Comp Selection Criteria: A Checklist for Analysts and Appraisers

The right comp set comes down to six filters applied in order: industry or property type, size and scale, geography or submarket, timing and recency, profitability (or cap rate and cash flow for real estate), and capital structure or lease terms. Get these six right and your valuation holds up in committee. Get even one wrong, and you’re defending a number instead of a method.

Before you dig into methodology, copy this checklist into your workflow:

• Match sector/property type first, then size, then geography — never the reverse.
• Cap your sample at 8 to 12 comparables unless a thin market forces fewer.
• Require sales or transactions within a recent time window, typically a few months, extending only with a documented reason.
• Exclude non-arm’s-length transactions and flag outliers before scoring.
• Write a one-sentence inclusion rationale for every comp you keep.

Pro Tip: If you can’t write one sentence explaining why a comp belongs in your set, it doesn’t belong. That single discipline catches more bad comps than any spreadsheet filter.

A defensible comp set is a narrow, disclosed group, not the biggest list you can pull. Eight to twelve comparables, each with a documented reason for inclusion, beats thirty comps nobody can explain.

Key Takeaways

Defensible comp selection depends on disclosed screening criteria, a capped sample of 8 to 12 comparables, and a documented one-sentence rationale for every comp included.

Point	Details
Screen in order	Filter by sector/property type first, then size, then geography and timing last.
Cap your sample	Keep 8 to 12 comparables and document why if you use fewer.
Adjust visibly	Normalize accounting or physical differences separately and show each adjustment’s impact.
Verify before finalizing	Confirm deed, financing type, and tax records to catch non-arm’s-length sales.
Automate the screening	DealAnalyzerAI generates ARV ranges and adjustment grids from screened property comps, with human sign-off still required.

Table of Contents

• What Comp Selection Criteria Actually Change in a Valuation
• What Are the Core Selection Criteria for Comparable Companies?
• What Criteria Matter Most for Real Estate Appraisal Comps?
• How Do You Build a Comparable Universe Step by Step?
• How Do You Normalize and Adjust Comps Fairly?
• Which Data Sources Should You Trust for Comp Research?
• Building a Scoring Matrix for Ranking Comps
• How Automation Fits Into a Repeatable Comp Workflow
• Sources
• FAQ

What Comp Selection Criteria Actually Change in a Valuation

Comp selection criteria move your indicated value more than any rounding decision on your multiple ever will.

Three quick examples: adding one high-growth outlier to a SaaS comp set inflates your median EV/Revenue multiple. Extending a property search from 3 months to 12 months in a rising market drags older, cheaper sales into your average and understates value. Dropping a lease-adjusted retail comp because its capital structure looks messy often removes the one comp with the most similar risk profile.

Pro Tip: Lenders and investment committees almost always ask the same question first: “Why these comps and not others?” Have that answer written down before you present, not improvised in the room.

What Are the Core Selection Criteria for Comparable Companies?

The top screens for trading comps and precedent transactions are sector/subsector, product or service mix, size band, growth band, margin band, capital structure, geography, and transaction type. Corporate Finance Institute identifies industry classification, size, geography, growth rate, profitability, and capital structure as the factors that materially move valuation multiples, and that list holds up across sectors.

Numerical bands give you something to defend instead of a gut feeling. A workable starting point:

Criterion	Recommended band	When to widen
Size (revenue/ARR)	0.5 to 2 times target	Thin sector population
Growth rate	Within a moderate percentage range	Early-stage or volatile category
Gross margin	Within a small range of points	Mixed business models in sector
Capital structure	Similar leverage tier	No exact leverage match exists
Geography	Same primary market	Global comps if sector is niche

Trading comps and precedent transactions answer different questions. Trading comps reflect current public sentiment; precedent transactions capture control premiums paid in real deals. Weight precedent transactions more heavily when you’re valuing a control sale, and lean on trading comps for a minority stake or a quick sanity check.

Here’s a worked example.

Pro Tip: Write down your three screening filters and one sentence per comparable explaining why it made the cut. That habit alone is what separates a defensible model from a spreadsheet nobody can audit six months later.

What Criteria Matter Most for Real Estate Appraisal Comps?

