Arista Networks, Inc. (ANET) Earnings Call Transcript & Summary

Thank you, Ita, for being with us today. I don't think I need to introduce Arista for anybody here. So I'll just jump right in.

Just I have a list of questions obviously here, but I'd be remiss if I didn't ask about NVIDIA's commentary this morning in terms of a negative preannouncement. This has nothing to do with NVIDIA. It's more about -- they've lined a lot of things, gaming, which has nothing to do with you guys whatsoever, but they did touch on data center. And given that you're the bellwether in terms of data center would be wanted to just maybe ask if you have a couple of comments or 2 on that.

Yes. I mean I'm not sure I have a ton to add at this point. I mean we just saw the announcement this morning. And again, different drivers, obviously, they seem to be very focused on gaming and the impact was coming from that part of their business. I think like we've talked about on the conference calls and stuff, we really haven't seen any softness across kind of our customers and our customer set, maybe that starts to help on the supply side a little bit, but we haven't seen that yet either. So I think we'll just have to watch and see kind of where things go from here.

That's actually -- it's actually pretty comment about supply chain. We'll get to that. But maybe just backpedaling, and thank you for addressing that for us here. With regard to the macro, again, maybe this is a bit redundant from the most recent quarter, but you guys have been doing great in terms of managing the supply chain. So those are kind of 2 separate questions. Maybe we just talk on the macro. What should we be looking for in terms of maybe your macro exposure? I mean enterprises early or in the earlier stages of growing there, so maybe there's nothing there. Cloud has secular growth. But where is Arista kind of exposed in terms of maybe potential macro exposures?

Yes. I mean, again, I'm not sure that we've seen anything different in the business to date, right? But obviously, we're cognizant of all of the same signals and drivers that everybody is talking about. The model is pretty resilient. Again, if you have a long and deep recession then it's going to impact everybody. I think when you look at the pieces of the business that we have, we're in the enterprise space where we're focused on larger enterprises. We have -- tend to have maybe longer views of investment cycles. We saw certainly with COVID and stuff that were reasonably resilient there as we went through that cycle. But again, it depends on severity, obviously. But I think we have seen some resilience from that piece of the business that focus on kind of larger enterprise helped a little bit in the past. On the other -- on the hyperscale end of the business, I mean, they certainly have the ability to continue to invest. They have come off of a period where perhaps they were -- they had underinvested. There's certainly a very acute focus on supply and meeting their supply needs over the next period of time, and they are giving us more visibility because of that. But again, I think when you think about a recession or you think about a macro impact, it really depends on the longevity and the severity of in terms of what impact it would have on the business.

That's fair. And you mentioned the unique position you have with cloud titans we're well aware of that and supply chain. Are there any -- we've heard from some of your peers that the supply chain might be marginally getting better. Again, this is not all clear sign, I think maybe there might have been some comments on the most recent call. Can we just maybe talk about any glimmers or any pockets of loosening or any signs of improvement?

Yes. I mean we're in an odd situation where we did a lot of work to, I would say, add manufacturing capacity and you've seen our purchase commitment, so secure supply of a broad set of components. And I think all of that's gone reasonably well. What we're struggling with and have been certainly with for the last couple of quarters is really more these decommits of kind of smaller piecemeal parts, which is harder to predict and honestly harder to resolve in some ways, right? We're -- you saw us -- our guide for Q3 kind of encompasses some improvement from a shipment supply perspective. We had line of sight to that, obviously, when we guided. And then we'll see, right? I think it would be good to see some continued momentum around some of those decommit parts, et cetera. then we could start to feel better about kind of gaining some ship momentum. But I think for now, we're just taking kind of a quarter at a time until we see that, but we did see some improvement. And those are really the gating factors for us right now -- are those piecemeal parts?

Yes. I definitely don't want to be reaching too much with a hope here, but you definitely indicated the better shipments, less deferred revenue drawdown, et cetera, in terms of the third quarter. The follow-up question might be a little redundant, but you -- because you have been very lean forward in terms of the purchase commitments. Is it -- is that improved shipment view on 3Q, a, like more due tied to the purchase commitments that you've made in the last 6 months. B, more because of you're seeing a little bit of an improvement in those decommits or combination of both?

It's really -- three folds to make it work.

[ I should maybe give them to see ].

We are kind of obviously leveraging the investments that we've made. And then again, getting some relief on those decommits is really critical to be able to kind of improve shipments for customers. But it's both. It really is both.

Let's just shift gears then. Thank you for that on the macro on supply chain. On competition, we were talking just moments before in terms of you coming in here that the bellwether and cloud and cloud titans is your market, but you entering the enterprise market, do you want to just talk a little bit more about competition and how -- I guess, from related to competition go-to-market for you all? And what's kind of changing at Arista investment-wise and how should we kind of watch you guys in the coming years in terms of competition and expanding your TAMs?

