Business data rarely has one perfect identifier. Names vary, phones are formatted differently, domains redirect, and several locations may share a brand. A defensible deduplication workflow normalizes fields, creates match tiers, and sends ambiguous pairs to review before records are merged or suppressed.

VISUAL LESSON

What you will learn

  1. 01Normalize common business identifiers.
  2. 02Create strong, probable, and review match tiers.
  3. 03Preserve branches and an auditable merge history.
Business records pass through domain, phone, and address matching into unique, duplicate, and review queues
Layered matching separates confident duplicates from legitimate multi-location businesses and uncertain pairs.

ILLUSTRATIVE WORKED EXAMPLE

Classify an illustrative duplicate-review queue

Exact domain + phoneHigh-confidence same-business signal
Strong
Exact domain onlyCould be multiple locations
Review
Name + addressNormalize both before comparison
Probable
Similar name onlyNever auto-merge on this alone
Weak
Illustrative example—not a benchmark. Replace the sample values with your own campaign, market, and measurement data.

PRACTICAL INTERFACE MAP

Move records through a match queue

Normalize01Create comparable fields

Preserve originals while standardizing domain, phone, business name, and address into helper columns.

Match02Apply rules from strongest to weakest

Assign exact, probable, and possible tiers; exclude already resolved pairs from lower-confidence passes.

Resolve03Review and log the decision

Choose keep separate, link as locations, merge, or suppress, and record the evidence and surviving values.

Conceptual walkthrough. Labels, controls, and availability can vary by account, region, plan, and interface version; verify the current screen before acting.

STEP-BY-STEP LESSON

Normalize → tier matches → review → preserve history

StandardizeComparable helper fields
ClassifyStrong, probable, or weak
ResolveMerge, link, or keep separate

THE LEAD ATLAS METHOD

Lead Atlas Data can build a current custom business-contact list around specific categories, markets, and locations, giving the deduplication process a clearly defined campaign scope.See how custom list research works ↗
01

Protect the source list

Keep an untouched copy with source, acquisition date, search scope, and record IDs. Create normalized helper columns rather than overwriting the public business name, phone, address, or URL that was researched.

Define what a duplicate means for the campaign. Two locations under one brand may be separate prospects, one account with several sites, or out of scope depending on the offer and ownership model.

02

Normalize stable identifiers

Convert websites to a comparable registered or chosen domain field after removing schemes, common prefixes, tracking parameters, and trailing paths as appropriate. Standardize phone country codes and digits while retaining the original display value.

Normalize whitespace, punctuation, common business suffixes, address abbreviations, and case in helper fields. Do not erase suite numbers, branch identifiers, or locality details that distinguish real locations.

03

Build match tiers

Start with high-confidence combinations such as exact normalized domain plus phone, or domain plus address. Use single-field matches as candidates rather than automatic merges when shared websites, call centers, or multi-location brands are possible.

For probable matches, combine normalized name, address, locality, website, and phone evidence. Assign each rule a tier and reason so reviewers know why a pair entered the queue.

04

Review branches and ambiguous pairs

Open the public website and compare location pages, contact details, legal identity where relevant, and the campaign's definition of a prospect. Mark same entity, separate location, related entity, unrelated, or unresolved.

Never choose the survivor only because it appears first. Prefer the record with clearer provenance and more recently verified fields, then preserve source-specific values and links to related locations.

05

Merge with an audit trail

Record the kept ID, retired ID, rule, evidence, reviewer, date, values retained, and whether the record was merged, linked, suppressed, or left separate. Re-run duplicate checks after imports and major enrichment work.

Deliverable: protected source file, normalized helper fields, tiered match rules, duplicate-candidate queue, branch review notes, survivor policy, merge log, and post-merge reconciliation counts.

LEAD ATLAS WORKBOOK

Turn this lesson into a research brief.

Apply “Deduplicate a B2B Prospect List by Domain, Phone, and Address” to one campaign before requesting or using a list.

  1. 01Market boundary

    Name the locations and business categories this decision applies to.

  2. 02Fit evidence

    Write the public signals that would make a business relevant enough to review.

  3. 03Exclusions

    List the business types, markets, and records that should not enter the campaign.

  4. 04Outreach use

    State who will review the list, personalize the message, and record outcomes.

THE TAKEAWAY

Use exact domain and phone signals as strong evidence, combine weaker name-and-address signals carefully, and preserve every merge decision.

OFFICIAL REFERENCES

Check the platform’s current instructions.

Platform labels, eligibility, and workflows can change. These official help pages were used to validate this lesson.