Directly Adds Predictive Routing to On-Demand Help Desk Solution

By Paula Bernier March 22, 2016

Directly today is introducing a predictive routing that identifies what experts are most likely to be able to resolve when it comes to customer service tickets. It does that by identifying similarities between incoming tickets and past tickets that specific individuals were able to resolve.

The solution was designed to be used situations in which those people involved in resolving tickets are users of the company’s products. The Directly technology plugs into help desk solutions such as Oracle Service Cloud and Zendesk, or into a website, to allow for that.

Companies like Pinterest and FreedomPop use Directly to scale online support by routing customers’ questions to their expert users, who earn money or other rewards by resolving the questions. Unlike hourly call center workers, experts know and love the products already, and they answer questions on demand, where and when they want to. This method delivers significantly faster response times, customer satisfaction ratings of 92.5 percent on average, and higher resource efficiency, according to Directly CEO and co-founder Antony Brydon.

Now Directly, is backed by blue chip VCs including Costanoa, CrunchFund, and True Ventures, is adding the predictive routing feature to its solution.

Eugene Mandel, Directly’s lead data scientist, said sending the right ticket to the right agent is really an agent problem. Yet, he added, customers today expect very fast response times. So Directly is leveraging machine learning to do predictive routing. The system watches human experts to see which tickets they serve effectively and which they cannot. Predictive routing understands if a given question is likely to be addressed successfully by a given individual.

Businesses have long been working on how to more efficiently schedule and use the time of contact center agents and other employees and interest groups that can help customers resolve their problems and answer their questions. This is just one more way this is happening.

Edited by Stefania Viscusi