August 17, 2023

List Matching: How We Improved Our Match Rates by 81%

Clients often rely on m360 Research’s recruiting expertise to find market research respondents from their own database.

To facilitate list-based recruitment, we employ a matching solution to determine which contacts provided by the client are also present in our panels. The process typically involves matching entries based on certain attributes, such as names, email addresses, and other relevant data points. In the past, small name variations such as Anthony and Tony would go unnoticed. However, by implementing a new specialist tool able to catch data inconsistencies, we maximised list efficiency improving our match rates by up to 81%.
We run list matching during the bidding stage, as feasibility is often significantly impacted by whether these contacts are active members of our panels and have previously agreed to take part in market research. By launching our new advanced list matching tool with our industry-leading recruiting solutions, including dynamic profiling and custom recruitment, we are empowered to provide improved feasibility assessments to our clients, and a seamless experience for our panellists, resulting in fewer screening outs.

The Challenges of Traditional List Matching

The main challenge for successful list matching has been that contact lists rarely contain standardised data, and conventional solutions have been limited by this. Lists always have variations – spelling, abbreviations, and formatting – and different order of inputs would typically result in a non-match, even though records refer to the same ID.

Inconsistent Data

Dr J R Gonzalez
Univ of Central Florida

Jose Rogerio Gonzalez
Central Florida State University

Dr Gonzalez
UCF

Order of Inputs

Chan Li
PO Box 123, 45 Main Street

Li Chang
45 N. Main, Post office box 132

New Ways of Finding the Right Match

To mitigate challenges associated with list matching, we developed a new tool able to maximise list efficiency. Our new solution scans and assigns scores to each record, a much more nuanced approach to determining potential matches. The innovation lies in the possibility of making matches regardless of multiple variations and irregularities related to phonetics, abbreviations, non-phonetic similarity, transliteration, parsing and restructuring, misalignment, noise, synonyms, and typographic differences.

Phonetics

Phonetics

Bower, Bauer
Hernandez, Hernades
Muhammad, Mohammad

Abbreviations

Abbreviations

United States Postal Service
USPS
U.S. Postal Service
U.S.P.S

Non-Phonetic Similarity

Non-Phonetic Similarity

1, No. 1, First, One
1st Street, St, Str.

Transliteration

Transliteration

التأسيس
Altaasis, Incorporation

Parsing & Restructuring

Parsing & Restructing

Inc, Incorporation, LLC

Misalignment

Misalignment

500 Main St., S

Noise

Noise

Inc, Incorporation, LLC

Synonyms

Synonyms

Michael, Mike, Michel

Typographic

Typographic

Wilson, Wislon, Wilsn, Willsom
By the Numbers

We applied both solutions to a recent project:

  • The total number of active panel members matched increased 81% (from 411 to 744)
  • The total number of completes from matched active panel members increased 48% (from 39 to 58.
Unlike traditional solutions, our new specialist tool can find matches even amongst inconsistent data, and disorganised data. This transformation has not only addressed the challenges of list matching but has also elevated our project feasibility assessments and participant experience to new heights. As we look to the future, our commitment to innovation remains unwavering, fuelling our drive to deliver exceptional results for our clients. Share your specific criteria with us to see how we stack up. We commit to full transparency on our feasibility assessment.

Partner With Us

If you’ve been searching for solid partnerships to drive innovation, then look no further.

Latest Articles