Second Nature uses natural language processing and subject matter experts to build sales skills and product knowledge.
Everyone knows Alexa and Siri. Another assistant just got a significant chunk of money to make her name just as familiar. Jenny is a virtual assistant who specializes in training salespeople. Second Nature, the company that created this conversational robot, announced a $12.5 million round of investment Wednesday, Jan. 11 to build out the conversational artificial intelligence service.
This Series A money came from signals Venture Capital, StageOne Ventures, Cardumen Capital and Zoom Video Communications, Inc. via the Zoom Apps Fund. Second Nature has raised a total of $15.5 million.
Ariel Hitron, chief executive officer, and Alon Shalita, chief technology officer, founded the company in 2019. Hitron said his professional experience in both product marketing and sales inspired him to build the service.
"When I worked in product, I thought that the messaging sheets and the battle cards were the best thing ever," he said. "Then I switched sides into a sales leadership role and realized no one talks like that."
Second Nature's platform uses conversational AI to coach a salesperson and improve selling skills as well as product knowledge. A salesperson in training runs through a simulated sales call with Jenny serving as the customer. The video conversation is recorded and each trainee gets a score on his or her performance with advice on how to improve various elements of the pitch.
Hitron said people working for multinational companies from more than 54 different countries have used Second Nature. Customers use the virtual trainer when rolling out a new product or new corporate messaging and to support a sales certification process.
"With Zoom, their motivation was launching a new product," Hitron said. "They wanted everyone to speak proficiently about the phone product."
Colin Born, head of Zoom Ventures, said in a press release that the venture arm of the company decided to become a strategic investor based on firsthand experience with the software.
"After more than a year of our global sales team training and certifying on Second Nature, we truly understand the enormous potential in AI-driven sales training," he said.
Hitron said that the Second Nature training platform combines sales training and product marketing education in one place.
"What happens on the sales call is that you have maybe five minutes to gain the trust of a customer, and they don't distinguish between your skills and your knowledge," he said.
In addition to moving all sales coaching online, the pandemic erased the in-person relationship building element of sales. The loss of that socializing time meant that salespeople have to be ready to spend most of their time talking about the product, Hitron said.
"That five-minute part goes very fast and then you have to get to the point and you have to be very proficient on the product and know the case studies and the benefits of the product," he said.
Ian Campbell, CEO of Nucleus Research, said AI-based coaching has the potential to lower the burden on managers while increasing the effectiveness of coaching by not only being more targeted to the salesperson's needs, but also delivered on the salesperson's schedule.
"By increasing the efficiency of sales training, and ultimately sales productivity, Second Nature is addressing a problem that can deliver in significant long term ROI for its clients," Campbell said.
A training solution built to scale
Nancy Nardin, founder of Smart Selling Tools, said that Second Nature's ability to scale sales training is one of its strengths.
"You can create some scenarios once and then you get to use those with all of your reps," she said.
Second Nature also makes it easier for managers to coach a large team, Nardin said.
Nardin's Enterprise SalesTech Landscape shows how crowded the field is for sales coaching software but Jenny is another competitive advantage for the company, according to Nardin.
"Second Nature can say, 'We can give you a real-life situation so it feels like it's real life, not like you're practicing," she said.
The standard approach to sales training is to practice internally with other team members or practice on real clients, according to Nardin. That means a sales rep is practicing her pitch on live leads.
"If you're learning, you're burning the opportunities that you had because you might not be ready yet or you might be saying the wrong thing," she said.
Nardin said sales managers have to decide whether to let people learn as they work with customers or find a different training method.
She said that the sheer volume of competitors could make it challenging for Second Nature to gain ground.
"One company might prioritize sales coaching and another could say send more emails, so from that perspective they're competing against all solutions, not just the ones in their space."
Natural language processing and conversational AI
Second Nature combines off-the-shelf tools to handle the basics of natural language processing with internal modules that can identify these elements of a human conversation. The platform relies on both training data and best practices about how conversations work to accomplish that.
Hitron said that Jenny has to navigate more sophisticated conversations that most conversational AI platforms conduct, such as booking a table or buying a specific product.
"It's an open conversation but also semi-structured so Jenny has to know when to interject and when to guide a person," he said.
The virtual trainer has to distinguish rhetorical questions from an actual one.
"Jenny also has to know based on tone, intonation and pauses if I still have something to say or if I'm done," Hitron said.
The current product is designed to work for sales teams but it could be adapted for other groups, Hitron said.
Jenny's secret to success is working with subject matter experts at a company to understand "what good looks like" for a particular salesperson. When SAP launched a new sales simulation with Second Nature, several SMEs taught the virtual trainer about the new product and how to work with salespeople.
"As the training continues, Jenny continues to learn based on feedback from the reps," Hitron said. "A manager also could come in and say, 'This is a good response,' and add that to the definition of good."
How the sales simulation works
A sales rep starts the training simulation with a 15-minute conversation with Jenny. The rep has a few goals for the conversation that can cover product information as well as selling tactics. Jenny shows up as a static image while the trainee uses a video camera.
The rep gets an overall score as well as individual ratings for style and knowledge. A transcript of the recording highlights opportunities for guiding the conversation in a particular direction. The analysis also tracks how many filler words a person uses in the conversation as well as the energy level. The trainee has the option to run through the simulation multiple times before submitting it to a manager for review. Managers have the option to score the session as well.
Individuals also can provide feedback to improve Jenny's overall understanding of the topic and the algorithm's assessment of conversations.
Salespeople can use the analysis of the conversation to pinpoint the timestamp where they gave a good answer. That feedback becomes part of the training process for the algorithm, Hitron said.
- AI in closed-loop manufacturing can benefit edge computing systems: 4 things to consider in IIoT
- Bosch tests deep audio AI developed on the ISS to detect asthma in kids
- Our most technical people are down on AI ... and that's a good thing
- CXOs: Are we ready for AI to assist human decision-making?
- Why companies should use AI to fight cyberattacks
- How much can we trust AI? How to build confidence before a large-scale deployment
- AI trends for 2022: Four predictions from experts at Sony