A practitioner-tested list of top search criteria includes school district, GLA, year built, date of sale, neighborhood, site size, condition, and money-maker features, and that order matters: start with location and physical match before you ever look at price.

Standard practice calls for closed, arm’s-length sales within 3 to 6 months and within roughly a 1-mile radius or the same subdivision, according to appraisal guidance on comp selection. In thin markets, extend the window to 6 to 12 months, but document why and adjust for time.

Filter	Standard threshold	Expansion rule
Sale recency	3 to 6 months	Extend to 6 to 12 months in thin markets
Distance	Same subdivision to 1 mile	Widen to submarket if inventory is sparse
GLA match	Within a generally narrow percentage range	Adjust with $/sq ft if outside range
Condition	Similar renovation state	Adjust dollar-for-dollar on documented rehab

Here’s a quick adjustment calculation. Your subject property has 1,800 square feet; your best comp has 1,650 square feet and sold for $330,000. At an area rate of $150 per square foot, you add $22,500 for the 150-square-foot gap, adjusting the comp to $352,500 before any condition adjustments.

Pro Tip: Bracket your comp set with one superior and one inferior property whenever you can. It shows a reviewer your subject sits inside a defensible range, not outside a curated best case. Always exclude foreclosures, short sales, and family transfers unless you can prove they were arm’s-length.

How Do You Build a Comparable Universe Step by Step?

Screen in three stages: sector or property type first, then size or stage, then transaction date and geography last. At each stage, record how many candidates you started with and how many survived. That population-reduction trail is what Opagio recommends for a defensible comp set, and it’s the single fastest way to show a skeptical reviewer your logic.

The methodology, step by step:

1. Search broadly using sector/property-type and transaction-type filters.
2. Apply size or scale bands to cut the population to a workable list.
3. Filter by geography and transaction date, tightening or loosening as needed.
4. Flag outliers, non-arm’s-length deals, and duplicate listings.
5. Score remaining candidates against weighted criteria.
6. Document a one-line rationale for every comp you keep or drop.

Rank survivors with numeric weights rather than gut feel: industry or product match, size, geography, timing, and profitability each deserve a score, not a shrug. Export your working list into a table with these fields: comp name/address, sector or property type, size metric, date, distance or geographic tier, adjustment notes, and inclusion rationale. That schema doubles as your audit trail later.

This discipline mirrors a broader lesson from technology vendor selection: skipping a defined objective before searching is one of the most common procurement mistakes teams make, and the same failure shows up when analysts search for comps before deciding what “comparable” even means for their specific case.

How Do You Normalize and Adjust Comps Fairly?

Normalize before you compare, never after. Strip non-recurring items from earnings, align accounting bases across entities, convert between per-unit and aggregate metrics consistently, and time-adjust every value for market movement between the comp’s transaction date and your valuation date.

On the corporate side, if a target’s EBITDA includes a one-time $500,000 litigation settlement, back that out before calculating your multiple.

Keep adjustments visible, not buried. For corporate comps: normalize EBITDA, adjust for lease accounting differences, and note any add-backs explicitly. For property comps: adjust GLA, condition, lot size, and time separately rather than lumping them into one number. See how paired-sales adjustment methods apply this line by line.

Pro Tip: *Run a quick sensitivity check: remove your single largest adjustment and see how much your conclusion moves.

Which Data Sources Should You Trust for Comp Research?

For corporate comps, Bloomberg and S&P Capital IQ remain the standard for pulling trading multiples, precedent transactions, and capital structure data with audit-ready sourcing. For property comps, the MLS (Multiple Listing Service), county public records, and CoStar cover the bulk of verified transaction data, with platforms like Zillow useful for a quick sanity check rather than a primary source.

DealAnalyzerAI automates much of the property-side workflow, screening sales pools and flagging problem transactions faster than manual MLS searches.

8 to 12 comparables, cited with disclosed screening criteria, is the range that holds up under partner or lender review.

Before finalizing any comp set: verify the deed, confirm the financing type (cash vs. conventional vs. distressed), and cross-check against tax assessment records to catch non-arm’s-length transactions your first pass missed.

Building a Scoring Matrix for Ranking Comps

A weighted scoring matrix turns subjective comp selection into a repeatable process.

