Well, thank you so much for that introduction, and welcome, team. We've got an action-packed half-year results for you. We've done a lot, so we're going to charge through it. This is the FY 26 half-year results for Megaport. Now, we're just going to open. We have not run this before, but we've got a bit of a slide key here. Obviously, in the half, we made two acquisitions, and as we present through the slide deck, what you'll see is each one of these will represent. At the top right, you'll look at the top right of the screen, and you'll see if that's Megaport Network only, Latitude only, obviously Megaport Plus, et cetera, and this is the total group. So just remind all folks, look at the top right, think of this, and hopefully it's pretty self-explanatory.
So we'll be going through company highlights, financial results. Tish is sitting next to me. We've got an acquisition and strategic update. We've combined the two of those, and we're actually going to spend a bit more time than what you normally would expect for a half year, given the two acquisitions that we've made. And we're also going to walk through how that aligns to the strategic view of the business moving forward and why we made those acquisitions. We're also going to spend a bit of time on guidance. Guidance has been a little bit trickier. Obviously, we've had movements in FX, but we've also got two other companies coming into the business.
So we're going to give you a comparison of what we had with assuming there was no acquisition, and then we're going to break it down and finish the year with the acquisitions included for full transparency. In terms of, if we look at the top right, what you'll see, just, I won't keep droning on about this, but in the top right, you'll see that key. So this represents Megaport, Latitude, and Extreme, the two acquisitions inside the business. This is the Megaport Group annual recurring revenue. So we are a $338 million business, $263 million from Megaport, $68 million from Latitude, and roughly AUD 7 million in ARR inside the business. We've broken two highlight pages, one being the Megaport group highlights, and the next will be the Megaport standalone.
So we're giving you full transparency of the underlying business, and we have had an incredible run in that first half. Let's start with the group business. We said before, AUD 338 million in annual recurring revenue. That's up 49%, AUD 112 million year-on-year, and we closed, as you know, 2 acquisitions. We acquired, we announced and closed in the half, as we came to market in November, did the raise, and actually announced both acquisitions. Latitude.sh, which is the $68 million of ARR, 22 locations, GPU, CPU-as-a-service business, and actually adding more and more innovation that I'm going to show you through a demo soon. We actually announced the acquisition of a company in India that we didn't share the name, which was 40 data centers in India. That was for regulatory purposes.
We acquired Extreme IX, which is the largest internet exchange platform in India, something that we do in many, many countries. That gave us 40 different data centers with network, and we're going to talk through that further on. Comes with AUD 7 million in ARR and 400 customers. If we look at guidance, so let's. There's a lot of-- we've spent a lot on two guidance slides, and we've also got detailed guidance in the appendix, specifically covering off FX and also the breakdown of each of the companies in there and how it's made up, so that I think all analysts and investors can get their head around it. So Megaport Network original revenue guidance has been revised upwards.
The guidance that we gave at the start of the year, assuming no acquisition, we have revised the bottom end and tightened that range based on the success that we've seen through the business, and we're going to walk through that. The Latitude.sh revenue guidance that we gave in November, we are reaffirming. Let's look at the highlights from a Megaport Network perspective. This is, as you see, top right, the Megaport logo sitting there, AUD 263.4 million in ARR. That's up 19% on a constant currency basis, AUD 36.8 million year-on-year. There has been an FX headwind as the move between 65-70 cents, which is why it showcases at 16%.
The importance is that the underlying business is growing at 19% in constant currency, an incredible result from the team. And how about this? Net revenue retention by logo, 111%, up three percentage points year-on-year. Our net retention continues to grow. I'm going to talk about why we've seen that growth inside the business as we get through the deck. And also, let's talk about this customer lifetime. As we've changed the product mix and the types of services we offer customers, what we're seeing is customers taking longer-term commitments, and instead of just buying cloud connectivity, long-haul data center to connect, complicated global WAN structures, and you name it, and they're actually bringing all these new products, including internet, together.
Our lifetime for the customer, this is a big deal, and this is, for those who are unaware, it's basically one divided by churn, has increased by 3 years. In effect, the churn is lower, and so our customer lifetime has increased. It's 13 years, which is outstanding. Now, when you flow that number forward and you look at your lifetime value, your total lifetime value, which is a mixture of the lifetime and the average spend per customer and your margin wrapped into that, we're up 57%. That's a $2.5 billion total lifetime value, and we've got a slide to sort of share that, and we'll talk more about it. You can see our annual recurring revenue breakdown. Again, this is Megaport's standalone business or the network business, excluding acquisitions.
