I have written a lot about customer experience (CX) recently, as it has become a growing area in which Splunk’s own customers are looking for support. Reflecting this increased interest, a number of the presentations at the Splunk annual user conference, taking place later this month, will be focussed on improving customer experience. Sessions of particular interest will be:
- Conquering Perception With Data – A Story of Increased Customer Satisfaction, Tuesday, September 26, 2017 | 12:05 PM-12:50 PM, Hagop Hagopian, Sr. Product Manager, Charles Schwab
- Splunk and Machine Learning for Sales Efficiency, Wednesday, September 27, 2017 | 12:05 PM-12:50 PM, Chandra Vaughan and Michael Cormier, Ferguson Enterprises
- Improve Customer Satisfaction by Understanding User Feedback with Splunk Machine Learning Toolkit (MLT) and Splunk DB Connect, Wednesday, September 27, 2017 | 4:35 PM-5:20 PM, Sebastian Fernandez, Digital Analytics Manager, LATAM AIRLINES GROUP
We will be showcasing the latest iteration of our CX demo at the event, but in the meantime, you can learn more about the demo in my blog ‘Linking operational data with customer feedback to drive improved customer experience’. Its new feature is the ability to bring data from Enterprise Feedback Management (EFM) platforms into Splunk. We have built an EFM integration with the Maritz CX platform, which we will be demonstrating for the first time at .conf2017.
From a technical perspective, we have done this by setting up a TA using Maritz CX’s API to bring survey data into Splunk. Via this, we can ingest data from any survey, including both open and closed responses. Our demo now includes dummy data created by Maritz CX for a fictional airline, an extract from which is shown below.
Those familiar with customer feedback will recognize the structure of this data, although the format in which it is presented here will be slightly different than you are used to. It still includes:
- Survey ID
- Respondent ID
- Question ID
- Response (value)
In this example, Splunk has indexed JSON data, which is our equivalent of storing data into rows and columns in a structured database. Although this data format may be slightly different to traditional market research, the analytics generated will be familiar. As an example, here is a dashboard showing survey NPS taken from our demo.
The key benefits of being able to import EFM data into Splunk are:
- The ability to analyse customer behaviour (what happened) alongside customer feedback (how it made consumers feel); and
- Where unique customer IDs are present, we can join customer feedback with individual customer interactions, or end-to-end (omni-channel) customer journeys.
This has significant implications for the any company looking to identify CX improvements. Specifically, it will enable these companies to:
- Understand the impact of what happened during a customer interaction / journey on NPS (or indeed any other KPI);
- Extrapolate any insights generated from the sample of customers giving feedback to the population of customers interacting with the brand having the same interaction / journey; and
- Use alerts (with Splunk’s inbuilt alerting capabilities) to reconnect with individual customers who may have had negative experiences, including those who did not provide feedback.
I think that the ability to analyse operational performance alongside customer feedback will allow our customers to make big strides in improving their customer experience. This will translate into a competitive advantage for those who can harness their data, the most effectively.