The top tech companies have heavily invested in voice UIs: Amazon’s Alexa and Apple’s Siri are now the face of voice-powered artificial intelligence. It makes sense, as I've written before, since voice is likely the most natural, organic and simple UI for a post-mobile era.
Naturally, secondary voice-based services will now emerge: Voice-based SEO, for instance. But that's just on the consumer side. Enterprise voice data must be studied, too, which is why AI-based voice analytics platform VoiceOps has just left stealth to a seed round with the participation of Accel, Founders Fund and Lowercase Capital.
VoiceOps works by transcribing phone calls and using their artificial intelligence engine to parse the information in order to provide insights such as an individual employee's core skillset and any week-over-week trends.
“VPs & managers now know if their sales reps are highlighting benefits or just feature dropping on their calls,” the company explains. “Asking probing questions, if they’re making upsell attempts, or asking for the close at the appropriate moment. VoiceOps then compares individual performances to team averages, and tracks progress longitudinally. VoiceOps is doing in an hour what used to take managers days.”
Here's my interview with the co-founder and CEO of VoiceOps, Daria Evdokimova. In it, she discusses the company's business model, its foundation in cutting-edge machine learning algorithms, and what lessons any entrepreneur can draw from the company's challenges and successes.
Can you explain what VoiceOps does?
“In its simplest form, VoiceOps is a call analytics platform. Designed for enterprise sales teams and leveraging a form of AI (natural language processing), we’ve built a product that does in an hour what currently takes sales managers days – understanding what their team is actually saying on the phone.
We transcribe calls, parse at better-than-human quality, and then break down those conversations into core skills. That helps us analyze what behavior is most successful (e.g. how many probing questions to ask on a call, what's the value of describing benefits vs. features, when to time an upsell or close attempt, etc.). Once we understand what works, we help managers scale that out, improving performance across their entire team.
In short, we help sales teams close more deals.”
What makes now the right time for an audio data AI?
“Good question, Adam. Part of it is that the nature of sales teams is changing: while inside sales is growing, attrition rates for salespeople keep going up, which means that there's more need for efficient training on what reps should say on the phone. Every other part of the sales stack (email especially) are already data driven, but sales reps spend most of their time on the phone, so it makes sense that it's the most important activity to have analytics on.
The other very important piece is the more recent advances in Machine Learning – it's only now becoming possible to have a large scale voice analytics operation because of the recent advancements in speech to text, machine learning algorithms, and the cost of computing power becoming extremely cheap.”
What challenges have you overcome in your path to cofounding this company?
“I think in general the hardest part in starting a company is building something valuable enough that people or companies are willing to pay real money to use it.
From a very early stage, we were extremely focused on not pricing ourselves out of business, which meant we needed to build real value into our product, i.e. the output needs to have an impactful return on investment. For our users, saving time by automating and providing accurate data for sales coaching is great. But by improving core sales processes we’re able to show an increase in metrics like successful close attempts and revenue. That’s where we become indispensable for our customers.
Understanding how we can provide the most value to our customers, and then executing on it will always be the challenge we’re most focused on.”
Any startup lessons for any young entrepreneur who might be interested in a similar journey?
“For anyone considering starting a company, I'd absolutely suggest finding co-founders that you trust and that accel in the areas you’re weakest. That means first understanding that you can’t do everything, nor should you try. I’m extremely lucky in that regard — both (cofounders) Ethan Barhydt and Nate Becker have put in heroic efforts to get us where we are today.
The other piece of advice I can give is to be the expert in whatever space you’re attacking. So many parts of starting a company are sales oriented, whether it’s recruiting, fundraising, or actually selling your product or service to clients. Having a passion for what you’re doing helps with selling, but being the most knowledgeable in the room about your subject matter builds credibility, and having credibility as a first-time entrepreneur is no easy feat.
In other words, make sure no one else knows more than you about your business.”
What’s in VoiceOps’ future?
“For us, our wedge in the door is the call analytics platform we’ve built, which is essentially a very powerful sales enablement tool.
However, almost every customer we’ve onboarded has rolled out VoiceOps to other teams within their company. Eventually our goal is to serve every single function of a company that relies on calls to be successful, be it sales, support, customer success, or compliance.
Eventually, we believe we can become the hub of all voice data for the enterprise.”