You can see we continue to grow that, annual recurring revenue fast. You can see the breakdown of Americas, Asia Pacific, and EMEA as always. Let me state this: the Americas, and specifically the United States, really is on fire. They're growing at 24% inside that business, and that's where we've been investing for a lot of growth. We're seeing an incredible, opportunity for us there, and it's continuing to grow, and we're not even scratching the surface. If you look here on the right, what we've shared is the net or the incremental ARR additions, and you can see this is the largest ever year-on-year ARR increase. This is astounding. You can see this just pushing up and up and up and to the right.
So we are incredibly happy with how the team's performed, and we're going to go through more of those details. It's not just about expanding the existing base, and that it's super critical that you do that, but if you're not adding net new logos, you've fallen out of product market fit. And this is where we've seen that in H-- sort of the full year, you saw that massive jump in new logos. We have actually had a 100% increase in new logo growth compared to the prior comparative period. That's important because seasonality rolls throughout the year, and this is a representation of the fact that this wasn't some lucky sort of H2. We're actually growing and continuing, and year-on-year, we're up 100%.
That is an incredible result from the go-to-market team, but it's a mixture of two things: products that actually resonate with the market and the go-to-market team that we've invested to take that out to market, and that crosses all areas: channel, frontline sales, SDRs, customer success, you name it. So congratulations to the Megaport team there. This is a slide we shared in the full-year deck, and it's actually just continues to astound. It's actually, I think, what we've been hoping to see, but it's, it highlights the fact that the investment in engineering talent and all of those products that you've heard us sort of constantly refer to bringing out, in that period of time, sort of FY 2024 period, all of that has come to fruition, and 30% of our ARR growth is now attributed to what is new products.
The underlying business is still growing fast, but on top of that, you get this growth from new products, which just highlights the importance of continuing to invest in new products to add to the space. I've got a slide in the strategy section that will sort of drum this one home, but you can see what a dramatic increase that has had in H1. Outstanding result from the engineering and product team, and great job again from the sales team for taking that out. Total lifetime value, we shared that prior, and we highlighted the fact that customer lifetime's gone from 10 to 13 years. You can see that that is a marked shift upward.
That is, if you look at it, we made quite significant changes in the business in terms of how we offer contracted services, the types of services we take to market. We did some significant investments in the network and, and sort of... I'll show you on the next slide. But we, we've sort of been going on about the investments that we've made in 400-gig backbones, 400-gig networks, so we can offer a 100-gig just about everywhere on the planet, as well as rolling out massive, massive speed in internet. All of these things bring through significantly longer contracted business and also much higher ARR lands. And so that's why you've seen our ARR per customer, of the 3,000 customers inside the Megaport business, this is excluding the acquisitions, is up 6% year-on-year.
You couple that with that lifetime increasing, and then you actually have this $2.5 billion increase. So, again, testament to the health of the business in every metric you can think of, going incredibly well. This is a bit of an eye chart. It's always astounding to sort of think that this isn't a full year. This is only six months of execution, but worth pointing out that we continue to execute against the strategy, which we'll talk through further, but we break that down into build, innovate, and invest. These provide the pillars to increase TAM and go to market. If you look at build, 51 data centers added. We're now. We crossed in this half, actually, we forget because we celebrated in August, but we crossed 1,000 data centers globally.
We're now at 1,034. We added five IX locations. We've actually on-ramped, in effect, 11 direct connects into Latitude, so the compute platform that we now have. We've got 11 new cloud on-ramps to 344. It really is cooking with gas there. We added two internet markets, Italy and Sweden. That's all the regulatory components coming off. We've got pretty much the majority now. We lifted internet to 100 gig in 16 metros. In fact, when we first rolled out internet, we weren't sure how successful it would be. It's been so successful, and we thought it would only ever be at sort of 10 gig levels.
But for the last year or so, we've been rolling out big 400-gig machines, and so you can get a 100 gig in 16 metros, and it's, it's astounding to see how many enterprise customers are taking that up. So we're gonna not-- we're not stopping. We're continuing to, to build that out globally. A 100-gig connectivity. So this is... When I first joined Megaport, the fastest we could deliver was 10 gig anywhere. And then we had-- now, if you look at it, from 802 data centers, we can deliver a 100 gig, in 60 seconds. Quite astounding. There really is nothing else like it. Let's talk about innovate. Cam and the team have been working incredibly hard to deliver some pretty impressive, enterprise-grade security features.
