Cogito raises $5.5m to monitor the "tone" of call center workers
“This call may be monitored for quality assurance purposes.”
Every day millions of miserable human interactions begin this way, usually pitting a customer recently escaped from “voicemail jail” against a weary phone professional running through a checklist on the screen in front of them.
Today a cloud software outfit from Cambridge, Mass. called Cogito ("I think...") announced it has raised a $5.5 million Series A from Romulus Capital and Salesforce Ventures to fund its mission to improve this experience for all involved.
The Cogito solution is to pass the audio signals of the calls through voice analysis and behavioral models to give the agents and their supervisors real time feedback in a dashboard on their screen.
“The technology is based on behavioral analytics. We can analyze all the richness of the human voice, not the words themselves but things like pitch and tone and texture and pace and overlapping – all the rich components of the human voice – using that to understand things like human intentions,” said Steve Kraus, Cogito VP of Marketing, by phone.
The Cogito dashboard displays a continuous reading of those human qualities, giving alerts and prompts to the agent so that they can adjust their tone and approach in order to sound more empathetic and develop better rapport.
“The beautiful thing is that on 100 percent of those calls we are gathering information. Traditionally, agents have these one-off, once-a-month review sessions with their supervisor, now the agents can be much more involved in the process,” said Kraus, “we can provide objective feedback on 100 percent of calls.”
That’s a promise that will surely be music to the ears of call center workers, already amongst the most measured and monitored employees in the world. Now even their tone can be analysed: One imagines a constant feed of digital concern trolling: U mad, bro? You seem stressed. Y u mad tho?
Still, there is at least real science at work here. Cogito began its life in the MIT Media Lab, before spinning out in 2007. Since then it has been developing its models and underlying architecture, first validating the products of this research in pilot programs funded by DARPA and the National Institutes of Health in studies that attempted to detect depression and other mental health disorders in the voice signals of patients and veterans.
That may also come in useful in call centers, reports over the years have claimed phone professionals exhibit a high incidence of emotional fatigue. And customer service is big business. For over twenty years, the world’s biggest companies have been cutting costs by outsourcing this work, resulting in a substantial industry. In the Philippines this sector, which falls under the rubric of Business Process Outsourcing, is the fastest growing in the entire economy. It accounts for 10% of GDP and has grown fivefold to some $15 billion in revenues since the early 2000s, employing over one million people. Roughly a third of the estimated 13 million global call center employees work in the United States.
Low-hanging fruit has been picked off with self-service mobile apps, speech-recognition and chat interfaces, leaving only the thorniest problems to the humans in headsets. Increasingly, these remote workers are the only point of human engagement between a business and its customers.
Though most of these millions of calls are monitored, that usually means the interactions are sampled randomly or passed through speech recognition software to detect key phrases, later to be used in call center agents’ performance reviews. Customer satisfaction is measured by asking “were you satisfied” or some variant thereof, assuming the customer hasn’t already slammed the phone down in a paroxysm of existential angst.
Cogito’s Dialogue product is compatible with Avaya – one of the big software suites for call centers in the market. Kraus said that Salesforce is a “channel partner”– “they have so many deployments and this is a nice way for us to get into those deployments with them.”
For now, Kraus says the company has its hands full “penetrating into the customer service space,” by investing in sales and marketing. From there, he said it’s natural that the technology will spread into other parts of their customers’ businesses. Will Cogito’s sales force be utilizing their own product in that process?
“Yeah, our guys will use it,” he said.
Throughout our phone call yesterday, Kraus seemed interested and engaged. He began his answers with validations – “that’s a good question” – and paused before speaking to make sure I had finished. He spoke in a variety of tones depending on the content of the conversation, ranging from affable to informative. It all seemed very natural, spontaneous, and authentic.