Quick Definition
Manual Screening — Manual Screening is the critical step in benchmarking where potential comparable companies identified through database searches are individually reviewed for functional comparability.
Our glossary is continuously expanding. For deep-dive guides on specific topics, explore our blog or contact us with suggestions.
Manual Screening is the critical step in benchmarking where potential comparable companies identified through database searches are individually reviewed for functional comparability. After quantitative filters narrow the initial search, each remaining company must be manually evaluated to determine whether its business profile genuinely matches the tested party. Manual screening distinguishes rigorous transfer pricing analysis from mechanical database queries.
Software can assist with collecting company information and preparing screening suggestions. A practitioner still needs to assess functions, assets and risks against the tested party and approve the decision. See how AI-supported comparable screening fits into that review.
The OECD Transfer Pricing Guidelines (2022) emphasize that database searches alone are insufficient in Chapter III. The Guidelines note that database searches are only one part of identifying potential comparables—it is important not to over-rely on quantitative criteria for comparable selection.
The Guidelines also note that a manual review or qualitative analysis of information relating to each potential comparable is needed to confirm comparability and form a reliable comparable set.
US Treasury Regulations §1.482-1(d)(3) similarly require evaluation of comparability factors beyond what databases can capture, implicitly mandating manual review.
Manual screening typically occurs after quantitative filters reduce the database population from thousands to hundreds of companies. The process:
Manual Screening Workflow:
| Step | Action | Tools/Sources |
|---|---|---|
| 1 | Read business description | Database profile, company website |
| 2 | Verify industry classification | Check actual activities match NACE/SIC code |
| 3 | Assess functional profile | Is this genuinely a distributor/manufacturer/etc.? |
| 4 | Check for disqualifying activities | Manufacturing by "distributors," IP ownership |
| 5 | Review ownership/independence | Verify no related-party issues |
| 6 | Document accept/reject decision | Create audit trail |
Common Rejection Reasons:
| Rejection Reason | Example |
|---|---|
| Wrong functions | "Distributor" that actually manufactures |
| Different business model | Retail instead of wholesale |
| Unique intangibles | Owns proprietary technology or brand |
| Not independent | Related-party transactions affecting results |
| Abnormal circumstances | Restructuring, bankruptcy, start-up |
| Insufficient information | Can't determine actual business activities |
Databases Can't Assess Comparability: NACE codes and financial ratios tell you industry and size, not functions. A company coded as "wholesale distribution" may manufacture, own IP, or operate retail stores. Only manual review reveals the true business model.
Tested Party: Limited-risk distributor of industrial chemicals in Germany.
Post-Quantitative Filter Set: 45 companies passed database filters (NACE code, geography, independence, revenue, data availability).
Manual Screening Sample:
| Company | Business Description | Decision | Reason |
|---|---|---|---|
| Chem-Dist GmbH | "Wholesale of chemical products to industrial customers. No manufacturing." | ✅ Accept | Functions match LRD profile |
| ChemTech AG | "Development and production of specialty chemicals with distribution." | ❌ Reject | Manufacturing + R&D—not comparable |
| EuroChemicals | "Import and distribution of chemicals; also provides blending services." | ❌ Reject | Blending = value-added transformation |
| Industrial Supply BV | "Wholesale distributor of industrial supplies including chemicals." | ⚠️ Review | Broad product mix—may include non-chemical items |
| Green Chem Ltd | "Distribution of eco-friendly chemicals. Owns proprietary formulations." | ❌ Reject | Owns IP—not limited-risk |
Final Result: Of 45 companies passing quantitative filters, 12 were accepted as comparable after manual screening.
Every manual screening decision must be documented—this is critical for audit defense:
| Documentation Element | Purpose |
|---|---|
| Company name and database ID | Identification |
| Business description reviewed | Source of information |
| Key activities identified | What the company does |
| Accept/Reject decision | Clear conclusion |
| Rationale | Why accepted or rejected |
| Information source | Database, website, annual report |
Accept/Reject Matrix: Create a spreadsheet documenting every company reviewed, the decision, and reasoning. Tax authorities commonly request this during audits. A well-documented matrix demonstrates rigor and protects your comparable set.
Databases capture structured data—industry codes, financials, ownership percentages. They can't assess functional comparability—whether a company actually performs similar activities to your tested party. A "distributor" by NACE code might manufacture, own IP, or operate a different business model. Only human review can evaluate these qualitative factors.
Screen all companies passing quantitative filters. If that's impractical (hundreds of companies), consider tightening quantitative criteria first. Common practice: screen 50-150 companies to arrive at 10-20 accepted comparables. Document the screening process even for rejected companies.
Primary sources in order of reliability: (1) database business descriptions (Orbis, Amadeus), (2) company websites, (3) annual reports if available, (4) regulatory filings, (5) news articles for context. Cross-reference sources when business descriptions are ambiguous or incomplete.
Generally no—reject based on functional differences, not financial outcomes. If a company passes functional screening, its profitability is a valid data point. Excluding companies because their margins are unfavorable creates bias. Exception: persistent losses without business explanation may indicate structural issues warranting rejection.
Investigate further using company websites and annual reports. If you still can't determine comparability, you have two options: (1) reject with documented uncertainty—"insufficient information to confirm comparability," or (2) include with caveat—acknowledge the uncertainty and note it may be a comparability defect. Conservative approach: reject when uncertain.
Not necessarily. Once accepted, a comparable typically remains in the set unless circumstances change (acquisition, restructuring, business model change). Monitor for material changes annually but don't re-run full manual screening unless the comparable's business profile has shifted. Consistency demonstrates robustness.
One to two sentences explaining why the company isn't comparable. Examples: "Rejected—company manufactures chemicals in addition to distribution." "Rejected—significant related-party transactions (subsidiary of multinational group)." "Rejected—business model is retail sales, not wholesale distribution." Brief but specific.