We rolled out IPsec, which is encrypted tunnels on the MCR, packet filtering, which is like control lists on that platform. We're going to continue to expand our investments into security, and you'll see that in the second half. We added console access for MVE. We added, we added probably one of the most requested MVE images, which was Meraki. We've got Juniper and Aviatrix. 26 countries, seven languages. We added 400 gig ports, which is really interesting. If you remember, when we launched, we were always 10 gig and 1 gig. We added 400 gig ports and actually had immediate customer adoption. So the amount of, utilization out there is quite astounding, and, and we'll talk through that even when we look at sort of what AI is providing from a tailwind, but just quite astounding.
Now, what's important is that we continue to build to support that. Acquired two companies, we're going to go through that. We continue to invest in go-to-market. We're actually doing some pretty cool stuff with DWDM, which is like if you get your propeller and want to hear that spinning, we're looking at rolling out DWDM for the first metro and then doing that across many. Cam tells me this is a really big deal. And then we've upgraded the backbone capacity to 400 gig to 8 countries. Now, what does that mean? We're getting subsea connectivity at 100 gig, and this is where we get massive ARR for single VXCs, and it's really, really, it's exciting.
And the last thing, and I'll point this out just more broadly, is, you know, people, I think, carry on that AI is changing all their business and, and it's hard to know what, what I think, some businesses is real and what is not. In our case, we're not someone that sort of just spruiks this out there. We are seeing tremendous assistance, like most companies, for AI in the development space. So when you look at adding Claude into this, the efficiency of the development teams is phenomenal. But also what we're seeing, and this is more for everyone's kids out there, if you want to study something, I'd call it prompt engineering, AI prompt engineering. Folks that don't even know what, really what they're doing, but they know how to prompt the AI.
We're seeing these folks come in and just do some incredible speed and change so many things inside the business very, very rapidly, particularly on the engineering front. All right, I'm going to pass over to Tish, our fearless CFO. She's going to slide in here, and we'll go through the financial results.
Thank you, Michael. So financial results for the half. So, a fair bit happening. We had two acquisitions, as well as the underlying business of Megaport. So what we've highlighted here within the EBITDA is largely Megaport and the Latitude acquisition. Now, Latitude, we acquired, we announced, did a capital raise early November. On the 26th of November, we closed it. So you will see one month of results in here as well. So I did want to highlight that. Revenue, I know Michael has talked about that. However, it does come through in the numbers clearly. You've got the strong expansion with NRR, with the continuing investment within the existing customer base. We've got the growth in new logo acquisitions, and those two combined results in strong revenue, plus the inclusion of the compute revenue from Latitude.
Partner commissions continues to stay relatively consistent at 11%. Direct network costs, now we've got IFRS 16 in there, which I know some of those on the call do enjoy talking to me about, so I'm always happy to talk about accounting standards with anyone. However, what we look at is across kind of a net-net basis of that, regardless of accounting treatment, because the focus of the business over the last 18 months has been that global backbone rollout to 400 gig, which has been led by North America in particular. So if you exclude IFRS 16 from both this half and last half, it's consistent. Employee costs, we've been, and I think we've hired quite a few folks, and so you will see in here that is continuing to be planned.
You can see that rolling through in the numbers, and that is to support the accelerated revenue growth and then the related spend associated with those, those new folks coming on board. In terms of other operating expenses, you've got your sales and marketing event activities, which you can clearly see in there, as well as travel, and some IT costs associated. Now, EBITDA is one thing to notice. Great, Tish, you've got an EBITDA margin of 26%. That does include the Latitude one month. And I just wanted to highlight that the exit margin for Megaport network business of 21% is in line with guidance and really reflects that timing of our planned investment, which we'll talk about further in the guidance slide, and happy to take any questions. However, I did want to highlight that as well.
Now, the cash flow, a lot of things happening in here, so I've tried to break it down clearly within the text, but I'll just go through it quickly. The operating cash inflows have increased. That is purely due to the higher customer revenues during the period and the one month of the Compute or Latitude.sh results. The investing activities, you've got some acquisition payments, and you've got also the CapEx payments, which are related to the planned delivery of equipment. Within the appendix, we've provided the breakdown for CapEx for the group, so you can clearly see what that has been on, and you will see that it's to do with supporting the expansion plans, and including the planned delivery.
