Machine learning is defined as the application of artificial intelligence (AI) that provides systems the ability to automatically learn and improve from experience without being explicitly programmed.
At Colt we’re starting to tap into machine learning to get us closer to our customers.
The bespoke machine learning approach Colt have developed enables analysis of large volumes of qualitative data with ease and, in our pilot, we analysed 45,000 comments from 9,500 customers.
We took a number of interesting learnings from the pilot which will be sharing with various teams throughout the business over the next few months.
Whilst speaking at a recent industry conference, The Future of B2B Research, I shared the platform with the audience and during a breakout session, a number of industry peers were curious about the idea of the ‘horizontal customer’ that has been unearthed by our machine learning pilot.
Businesses rely on a set of specialists to deliver a product or service with the business often organised in teams related to their expertise. Over time, the business shapes itself into a series of units servicing vertical customers with the analysis of customer intelligence often mirroring these verticals.
Machine learning is helpful to unearth feedback that is horizontal in nature. It is able to do this as it looks at the data points without any bias or preference, this effectively providing the customers view on the whole business, rather than a specialist area.
In the pilot, we identified the following horizontal customer truths:
- Customers want us to take ownership when things go wrong.
- Customers want us to communicate clearly.
- Customers want us to improve our speed of response.
- Customers want signposting on who is responsible for helping.
The pilot data range was from January 2018 to January 2019 so effectively, we were spotting patterns in old data with the issues above now having initiatives linked to them in order to solve them.
In the future, the aim will be to incorporate additional types of data such as inbound call transcripts and Colt Online ticket descriptions. We then will look to expand the scope of the tool to go beyond the survey data and work with a large range of customer experience touchpoints.
This is early days for machine learning but there are exciting times ahead.
Watch this space!
Noah Roychowdhury is the Head of Customer Intelligence at Colt