We aim to provide better customer service through data. What this means we look at the questions, problems and concerns that customers have had in the past, to find patterns in these, and using insights from these patterns to provide better service in the future. The big challenge with this is that customers do not say the same words when describing the same problem.
This is the challenge of organizing and categorizing unwieldy data. The fact that we have a database of 10s of thousands of inquiries helps and one small insight recently produced a major breakthrough. In the inquiry form on the front page we have a field called “description of problem”. Until recently, every person was putting in some different combination of words to describe a cracked laptop screen. From this, on a whim, I decided to put “cracked screen” in the field ahead of time.
What then happened was that ~80% of inquiries did not change this. So then all I needed to know was the laptop model number to send the email back with the appropriate response. This allowed for computer generated responses for about 30% of all inquiries which gave the customer the info needed to make a decision in 30 seconds instead of 1 hour. This is a large improvement in quality of customer service.
We hope learn more from and better organize unwieldy data based on customer requests to have more similar breakthroughs in the future.