Financing activity inflows, you can clearly see that that's largely driven by the capital raise that we provided earlier in the year, late, late last year. One thing to highlight is the net cash flow was an outflow, if you ignore all of the capital raising and acquisition activities, was an outflow of AUD 10 million or under AUD 10 million. Now, that reflects the planned expansion, hiring of the go-to-market, as well as the CapEx spend and the investment in the front-ending of the ordering of equipment. So that is one thing I do want to highlight to the group. We've talked around headcount previously, particularly at last, at the last year-end time that we came to talk to you. We've got here the sales and marketing continues to be a clear activity that we invest in, at, and that's moving upwards, continues to.
Product and engineering is a big focus area, particularly with Cam hiring across the world, hiring the right talent. G&A continues to stay steady as a percentage of revenue. This is how we look at the investment of the business for headcount, and this is just Megaport standalone at this point. I'm going to hand back to Michael. I've done a quick snapshot of financials, but over to you.
Beautiful. Great job. Thank you, Tish. All right, so we're going to go through the acquisition and strategic update. So this is where we're going to look at some of the strategy around the business, and also talk through a bit more detail on the acquisition. We're gonna run a bit longer than normal, just because there's a lot to go through. The first slide is something that I, I probably didn't expect to be adding to the deck, but this is a lot of inbound questions. These two questions have been coming, certainly with the way markets have moved, and sort of everyone's trying to wrap their head around, I think, the, are you a company that benefits from AI, and are you a company that can be disrupted, rapidly by AI?
I just wanted to sort of draw this for all shareholders and anyone that's either new to Megaport or even knows us well. First question, I'm gonna sort of just, you know, point this out: Is AI benefiting Megaport? The answer is yes, and that is why you've seen such outstanding results in that first half, particularly driven in the United States. AI is very, very strong there, and the rest of the world is sort of starting to catch up, but really, in the US, it's cooking with gas. It is providing what is a very strong tailwind for us. So AI is driving significant movement of data.
So if you think about it, there's huge amounts of data movements as you start to either send your data to AI, to models, whatever it may be, and the other piece is it requires large amounts of processing by compute and GPU. So the two parts of our business, both network and the compute and GPU businesses, benefit strongly from AI. I think that's obvious, but it's just worthwhile sharing it. The other piece I want to share is that Megaport is a software-defined physical network and compute. We use software to automate physical things that live inside data centers all around the planet. The last one on this is you don't have to pick a winner. Megaport is never looking to pick the winner, and there's lots and lots of changes that we see all the time in terms of deals.
We are the picks and shovels for AI. As often people would say, you're like the overnight success, but 13 years in the making, and so we've been building this platform for the past 13 years to land literally in the exact right place with the right platform at the right time. There's almost no one else on the planet that can do what we do, and certainly no one that offers what we do on a global scale. So we are absolutely a beneficiary of that space. We built the cloud component, and the AI piece is actually playing out in a very similar way. So the second question, which is, well, whoa, hang on, can AI disrupt Megaport like these software-only providers?
Now, for a few years now, I've always said it'd be always tough being a software-only CEO with all the disruption happening in AI, and that's why I'm a huge believer in Megaport's business because it is the beautiful mix of both software and hardware distributed at scale. So the answer is no, we won't be disrupted by AI because of one important fact, and that is we are physical. And it's a bit of a, sort of, a lot of physical words here, but just to sort of drum that message home, we move data via networks, and we process data via CPU and GPU. Both of those are physical. And to put that into perspective, we have a software layer that runs across this physical IT infrastructure, which can't be replaced by AI. AI is not so good with atoms, not yet anyway.
If you look at it, we've got 320,000+ routes of physical fiber, 320,000 physical routes. We have 3,000+ physical network devices. They live in 1,000+ different data centers. We now have 7,700+, and growing fast, physical compute servers and GPUs, and we deploy them in 30 physical countries, not just deployed via some SaaS platform that lives on AWS, but actually in the country, and then on top of that, we have all the telco regulations and licenses, which is not easy. All right, I think I've driven that message home. I've shared this slide a lot. I won't labor it too much, but we, I wanna keep this, keep reminding folks, we make all strategic initiatives based upon these guiding principles. That includes innovation in product and acquisitions.
