I recently interviewed Sebastian Jimenez, the founder and CEO of Rillavoice. Sebastian was hands down the funniest founder I’ve met – he was a prominent standup comedian during his time at NYU. Originally from the Dominican Republic, Sebastian came to the US for college. Graduating in 2018 with a Bachelor’s in Econometrics and Quantitative Economics, Sebastian claims to be the “class clown” of his grade. “I did a show for all the freshman at the time when I was a junior,” Sebastian recalls, “and I was doing seven shows a day, six days a week, just back to back. I learned more about business doing standup than doing Econ at NYU.”
Sebastian describes standup comedy as “the art form of thinking of something and testing it with a crowd”. This is strikingly similar to testing product-market fit in the startup world, a process that kills 42% of startups. Sebastian’s passion for standup led him to be a creative thinker and a great iterator. It also pushed him out of his comfort zone, which led him to find his calling in the startup world when he founded his first startup, Ballotbox – a tool for political campaigns to do door-to-door canvassing. “It turned out to be not a good idea,” Sebastian recalled, “because we realized, ‘oh my god, when elections end there’s no money.’” So, Sebastian, with his other two immigrant founders who were working on this startup to reform American democracy (none of them could vote, by the way, Sebastian pointed out), had to pivot to selling their software to companies that do field marketing. “We started working with a lot of CPG (Consumer Packaged Goods) brands to sell and distribute their products offline,” Sebastian explains, “and we noticed the big aha moment for Rilla.”
At this point, Ballotbox was doing pretty good selling to CPG brands – they were making around $100k ARR with only a team of three. But the problem Sebastian saw was that there was a massive visibility gap between the CPG brands, who sell their products to grocery stores and other distribution channels, and their end-consumers. “We noticed that field marketing brand ambassadors, the people who put on a red bull jacket and stand in Costco trying to get you to take a sample, was the only point of contact that the brands had directly to their consumers,” says Sebastian. After speaking to some of their field marketing managers, Sebastian realized that the CPG brands were having around two million face-to-face interactions with their end-consumers every month. Sebastian knew that large CPG brands are paying a lot of money to gather data on their end customers through surveys, Facebook ads, and focus groups, but what if they are able to capture the millions of direct interactions between their field marketers and consumers and create analyses of the conversations using AI. “That’s why we’re called Rilla, it stands for Guerilla Marketing, which is another term for Field Marketing,” Sebastian points out.
With this idea, Sebastian called up his friend Michael from Miami, who was working for a contractor for the Department of Defense doing signal processing at the time. Their friend from high school, Chris, who was working as a software engineer at Morgan Stanley, and Sebastian’s CTO from Ballotbox, were also recruited by Sebastian. Together, they started Rillavoice in October 2019. Starting with field marketers as their target consumers, Rillavoice had to pivot yet again when they realized that this go-to-market strategy made no sense. “The people who control the budgets for consumer insights are not involved at all with field marketing,” explains Sebastian, “so it was really hard to sell to field marketing people to use this product when it provides value to a completely different department.”
Sebastian and his team finally found their go-to-market strategy with retail, but then COVID hit. Despite the pressure to move online, Rillavoice refused to become yet another speech analytics company that serves online meetings and telecommunications. So they changed their go-to-market strategy once again to target CPG brands’ sales divisions, where the B2B salesmen were still going out to grocery stores during COVID to get their products on the shelves. Rillavoice was able to gain some traction pursuing this strategy and raised their first round of funding during COVID.
Flash forward to today, Rillavoice is a B2B solution that provides the first speech analytics tool for physical retailers to analyze the thousands of conversations that sales reps are having with customers. “We have four things for our ICP (Ideal Customer Profile): they have to sell face to face, they have to be sales-oriented, they have to have tablets and smartphones, and they have to be commission based,” Sebastian summarizes. Rillavoice uses a frontier Voice AI technology to help companies who fit this ICP understand the physical consumer experience, better train sales reps, and increase compliance with new campaigns, promotions, and products. Rillavoice’s customers achieve a 25x field training productivity, 1.5x conversion rate, and collect 2000x more consumer data. All of this comes from the core technology of Rillavoice.
Rillavoice’s AI was more challenging to develop than comparable algorithms that target virtual speech analytics. This is because in Zoom meetings, people use different input microphones and the process of identifying who is speaking becomes a lot cleaner. In the physical setting, there are many variables at play – background noise, multiple speakers using the same input medium, varying distance from the microphone, confidentiality, etc. Developing this technology also had more general barriers, one of which Sebastian calls “The Gossip Algorithm”. Right now, the salesperson goes into work, hits start on the Rillavoice app on the tablet/phone, and can just leave the app on for the whole shift. But what about casual conversations between coworkers? An exchange between the boss and the employee? Rillavoice’s Gossip Algorithm basically creates a “voice fingerprint” for each employee working at the store when they sign up and opts to not record any conversations between two known voices. This protects the employee’s privacy and decreases inefficiencies in processing useful interactions. Rillavoice also uses AI to protect the confidentiality of the consumers, where personal information such as phone numbers, addresses, and names can be blocked out.
Now, Sebastian is working on making the mobile app easier to use for salespeople and other UI/UX design problems. Rillavoice is also scaling rapidly to reach more departments and brands to capture data, which will help salespeople sell more efficiently overall. Rillavoice charges its customers using a SaaS licensing model, by hours of audio processed, per employee per month. They are seeing a very high retention rate, with some cohorts reaching Net Dollar Retention (NDR) of over 100%. Rillavoice has been growing around 15% week over week since they’ve narrowed down their go-to-market strategy.
“Ultimately we want to be the Google Analytics for offline sales,” Sebastian envisions. Rillavoice wants to make offline commerce as frictionless as possible, and for a young startup, it’s sure been through its share of pivots. This ties back to Sebastian’s background in standup, where he described his experience as “have an idea, iterate, iterate, iterate, and iterate again.” From standup to startup, Sebastian and Rillavoice take on a brand new segmentation of the speech analytics industry.