Yes. I think, look, we started investing in enterprise probably, gosh, 3, 4 years ago in a more, I would say, a robust way around kind of just the direct sales go-to-market, right, and starting to really build out an enterprise coverage model, if you like, expanding kind of geographically, expanding kind of the teams across the different geographies. And we've been consistently growing kind of that footprints, plus or minus 30% has been our goal, right, to continue, we've been much faster than that. It's not efficient. Obviously, we think it's worth kind of making those investments to grow. And we've been getting rewarded for that, right? You've seen good logo, new logo numbers. You saw kind of enterprise combined. All of enterprise combined was like 40% of the business last year. So we've made some good progress there. We obviously broadened out the portfolio. We were very much a data center switch focus. So you have to -- you could only enter into an account when there was a data center opportunity. Then we added routing on one end, and we've added the campus and some of the campus capabilities, layering in some security through some of the M&A stuff that we've done. So that portfolio has become much broader now, too, and we're seeing good acceptance of that. And that gives us other ways to enter into accounts. And I think that's been very successful as well. So you've seen Jayshree kind of have listed some customer wins on the last earnings call. And I think it's helpful just to see the breadth of those, right? Because we were very much started out of focus on financials, and we had to focus on high tech. But now you can see from those examples that it's much broader, and we're able to start kind of sometimes with the campus win, sometimes with the data center win. So I think that's definitely helped. We're still focused on kind of the large enterprise, say, larger mid-market type customers. I think that's where we see the biggest opportunity. That's where we think the market really wants a competitor to be able to kind of provide an alternative in that part of the market because there's a very dominant shareholder in that part of the market. So I think that's an opportunity, right? We have to continue to execute. We continue to layer CloudVision and some of the capabilities of CloudVision. We're continuing to add to that all the time. That is definitely a differentiator for customers. The automation that comes with that, the visibility that comes with that, it's clearly helping us to kind of drive those new logo wins. So over time, we'll do more with the channel. We've started to kind of make some investments in the channel. We are signing up channel partners that are supportive of kind of Arista, the more pull-through we get from end customers, the more broader we're adopting in the enterprise, the more successful have signing up those channel partners. And that -- so that will take some time, but that's kind of work that's in progress now.

It's just new. Okay.

It will help drive some of the campus growth, et cetera, into the future.

Would it be a fair characterization to say that the channel investments and the channel successes are going to be a little bit relatively more important for Arista, given the margin structure, the very successful margin structure that Arista has?

Yes. I mean I think in some ways, from a business model perspective, it's okay either way, right? You've got a sale that's kind of has the opportunity for more software content, has the opportunity for more services content. So it's a good, healthy margin kind of sale, whether we share that with the channel because the channel is selling or we resource sales and marketing that ourselves. I think that's either way from a business model perspective, we're okay with that. So I think there is an opportunity to drive sales through the channel. There's an opportunity to share margin there and have a sales-led kind of channel sale, but the direct sale into those accounts as well as also fits very nicely in the business model.

Is there a difference between -- from the enterprises, especially campus, you growing 100% the last year, there's a guide out there for 100% in this current year. Is there a difference between coming at it from, like, say, the wireless access point or from more towards the enterprise switch and more the data center that I would think you would be more inclined to be coming from? A technical question maybe.

I mean it's interesting. We had gone into this thinking that the majority of those new campus wins would come from existing data center customers and that would end up pushing from the data center into kind of the campus aggregation there and then to the wireless. But if you look at kind of the numbers, I think Jayshree has commented on this a couple of times, we're seeing a good percentage of those wins in campus, our actual campus with, right? Now whether they're starting with wireless and going back or wire -- starting with the campus wired products and going to the WiFi, it's hard to tell sometimes. But they are campus like we're starting with a campus win in a customer account. So that's exciting and probably different to what we thought. And I think it just speaks to the fact that people are looking for an alternative. And we're filling that position.

And just to reiterate, there's no difference from what you're seeing in terms of the go-to-market competition. I would say like another somebody that's more wireless focused versus more of a switch focus it doesn't matter from what you...

Yes. I think our bias is still towards that higher. Larger enterprise customer where -- it really is concentrated. If you look at the market share numbers, it's really concentrated to a single player. So that's a good focus for us. I mean there's obviously other players in the market kind of but not necessarily in that large enterprise space. And I think we have the brand to be able to kind of compete for that business. So that's a natural place for us to play.

Yes. Maybe of enterprise back to cloud. It's more of a same question. I mean I think you've been dealing with this phenomenally well the last how many years, you've been a rise. But this business is getting very large. The spending is very large. You're very strategically important to your customers. How do you -- everybody asks you a million times and they will all day and your one-on-ones about the cloud cyclicality. How do you kind of look out and manage that?

Yes. I mean, look, I think the cloud business is hugely important to the industry and to Arista, right? This is a piece of the business that will continue to grow over time. And if you look longer term, I think you're -- there are so many continued drivers for their businesses or AI, some of the things we've been talking about with AI, et cetera, that it's a very important part of the business and very special customers, as Anshul kind of talked about on the call, right? Is there a cyclicality, will there be cyclicality there spend? I think we've seen that once, right, to date. I think we've kind of accepted that there may be some cyclicality to their spend over time. I think sitting where we are now, supply is kind of giving us more visibility, the supply constraints more visibility to kind of what's happening there. But eventually, there may be some cyclicality in that spend, right? I think that's why it's so important to have the other pieces of the business that we're investing in those as well and continuing to have those grow as well and provide some balance, right? I mean the model is pretty flexible. You saw us kind of -- even with the COVID in 2022, you saw us manage kind of through that very well from a financial metrics perspective. So there is a lot of flex in the model. But again, I think we're in a cycle right now where we've got better visibility and we'll see how kind of this plays out into the next technology cycle.

It's sad that these questions get tougher as you have success.