Kenny Gunderman
Management
And David, on the hyperscaler question, first, just as a reminder, we've said this before, but I think it's worth reiterating that not all of these deals are the same. The hyperscalers are -- first of all, they're not all one single customer. They all have different buying patterns, of course, but also in some cases we're selling traditional waves or we may be selling traditional dark fiber IRU on existing network and it's just their typical network planning for current and future uses. And so we may be selling six or 12 strands of dark fiber, or we may be selling 100 or 400 gig waves in the traditional sense. I think you're -- and by the way, that's terrific business for us, and it's a growing part of the funnel, and it will continue to grow on a go-forward basis. And I highlight that because I actually -- we think once we start to hit the inference phase where people are using more of the AI at the edges where you can't -- where a lot of that data can't be cached and it has to be used real time, that's when a lot of the lit fiber on the edges will become even more valuable and I think that's when you'll start to see a lot of this business flowing through into MRR as opposed to now. You're seeing a lot of these lumpy type deals that we're explaining and it's either in bookings or it's not or it's a one-time sale or it's not. And I think that's the heart of your question. What's the sales cycle like for those types of deals where the hyperscalers are building new data centers, for example, in a unique location. And look, what I would tell you is, the fiber is not -- the fiber piece of it is not what's driving the decision initially. For the hyperscalers, it's procuring the land, and it's procuring the power and the grid capacity, and obviously building the data center. And then from there, the hyperscalers, I think, are looking for fiber providers that have network near the location or can justify building it off of an existing location. And so it goes back to my earlier question. It's not really a traditional sales cycle, if you will. I think the majority of the capital required for the hyperscalers to build these data centers is not the fiber. It's the data center, the land, and the power. And then fiber comes later. And so in some ways that's a good thing, right? Because there's a little bit less focus on getting the last dollar of economics and it's more about getting that fiber built on time and getting it turned up expeditiously. And I think in our markets, our core markets, we have an advantage when it comes to providing that.