Customer experience is what drives a lot of business operations today; it's the thing that ensures customers are sufficiently satisfied with an overall experience to keep coming back, time and again, to that same business. Making a better customer experience isn't always easy, and Zendesk recently announced its new Satisfaction Prediction tool, designed to help users find out how likely a customer is to be satisfied, and more likely to return.
Satisfaction Prediction uses machine learning capabilities to analyze individual tickets and reveal how likely that ticket is to end up with a good or bad rating. It's analyzed over two million interactions so far, and in the process, learned plenty about how a business responds to tickets of certain natures. Currently in use at companies like Easy Taxi and even Pinterest, the system is carrying out high-volume analysis and even replacing dedicated team members who did that job at Pinterest.
The system takes a variety of factors into consideration, ranging from the text involved in the trouble ticket, the total number of replies involved and the total time on the ticket to create a model that demonstrates how likely that ticket is to end well for the customer, and therefore, for the company. Dashboard analytics make insight into the overall process simple, and the whole thing can be presented in real time to make responses more timely. Those interested will be able to pick it up immediately, though it seems to be limited to those who are on the Enterprise plan that get at least 500 satisfaction ratings in a month.
Ovum Research principal analyst Aphrodite Brinsmead offered comment: “ With Satisfaction Prediction, Zendesk customers do not need to employ a data scientist or worry about gathering relevant data; the feature simply predicts and identifies more sensitive queries, prevents customers from churning, and improves satisfaction rates.”
It's a fair demonstration of the power of analytics; by knowing how other efforts have turned out in the past, organizations can get a handle on how similar efforts will proceed in the future. It's not a guarantee, of course, but the system makes a good point. When a ticket comes in with a lot of verbiage and a long wait time, and that ticket generally ends badly, it's a safe bet that tickets that come in looking like that in the future will also end badly. It's not exactly good news to hear that it already seems to be costing jobs at Pinterest, though more than likely that agent was just moved elsewhere.
In the end, however, a tool like Zendesk's is likely to make companies better and improve the chances for repeat business. That's good news no matter how it's looked at, and also improves the likelihood of future hires made down the line. A business that works better is more likely to survive long-term and continue contributing to the overall economy.
Edited by
Stefania Viscusi
