Customer Login  |  

Information Evolution is proud to be a part of the upcoming CrowdSourcingWeek Summit in Washington DC on June 15th. The one-day program will focus on the second wave of the crowd economy with domestic and international industry-leading professionals sharing how crowdsourcing is reshaping industries and transforming organizations today.

Open innovation, crowdfunding, government transparency, participatory journalism, are all part of the massive transformation going on across the world powered by crowdsourcing.

If you enter promotional code ‘CSWDC15’ on the Eventbrite registration page you’ll enjoy an exclusive 15% discount off of the already reasonably priced registration fee.

{ 0 comments }

posted by Shyamali Ghosh on May 15, 2017

by Matt Manning

When your company’s product is a database information service you are familiar with having “holes” in your offering. Must-have fields are populated 100% of the time, “need to have” fields are 95% complete, and there is a slippery slope of “nice to have” fields like URLs, phone numbers, and email addresses, that are progressively less populated.

There is a similar issue with “holes” in the universe of your coverage where databases purported to cover, for instance, every hospital in Canada except small regional clinics serving remote audiences or in-house clinics at large factories, or urgent care centers that aren’t government certified.

Welcome to the “fill in the blanks” zone.

The increasing reliance on apps, APIs, and web-based information services in the daily lives of professionals means that coverage gaps have become more apparent than ever before. Searches, for example, can’t return a valid record when one of the search criteria is missing. Size indicators are the most glaring examples of this “expectation gap” since even simple searches on all firms above a certain “size” will routinely fail to include big private firms that don’t disclose revenues and smaller firms that are reluctant to divulge head counts.

This is where estimates of any kind are better than nothing. These values, of course, need to be noted as estimates, but they really need to be populated 100% of the time. Head counts are easier to obtain via “inferential” data (i.e., hard data that strongly suggests facts like a company buying real estate indicating that it is planning to grow) and revenues can be fairly accurately projected with domain knowledge of given industries. An example of the latter is an assumption that if the largest firm in an industry sector publicly reports revenues of $X and a headcount of Y then it is safe to assume that this ratio will hold true for other firms in that same business, thus only requiring either X or Y to solve for the unknown value.

If the databases powering your information services have any gaps in “must-have” fields then it may be time to start estimating your way to a better user experience.

{ 0 comments }

posted by Shyamali Ghosh on May 4, 2017