

Do not just randomly pick "revenue between $200M and $1B" because it sounds reasonable. Explain why. If your target is $400M, the justification might be "±50% scale, or $200M to $600M, rounded to $200M–$1B to capture some modest outliers." This rationale can be discussed and improved.

Mistake 6: Over-filtering on profitability

Some companies with identical business models have different profitability due to operational maturity, scale, or management quality. Do not exclude a good peer just because margins are 20% vs your target's 25%. Instead, note the difference and be prepared to discuss it.

The feedback loop: screening to peer set to valuation

Once you have your final peer set, compute their valuation multiples. If the multiples range is very tight (e.g., all peers at 9x to 11x EV/EBITDA), your screening has worked well—you have a homogeneous group. If the range is very wide (e.g., 5x to 18x EV/EBITDA), your peer set is still too heterogeneous. You might need to:

1. Add another filter to narrow the set further
2. Manually exclude the outliers at the high and low end
3. Segment your peers into two sub-groups (e.g., "high-growth, higher-margin peers" and "lower-growth, lower-margin peers") and use separate multiples for each

A visual screening process

FAQ

Q: How many filters should I apply?

A: As many as needed to reach a homogeneous peer set, typically 4–7. Too few filters (1–2) and you have a diverse hodgepodge. Too many filters (10+) and you are over-optimizing and may eliminate good candidates. Start with 5 core filters (business model, scale, growth, profitability, leverage) and add more only if the initial set is still too diverse.

Q: Should I screen on market cap or revenue?

A: Revenue is better for most analyses because it is less distorted by market sentiment. Market cap can swing wildly based on sentiment, while revenue is stable. If peers have very different profitability, revenue-based screening is more objective.

Q: What if screening leaves me with fewer than 5 peers?

A: You have a few options: (1) Broaden one or more filters (e.g., loosen the growth band from ±5% to ±10%), (2) Include transaction comps instead, or (3) Supplement with manual peer selection for companies that are close fits but did not pass your screens. Acknowledge the smaller sample size and the limits of your peer set.

Q: Should I screen on net debt or gross debt?

A: Net debt is better because it accounts for cash balances. A company with $500M of gross debt but $400M of cash is much less leveraged than the gross debt number suggests. Use net debt-to-EBITDA.

Q: Can I screen on profitability using net income instead of EBITDA?

A: It is less ideal because different companies have different depreciation, interest, and tax situations. EBITDA or operating margin is more comparable. That said, if you are valuing on a P/E basis (profit multiple), you might screen on net income margins to ensure consistency.

Q: What if a major peer barely fails one of my filters?

A: Revisit your filter. Is it too strict? If Competitor X has 8% growth and your filter is "10% to 22% growth," and Competitor X is otherwise a perfect fit, consider loosening the filter to "8% to 22%." Transparency is key: note that you included a peer that technically fell outside your band and why you made that judgment call.

Related concepts

• Peer set quality: The foundation of accurate comparable company analysis
• Multiple dispersion: When your peers show a wide range of multiples, signaling heterogeneity
• Outlier removal: Identifying and handling peers that are statistical outliers in multiples
• Adjustment factors: Tweaking peer multiples to account for differences in growth, margin, or leverage

Summary

Screening is the systematic process of filtering a large universe of companies down to a focused peer set. Use a funnel approach: apply filters sequentially, starting broad and narrowing down. Start with business model and scale, add growth and profitability filters, and finish with leverage and category adjustments. Aim for 5–8 final peers with reasonable homogeneity. Document each filter and its rationale so others can follow and critique your work. If your final multiples show a very wide range, your screening was too loose; revisit and tighten. If screening leaves you with too few peers, broaden your filters or supplement with transaction comps. The goal is a defensible peer set that is "best available" even if not perfect.

Next

Read the next article, Direct vs indirect peers, to understand the distinction between companies that compete directly with your target and those in adjacent segments, and when each type belongs in your valuation analysis.