Feature Deep-Dive

Advanced Name Matching for Deceased Records

Standard keyword searches miss obituaries because names are spelled differently, nicknames are used, or the deceased went by a maiden name. Our multi-layer algorithm catches what manual searches—and Legacy.com—routinely miss.

Why exact-match searches fail for deceased records

Obituaries are written by grieving families, transcribed by funeral home staff, and published across thousands of different websites—each with its own inconsistencies. Names get shortened, maiden names reappear, and typos slip through. A system that only matches exact strings will miss a meaningful percentage of real obituaries for the names you're watching.

Our 6-layer matching algorithm

Each layer adds a different type of name-awareness so the system catches what the previous layer missed.

Step 01

Exact-match pass

We first check for exact name matches across all 2,500+ sources. If the obituary spells the name identically to what you provided, we flag it immediately.

Step 02

Phonetic & fuzzy matching

Names that sound the same but are spelled differently (Cathy/Kathy, Steven/Stephen, Bryan/Brian) are caught by our phonetic-similarity layer using Soundex and Metaphone algorithms.

Step 03

Nickname & diminutive expansion

A curated dictionary of 800+ common nickname pairs maps formal names to diminutives automatically: William↔Bill/Will/Billy, Robert↔Bob/Bobby/Rob, Elizabeth↔Liz/Beth/Bette/Lisa, and many more.

Step 04

Edit-distance tolerance

Typos and transcription errors in published obituaries are caught via Levenshtein edit-distance scoring. A one- or two-character variance (Smyth/Smith, Jonson/Johnson) still produces a high-confidence alert.

Step 05

Maiden-name & hyphenated-name support

You can provide alternate surnames (maiden, married, hyphenated) when setting up a monitor. We search all provided variants simultaneously, so you never miss a posting that used a different form of the name.

Step 06

Multi-factor confidence scoring

Name similarity is combined with location, approximate age, and family-member cross-references into a single confidence score. Only alerts above a configurable threshold are sent—drastically reducing false positives for common names.

How we handle the hardest cases

Common names, nickname variants, and non-ASCII characters are where most systems break down. Here's how we handle them.

"John Smith"

The problem

One of the most common names in the US. A basic name search returns thousands of results.

Our solution

We narrow with state, approximate age (±3 years), and optionally a spouse name. This reduces a field of thousands to a handful of high-confidence candidates.

"Mary Johnson"

The problem

Extremely common; often published under a married name that differs from records you hold.

Our solution

Supply both maiden and married surnames when setting up the monitor. We run parallel scans and merge results, flagging any match above threshold.

"James vs. Jim"

The problem

Obituaries often use the name the person was known by, not the legal name on records—completely missing exact-match tools.

Our solution

Our nickname dictionary automatically expands 'James' to include 'Jim,' 'Jimmy,' and 'Jamie' before scanning, so no manual guesswork is needed.

"María García"

The problem

Accent marks and diacritics are inconsistently applied across funeral home websites.

Our solution

We normalize Unicode characters during indexing. 'Maria' and 'María' are treated as equivalent, so accented and unaccented variants match each other.

ObituaryMonitor vs. Legacy.com vs. manual search

Legacy.com is useful for browsing. It's not designed for systematic name-matching across sources—and it doesn't monitor continuously.

FeatureObituaryMonitorLegacy.com
Fuzzy / phonetic matching
Nickname expansion (800+ pairs)
Edit-distance typo tolerance
Maiden / alternate surname search
Multi-factor confidence score
Common-name disambiguation
Diacritic / Unicode normalization
Continuous automated monitoring

Fewer false positives. Fewer missed matches.

The multi-factor confidence score combines name similarity, geographic match, age range, and family cross-references into a single number. You set the threshold. We only alert you when we're confident— so you're not chasing dead ends, and you're not missing the real thing.

  • Name similarity score (0–100) across all 6 layers
  • Location weight: city, state, or national-level match
  • Age-range verification (±3 years from provided birth year)
  • Optional: spouse / family-member cross-reference

14-day free trial · No credit card required

Who relies on accurate name matching

Related Resources