Export these fields into your worksheet: comp ID, raw metric values, weighted score per criterion, and total. Set a cutoff score (commonly 70 out of 100) and keep everything above it, expanding your parameters only if you land under 8 usable comps.

How Automation Fits Into a Repeatable Comp Workflow

Automation speeds up screening and produces the documentation trail reviewers actually want, but it doesn’t replace human judgment on the final call. A workable pipeline runs data ingestion, then rule-based screening, then outlier and non-arm’s-length flagging, then an adjustment grid, then human sign-off. That sequencing mirrors how AI-driven comp selection tools already operate in commercial real estate.

DealAnalyzerAI applies this to property analysis: generating ARV ranges from screened comps, building adjustment grids, and exporting a rationale per comp. See how ARV calculation approaches compare across methods.

Pro Tip: Record which reviewer signed off and which version of your rule set generated the comp list. When someone questions a valuation eight months later, that record saves you hours.

What I’ve Learned From Reviewing Comp Sets Repeatedly

Every defensible comp set I’ve seen shares three habits: the analyst discloses their screens up front, brackets with a superior and inferior comp, and annotates every exclusion, not just every inclusion. Watch for cherry-picking, stale comps stretched past their time window, and over-reliance on a single multiple or price-per-square-foot figure.

Pro Tip: Reviewers trust a comp set more when you show your rejects, not just your winners. Volunteer that story before they ask for it.

Screen Faster Without Cutting Corners on Documentation

Manually cross-referencing MLS listings, county records, and CoStar pulls for every property eats hours you don’t have when you’re screening a dozen deals a week. DealAnalyzerAI runs that screening automatically, flags non-arm’s-length sales, and builds the adjustment grid a lender or partner will ask to see, so you spend your time reviewing conclusions instead of assembling them.

The tool generates ARV ranges from your screened comps, exports a per-comp rationale, and calculates rehab costs from uploaded property photos so your adjustment grid reflects real renovation math, not guesswork on money-maker features. Neighborhood-level renovation value plays into this too; see how renovation choices interact with local comps before finalizing your feature adjustments.

Start a free analysis at DealAnalyzerAI and run your next property through the same screening logic this guide just walked through.

Sources

For corporate comp methodology, consult Corporate Finance Institute and Opagio, both of which include worked examples on screening and disclosure. For property comps, the New Jersey appraisal guide and Birmingham Appraisal Blog cover physical filters and time/distance bracketing in detail. For automation methodology, the AI comp selection primer walks through rule-based screening and flagging. Keep a saved snapshot of every source page you cite in a final report. Data providers update listings and multiples constantly, and you’ll need that snapshot if anyone questions your numbers later.

This article is general information, not a substitute for advice from a qualified financial advisor. Consult a qualified financial professional about your own circumstances before acting on anything here.

• How to Build a Defensible Comp Set | Opagio
• How to Choose Comparable Companies | Corporate Finance Institute
• How an Appraiser Selects Comps: A Practical Guide – New Jersey Property Valuation Insights
• My top 8 search criteria for finding the best comps | Birmingham Appraisal Blog
• AI Comp Selection: Build a Defensible CRE Comp Set

FAQ

How Many Comparables Should You Include in a Comp Set?

Eight to 12 comparables is the standard defensible range, with disclosed screening criteria and a one-sentence rationale for each. Use fewer only in thin markets, and document why explicitly.

What’s the Biggest Comp Selection Mistake Analysts Make?

Relying on price-per-square-foot or a single multiple instead of matching on physical and locational criteria first. Practitioners consistently warn against searching by estimated sale price rather than starting with property characteristics.

How Far Back Should Comp Sale Dates Go?

Standard practice is 3 to 6 months for arm’s-length sales, extending to 6 to 12 months only in thin markets with documented time adjustments.

Can AI Tools Replace Manual Comp Selection?

AI tools like DealAnalyzerAI speed up screening, flagging, and adjustment-grid creation, but human sign-off remains required for the final selection and judgment calls.

What Size Band Should You Use for Corporate Comparables?

Recommended

• Property Offer Decision Criteria List for Investors
• How to Adjust Comps for Property Differences: 2026 Guide
• Standardize Investment Property Scoring for Active Investors
• How Investors Evaluate Multiple Properties Efficiently