Thought Leaders in Artificial Intelligence: Conversica CEO Alex Terry (Part 5)

Sramana Mitra: Yes. I’m not saying it’s not beneficial. I’m just saying that, from a technology and accuracy point of view, it would be easier to do certain things when the range of questions and the range of engagement is more constrained. I think if you’re getting into the realm of enterprise software, the range of questioning may become a lot more complex. But maybe not? You’re saying it’s not.

Alex Terry: You’re correct, but part of what we do when we go into new verticals is leverage, I think, 85 million plus interactions at this point in terms of training, natural language, and AI. We’re pretty good at understanding what comes back to us. The other secret sauce is how we write the outbound messages. We’re not boiling the ocean. We’re not trying to recreate Siri where people can just ask us anything. Since we’re typically responding to some stimulus, we get to control the conversation a little bit if that makes sense. If you think of the company name Conversica, it’s a play on conversation and conversion. We’re having conversations with people where we get to put some parameters around how that conversation gets started. That makes the technology problem quite a bit easier.

The other thing is like any AI, we’re not 100% perfect in our interpretation. I don’t think we would ever be able to spend enough money to be able to interpret 100.0% all the inbound replies, but we’re a high 90%. We’re very high in terms of our ability to interpret the inbound. We have a team of people to turn to whenever we don’t have sufficient confidence in the inbound reply. By the way, humans can’t always tell. I look at these messages. Sometimes, you can’t  always tell what people are saying. There’s a point of diminishing return in terms of the technology. While we’re always getting better, we don’t think we’ll ever get to 100%.

Sramana Mitra: What does it cost for an organization to roll out your sales assistant?

Alex Terry: We charge on a monthly subscription. It depends on the lead volume obviously. It’s priced to be cheaper than a human assistant. The typical startup package is around $1,500 a month. Our prices do vary somewhat by vertical because the use cases are different and the number of conversations are different. The ballpark is $1,500 or so per month. One way to think about it is it ends up being around a buck or buck fifty per lead. With volume, it gets cheaper.

Sramana Mitra: It’s not tied to the number of sales reps. It’s tied to the number of leads that your sales assistant is processing.

Alex Terry: That’s right. Although the number of sales reps and the number of leads do have a correlation. Organizations may hire one sales assistant for their team. The team might be two or three people or it might be 50 sales people. That’s one of the choices that customers have. Do they want to have one sales assistant name? Do they want two? That’s one of the choices that people have when they set it up. We try to be flexible. It’s largely based on the volume. Someone who’s processing 100,000 leads a month is going to pay more than who’s somebody who’s processing 500 leads a month.

Sramana Mitra: What does that range? Let’s say a thousand leads a month verus 100,000 leads a month, what is the range of price?

Alex Terry: From between $1,500 to $20,000 a month. I’m just giving you rough numbers. When you get to large volumes, we tend to have package deals. They tend to have special needs. They want unique conversations created.

This segment is part 5 in the series : Thought Leaders in Artificial Intelligence: Conversica CEO Alex Terry
1 2 3 4 5 6