On January 28, GV (Google Ventures) led a group of VCs that already included Andreessen Horowitz (A16Z) and Citi Ventures and others in a $75 Series C round of financing for Pindrop. That brings the company’s total equity financing to $122 million and, even though the company is private and doesn’t disclose its top-line revenue, we believe it to be a very generous multiple of annual sales.
Pindrop calls its proprietary, core technology “phoneprinting.” It does not literally capture or measure characteristics of the devices that originate phone calls. Instead, it analyzes the line quality and audio characteristics of a phone connection after it has been answered at a contact center. It gauges dozens of different factors that serve as indicators of where a call originated, whether it is an IP-based call, traditional switched circuit or wireless. If anything is amiss (for instance if Caller ID indicates origination in the U.S., but a background hiss or echo indicates the call originates from overseas), the call is flagged and can be routed to the company’s fraud department.
Vijay Balasubramaniyan, the company’s co-founder and chief executive officer has explained that the company has seen impressive growth among major money center banks and card issuers. New funds will be used to bolster R&D and ramp up sales and support efforts in the U.K. and Europe. Yet, based on our tracking of enterprise spending interaction analytics in the contact center, it is difficult to see how audio-based fraud detection (at base, a form of call analytics) will justify such a hefty investment.
GV and its fellow investors clearly have higher expectations for Pindrop. Indeed, Gene Frantz, a GV Partner told CNBC viewers that “any company that interacts with consumers via the phone has a need to combat fraud using this type of technology.” Both Frantz and Balasubramaniyan teamed up to use the term “Voice Fraud” to describe the spectre that they are fighting, with Balasubramaniyan asserting that fighting it represents a $20 billion opportunity — attributing it, in large part, to savings on agent expenses.
While Frantz was talking, CNBC took the liberty of displaying a slide that might depict the real upside to voice-based fraud prevention. Under the heading, “Voice-Activated Future: Opportunities & Challenges” a bulleted list read as follows:
— Rise of Intelligent assistants: Siri, Cortana, Google
— Growth in voice as IoT interface
— Voice-activated car control tech at CES
— New opportunities for fraud and theft?
In contact centers revenues for Pindrop and its peers are driven by the number of calls and percentage that involve fraud. In the world at large, revenues for fraud-prevention solution providers has yet to be defined. At some point fraud prevention must involve user authentication. At that point revenues may be driven by the number of devices or driven by the number of enrolled users.
Voice Fraud Isn’t the Problem, Voice Authentication Is
Throughout the discussions on TV and in the trades, speakers used the words “phone” and “voice” interchangeably. Pindrop, which does a great job of preventing phone-based fraud in contact centers, is not really in the business of detecting fraudsters or imposters in the wild, meaning on smartphones, in cars or when talking to the billions of “sensors” on the Internet of Things (IoT).
One of Balasubramaniyan’s key talking points is that voice is destined to be the primary user interface (UI) for the connected devices on the IoT. If that is the case, voice-based authentication will be the most natural way to provide secure, trusted links to cars, kiosks, consumer electronics and smart appliances. Pindrop, as it is positioned today, is not well-positioned to battle Voice Fraud on the IoT. To move in that direction, the latest $75 million investment will be money well spent if it is used to make voice-based user authentication ubiquitously simple and effortless.
Categories: Conversational Intelligence, Intelligent Authentication, Articles