Dashbot Raises $3 Million from Bessemer Venture Partners – CEO Arte Merritt Interview with Voicebot
Dashbot.io closed a new funding round last month led by Bessemer Venture Partners. The round totaled $3 million and comes seventeen months after it’s initial funding of $2 million completed in August 2017 and some interim funding between the rounds of about $1.2M. In addition to Bessemer which participated in both rounds, the current round included earlier investors Samsung Next and Bertelsmann Digital Media Investments and new investors ff Venture Capital and Runa Capital.
The company says it now offers analytics services to “over 15,000 chatbots and voice skills” and has processed more than 60 billion messages. Funds will be applied to developing new data science tools and creating predictive algorithms designed to improve “chatbot response effectiveness.” Voicebot caught up with Dashbot co-founder and CEO to learn more about the recent funding round.
Why was now the right time to bring in another funding round?
Arte Merritt: We are big believers in conversational interfaces. We believe it is the natural way humans interact. With the advancements in AI and NLP, we are seeing enterprises increasingly adopt these technologies to increase user satisfaction, reduce costs, and, depending on the use-case, increase engagement, awareness, and monetization. As enterprise adoption is increasing, we see a lot of opportunities to provide tools, services, and reporting to enable brands to develop better user experiences in their chatbots and voice skills. We are building the team out more and building additional features, based on data science, to help customers with their initiatives.
Your blog post says the round was $3 million, but Pitchbook lists it as $4.28 million. Did you subsequently expand the round and accept additional funding?
Merritt: Our most recent round was $3 million led by Bessemer Ventures. We have raised $6.2M to date from a great group of investors including Bessemer, ffVC, Samsung Next, BDMI, Scrum Ventures, Rembrandt, R/GA, Recruit, and Runa Capital.
What is the breakdown of chat versus voice apps being used today?
Merritt: We are in an interesting position that we see quite a bit of data – over 63 billion messages across a variety of use cases since we launched a little less than three years ago. When we first started, customer interest from both enterprises and startups was primarily around Facebook chatbots. By the following summer, the interest changed quite a bit – brands were more interested in voice skills than Facebook chatbots. The industry continues to change as we see the interest from enterprises is skewing more to customer service chatbots. Voice is still popular, but tends to be coming out of the innovation teams at large enterprises.
You say in your announcement that you will use the funds to invest in new data science tools and predictive response algorithms. How will these new capabilities help your customers?
Merritt: Building an effective conversational interface is hard. It is difficult to know all the things a user may say to a chatbot or voice skill, or the way they may say them. It is important to look at the analytics to better understand how users are interacting and where the chatbot might be breaking down. To make the process easier and faster, we developed a number of tools, based on data science and machine learning, to help identify where the Natural Language Processing (NLP) is breaking down – including a phrase clustering tool that can help identify mis-handled and unhandled Intents.
We are working on additional tools to further streamline the development process and feedback loops to better optimize the NLP – and hopefully lead to a better experience for end users. These tools help our customers reduce costs and increase user satisfaction, and depending on the use case, increase monetization, conversions, and/or engagement.
You are now about three years into your Dashbot journey. What has surprised you most about where you are today and how has it changed the way you think about the conversational AI market?
Merritt: We are strong believers in conversational interfaces and we continue to believe this is the future of human-computer interaction. It is exciting to see how quickly the space is evolving and the interfaces are improving. As mentioned previously, we initially saw a lot of interest in Facebook chatbots, that shifted to voice skills, and now is shifting more to customer service chatbots. The space is still fairly early. As the AI and NLP continue to improve, we continue to be excited about the future possibilities and what enterprises and startups build.