We obviously made 2 acquisitions, but I'll just remind folks, this is the secret sauce that makes Megaport strong, and so, same, so too with Latitude and, and any other business or product that we want to look at. The answer is, let's start with automation. Automation is the key, and it's not easy to do, and that's the software automating physical infrastructure. You can automate it, you can make it instantaneous. If you can put it at a global scale, you can actually deliver a global service to customers. If you make it the most resilient, and then on top of that, make it flexible, you can buy it monthly, hourly, whatever it may be, 60 months, you name it.
If you then can make it self-service, but with a really cool GUI, and make it really easy, and then you provide the best support, the pricing is disruptive, and somehow you do all of that, and you make it profitable, you have got an incredible company, and that is what Megaport is, and it's also exactly what Latitude is from an acquisition standpoint. The strategy from Megaport's perspective remains the same, and we constantly share these, I guess, the rings of total addressable market. So for those who've not seen this slide before, on the left, if you look at these rings, each ring represents a product set and a total addressable market that we can go after, and we started with cloud connectivity, we added virtual edge, then we added global WAN, data center connect, DC Internet, NAT gateway security.
That slide I showed you before about the growth in new products, these are the new products that are driving that. We just added two new rings. This unlocks significant TAM for all compute as a service and GPU as a service, and we are gonna continue to add rings in the second half as we announce new products and new innovations that we're gonna bring. On the right, if you look at the pillars for growth, we talk about build, innovate, and invest. Building is expanding a ring. It doesn't necessarily mean that we're adding a ring. So a great example is new data centers. Every time we add a new data center, we increase the TAM of all of those products. Every time we add a new market, like India, it's the same component, or when we added Brazil.
And then we've got expand capacity by going, as I said before, from 10 to a hundred gig to 400 gig, and now we offer those services. That expands the TAM inside those rings, and then we're gonna continue to expand compute and GPU offerings, so different types of SKUs, and obviously significantly more, and distribute it in all those different locations. Then we look at innovate. We are going to constantly build, and you saw the hiring that Tish shared around the innovation team inside both network, and I've shared also security been a big play for us. Those two sort of become hand in hand. And then if you look at the CPU and GPU innovation, in the demo coming up, I'm going to show you some of the really cool stuff that's already coming out, particularly in the AI innovation space.
Then we look at investments. We are expanding product engineering, as I mentioned. We're investing in that space. We will continue to expand go-to-market, particularly as we service and take these products to market, which is why you're seeing so much great results in that side of the business. We'll always explore strategic acquisitions, but always those strategic acquisitions will line back up to what we shared on the previous slide, which is the strategy that we look at around the investments. So we're going to quickly go into the Latitude.sh acquisition. We shared this in November, but it was a quick session when we did the capital raise, very successful capital raise, and thank you for all your contributions. I think we had 30 minutes from memory to sort of bounce through this.
So Latitude.sh, it's a compute-as-a-service business. It provides high-performance CPU and GPU in key markets worldwide. It's very simple to use. It's very API-driven, just like Megaport, incredibly predictable billing, unlike what you would have in the cloud, and flexibility to deploy workloads literally on demand. Super important. Massive global scale, totally automated, rapidly growing, and then have actually been very strong from a product-led perspective. All of this lands inside those, that, that, the strategy that we had from a business perspective. If you look at the ARR, well, what does it look like? Where were they based, and how does that play out? Well, there's 10 countries and 20 locations when we acquired. I think they're at 22 today, will continue to grow.
The US represents the vast majority at 50%, Europe at 20%, Asia Pac at 17%, LatAm at 13%, and it's at $45 million ARR as at 31 December 2025. Okay, who does it service, and what do these customer use cases look like? There's 2 parts to the business today, and actually, I'm going to show you how that will change in the future as well. But there's compute, and then there's GPU. If you think of compute, what you need for a bare metal platform, such as what Latitude's produced, high-performance compute is not something you get delivered inside a cloud, and it's hard to run and build yourself. So in the middle, we have these incredibly high-performance compute and very large network stacks, which deliver blockchain Web3, as an example.
