Every good adventure begins with a quest, and our story is no exception. A long time before Steve Tootill became the CEO of 360Science, he was consulting to a life insurance company where he was presented with a challenge : “Steve, how can we be absolutely sure we don’t have any duplicate customer records?”
It didn’t take Steve long to realize that the available matching solutions were rudimentary at best and could only match on name and address data, not even close to all the other information a life insurance company has about its policyholders and lives insured, such as date of birth, payment details, social security number etc. But every single item of this data was prone to error.
If the effort to rid their database of duplicates was to prove successful, it seemed buying a solution wouldn’t be an option, it would have to be built. Spotting an opportunity, Steve suggested to his client that he would develop a solution that would dedupe their customer data using all the key information they knew – and he would just charge them a package price in return for them acting as a sounding board during development and as a reference site. This arrangement enabled Steve to ensure that what he built was:
• More flexible than anything else available
• Developed from the outset as a package, so it would be easy to upgrade
• Accessible, both in terms of simplicity and pricing.
When it came to accuracy, Steve’s client was pretty happy, but there was one thing that was causing a lot of manual review – the popular fuzzy matching algorithm for names that Steve had used, Soundex, proved to be totally inadequate for catching phonetic similarity. In fact, there wasn’t an algorithm available that could truly map the way names and words are pronounced. For help with this dilemma, Steve turned to his father, Geoff Tootill.
