The Failure of “Sample Data”

The Failure of “Sample Data”

by Yaron Gueta Posted on Jun 19, 2017

I work at a company that easily/seamlessly captures 100% of the digital traffic going through your web/mobile digital properties.  I ran into a competitor in a small deal who offered to collect only a portion of digital traffic at a lower price.  They had convinced the customer that with a sample of the data, one can make predictions that were true/accurate.   The “framework” of capturing certain details would be a way to predict outcomes.

Unfortunately, that is simply not the case.  There is never going to be a situation where a sample of the data reveals the true intent of the masses.  Collection of 100% of the data solves problems immediately that otherwise can result in total failure.   Turns out the “framework” is also deeply flawed and rooted in individualism.

For a few years, I was in charge of putting together the soccer leagues in my hometown.  Once the kids reached a certain age, the competition required “tryouts”.  Kids were graded by 5 objective judges and scores were tallied to determine the best players.  This is very much like a Voice of the Customer Survey that many folks utilize in their customer experience organizations.  Here is what happened EVERY YEAR.

50-60% of the kids would show up for tryouts.  Thus, we had a very large sample (by comparisons in technology to 10-20% of my competitive digital insight solution providers).   Also, by comparison we have now used 2 tools (VOC survey and Data Capture platform).  The resulting seasons were a weekly headache of complaints about the uneven-ness of teams and lop-sided victories…..exactly what we were trying to prevent.  Total failure.  Had I known the strengths of the other 40% of the kids, Had I driven the committee to agree on the framework by which the kids would be judged things would have turned out better.  BUT, I was skewed by certain surveys and forced to fill in the blanks in order to get the work done.

Sample data may be useful in limited situations, but in all situations, having 100% of the data is going to yield the results that empower the most positive impact.  There exists so much disruption in every industry coming from the digital space – and here is a great example of why that is.  If I go into a bank and complete a transaction, it costs the bank $70 per transaction but if I do that same transaction online it costs them $0.07.  That is 1000% less per transaction and banks alone have millions of transactions every month.  There are $billions of dollars in opportunity out there.  The number one priority is having all the data to drive the decision-making process and not have to “fill in the blanks”.

At Glassbox, we automatically capture and INDEX (so you can easily search it) all the digital data and provide various interfaces to get things done immediately such as find/fix web/mobile app problems, analyze customer behaviors/patterns of exactly where they are going and what they are doing, data they enter, time they spend hovering, etc. on your properties.  Why did they not convert a transaction?  We have the ability to find, in 2 seconds or less, any digital session so call center agents can walk through and upsell frustrated customers in less time, converting the psychology of the customer from frustration to joy.  So legal/compliance can have the undisputed evidence required for litigation.  Our product is also exact video replay as if you are looking over someone’s shoulder migrating through the customer’s digital experience.

Sample data is a Pandora’s box that creates more work and can decrease revenue, increase cost and increase risk.  The outcome of “wing and a prayer” is not acceptable in today’s marketplace.  Having 100% of the data opens up use cases such as garnering a 360 degree view of the customer, retroactively analyzing different environments down to the customer or specific server/API level, and being able to make immediate, confident, data driven decisions.  Our platform is best of breed, as noted by our Enterprise use in some of the biggest banking/insurance/retail/hospitality companies.  This theme of lacking 100% data capture is pervasive in many business problems from both the IT side and the Marketing side.  My next post will continue down this track with real world examples and better positive outcomes through newer technologies.

 

Sign up for insights

Sign up for insights