Jeff Lawson
Analyst · JPMorgan. Please go ahead.
Well, thanks, Mark. Obviously, it’s very early in this game, so it’s hard to tell exactly how things are going to play out. We set out our vision for customer AI for what we think is going to happen, and I kind of said that not only is this going to become possible, I think it’ll become inevitable. And the key to a lot, if not, all of those things I talked about, is companies having a really good handle on all the data about their customers, right? So if you’ve got your data spread across all the different systems and sitting in all these different places and not aligned, it’s very dirty. It’s been really hard to actually put AI to use solving some of the really big things that I think AI will be able to solve for companies. And so the first order of business here is getting customer data in order so that as these AI use cases come to maturity, they have the raw information that they need to understand who they’re talking to and how you can start going about optimizing these customer interactions, customer relationships, and overall, like, the business and the front office of every company. And so that’s how we’re thinking about it today. Now, the other thing I think that’s super interesting in the world of generative AI in particular, I think that SaaS businesses that license per seat have the opportunity to be very much disrupted in this coming world, because I think companies will need fewer seats. I think that the things that AI is going to need to latch onto is essentially data sitting in systems, and that data is going to be really used in a usage type model. And so I think our business is, generally speaking, well set up for a world where companies may need fewer seats, they may contract the number of seats they’re using. They may not grow with the same number of seats, but the data, the backend systems, the processes, the workflows that are triggered by AI, that’s what really matters in this coming world. And so I’m very happy that Twilio is not in a position to largely be monetizing our service on a per seat basis, but rather we have a usage based model based on our communications business and even the data business as well.