Familial DNA searching — a technique that searches CODIS not for exact matches but for close-relative matches — was proposed and debated in the law enforcement community in the early 2000s. California adopted a policy permitting familial searches of its state DNA database in 2008, and a partial familial search of CODIS was attempted in the Golden State Killer case. It did not produce a useful lead. The limitation was again population coverage: the close relatives who would appear in a familial search had to themselves be in the database, which required their own criminal history involvement.
The technique that ultimately produced the investigative lead — uploading the crime-scene profile to a consumer genealogy database — was different from familial searching in important ways. It searched a much larger and more genetically diverse population that included people without any criminal history. The matches it found were more distant — not close relatives but third cousins, fourth cousins, cousins of various degrees at a significant genealogical remove. But the database was large enough and the DeAngelo family’s participation in consumer genealogy was sufficient enough to produce usable matches.
The development of genealogy databases to the size needed for this to work was not complete until the mid-2010s. The technical standards for comparing DNA profiles uploaded to genealogy databases with crime-scene profiles required specific technical developments in how raw DNA data files were formatted and compared. The match process in genealogy databases works through comparison of single-nucleotide polymorphisms across large segments of the genome, a different technical approach from the shorter tandem repeat comparison used in CODIS.