Matching logic explained

Overview

Matching logic determines how Cognism identifies the correct Contact or Account when you search, preview, enrich, or sync data.

Cognism uses a confidence-based scoring approach. The platform evaluates the data you provide (such as name, email, job title, or company details) and only returns a result when it meets a high-quality confidence threshold.

If confidence is too low, Cognism returns no data rather than guessing. This helps protect data quality and prevent incorrect enrichment.

Matching helps you:

• Find the right person or company faster
• Avoid enriching the wrong record
• Reduce duplicate Contacts and Accounts in your CRM
• Maintain accurate and reliable go-to-market data

Why matching matters

B2B data changes frequently. People change roles, companies evolve, and CRM records can become outdated.

Matching logic helps Cognism:

• Identify the correct profile from a large dataset of global professionals
• Prioritise high-confidence matches
• Avoid returning uncertain or incorrect results

How matching works

Understand how match scoring works

Cognism calculates a match score to represent confidence in a result.

Each input field you provide contributes to this score. More reliable or specific data increases the likelihood of a strong match.

Use strong identifiers

Some fields consistently provide strong signals and contribute more reliably to matching:

• First name
• Last name
• Email
• Professional social profile (for example, LinkedIn)

These fields help Cognism identify individuals with higher certainty.

Provide specific supporting data

Other fields contribute based on how specific they are:

• Job title
• Company name
• Company website or domain

More specific values increase confidence.

For example:

• “Software Engineer” contributes less (common role)
• “Senior Machine Learning Engineer” contributes more (specific role)

Similarly:

• “Acme” contributes less (ambiguous company name)
• “Acme Corporation Ltd” or a company domain contributes more

Meet the confidence threshold

Cognism only returns a result when the match score meets an internal high-quality threshold.

If no profile meets this threshold, no data is returned.

This ensures results prioritise accuracy over volume.

Understand match score ranges

Match scores can be broadly understood on a scale from 1 to 100.

• 80–100: Very high confidence, strong alignment across fields
• 65–79: High confidence, minor gaps or variation
• 50–64: Moderate confidence, less specific or partially matching data

Higher thresholds prioritise accuracy. Lower thresholds increase the chance of returning a result when data is incomplete or contains errors.

Example matching behaviours

Use high-accuracy matching

{
  "firstName": "Jane",
  "lastName": "Doe",
  "jobTitle": "Senior Product Designer",
  "accountWebsite": "example.com",
  "minMatchScore": 45
}

• Returns results only when inputs strongly align
• Best for high-quality, structured data
• May return no result if details have changed

Use broader matching

{
  "firstName": "Jon",
  "lastName": "Doe",
  "jobTitle": "Software Enginer",
  "accountName": "Acme",
  "minMatchScore": 28
}

• Returns results even when inputs are incomplete or contain errors
• May match one of several similar profiles
• Useful when data quality is uncertain or inconsistent

Use unique identifiers

{
  "email": "john.doe@example.com"
}

• Uses a direct identifier to find a match
• Typically results in a high-confidence match
• Best handled through Search, where scoring adds limited value

Troubleshoot missing matches

Cognism returns no result when no profile meets the required confidence level.

This can happen when:

• Input data is too general (for example, common names or job titles)
• Key identifiers are missing (such as email or company domain)
• Contact details have recently changed
• Input data contains errors and the threshold is too high

Returning no match is expected behaviour and helps prevent incorrect data from entering your CRM.

Understand API and platform differences

Results from the API may differ slightly from the Cognism web application (Prospector).

This can happen due to:

• Ongoing data updates
• Platform improvements
• Timing differences between systems

Consider API-specific factors

Check your entitlements

API responses include fields based on your entitlements, which are linked to your API key.

Core Contact and Account data is always returned. Additional fields depend on your subscription.

Stay within rate limits

All API requests are subject to rate limits.

• Maximum: 1,000 requests per minute
• Exceeding this limit returns: HTTP 429 “Too Many Requests”

Summary

Cognism matching logic returns Contacts and Accounts only when there is sufficient confidence.

• Every match is scored
• Strong identifiers improve match quality
• Specific data increases confidence
• Thresholds balance accuracy and match rate
• Low-confidence results are not returned

This approach protects CRM hygiene and ensures reliable enrichment at scale.