These are financial service rails, as an example, using blockchain to transmit via Solana, as an example, say, foreign exchange for enterprises. They use up huge amounts of compute and network, and it needs to be distributed globally. Enterprise applications that are not optimized for cloud cost a fortune in the cloud or don't perform how you need them to perform. What you have is an ability to run incredibly high-performative applications that don't sting you with all of the API calls and so forth in the cloud. So then you've got SaaS applications that need high performance at the edge. If anyone has a kid who loves Fortnite or gaming, you'll know what latency is and the importance of having edge compute that process fast with very low latency for gaming.
We've got some incredibly interesting gaming companies that actually spin up actual compute platforms for the gamer themselves. Really interesting. And ad tech is. If you thought gaming required speed, ad tech's even faster. It needs to show the ad as fast as possible. It needs to know Michael Reid is looking at this website. He's interested in a new surfboard or whatever it may be. I'm not that good at surfing, to be clear, but let's just say a really long surfboard, 'cause he's not that good at surfing, and they need to show that as quickly as possible. That's ad tech. And then we go into the streaming element as well. So that is some of the use cases that are really, really big in the CPU as-a-service.
GPU is a big market, and so if we look at inference being how we access the AI, and that is like ChatGPT, when you access it, that is inference. When you train ChatGPT, that is the giant data centers that you're seeing build out, so sort of the hyperscale, and then you've got this fine-tuning element, where you take a model and you fine-tune it for enterprise. I'm going to show you in the demo something really interesting. Let's jump into that, 'cause it's like, well, Latitude's had this history of product-led growth, and I shared that, not only is the platform incredible, but you've got to be able to access it and make it simple. So we're going to pray to the demo gods and so forth and just see if I can switch this machine over, and I'll see if I can.
All right. Steve's telling me to keep the demo short 'cause he just, he knows that I love it. Can this be seen? All right, first things, you're welcome to the Megaport portal. You're used to this. You can see all these different locations in the U.S. Every one represents a data center that we're in. This is cool. So we've now landed in India with that acquisition, so 40 different data centers live, and you can start ordering there. Now, if you jump out of that piece and you look at the top left of our portal, we can click on this particular piece, and now you can see the option between network and then compute, so Latitude.sh, we can click that component, and now we're in the Latitude.sh platform. And here is one I have prepared earlier. We look at projects.
This is the portal that we look at. Here's a live demo to shareholders, so we'll go and click onto that project. And just like Megaport, it says, "Quick click to create a server," and we've got some really interesting things down here. So we can create a server. Now, when we acquired the business, there was bare metal, bare metal GPU, and GPU VMs, but very recently, and I don't know if you've noticed, but we now have CPU VMs. This is virtual machines running across the infrastructure, a super important innovation inside the product. So you can actually click on the CPU VMs and bring down... We only have a few locations today, and we'll roll these out to more.
But you can choose a very tiny virtual machine at $0.07 per hour, which spins up a virtual machine with Ubuntu, run it for hourly at $0.07, and you can click Deploy. That is as hard as it is to run a virtual machine. It's sticking to schedule there. I'll move back to the bare metal component and make it quick 'cause Steve's on me. And this is my fun bit. So we'll go to the bare metal component. If we look at North America, South America, Europe, and Asia Pac, these are all the different locations. We'll choose Ashburn, where the clouds live. We have physically installed compute infrastructure that sits in all of these locations. They are available to deploy. There are different levels. There are entry-level, core optimized, memory optimized, storage optimized, you name it. Let's choose a big guy.
Scroll down. What's amazing is the platform is constantly looking at what infrastructure is there and what is the likelihood of using a certain operating system. Remember, this is bare metal, so it is actually your entire server, and it will preload an operating system into the ones that are available or some, and then you can choose to deploy that. So we'll choose $3.52 per hour, no RAID, pick a name, blah, blah, blah. You click Deploy, and it will roll up on the top right here, and you can start to see that being deployed. So that one's getting deployed. Now, if we go back down here... Oh, you can see that's now on. So that's your 5 seconds of deployment. Last thing I want to show you. So what we've done is we've deployed a bare metal machine, physical.
We've deployed a virtual machine, which is now running over there, and we're going to go to this thing, which is so new that no one even knows it exists, and this is AI inference. But I did want to show the team what this thing looks like. So this is where we've actually deployed model as a service, and so if we click through to the different playgrounds and API, create an API key in here. So what we want to do is create an API key. We then grab a model, an AI model, our own open-source model, and we deploy it on our own physical infrastructure, so you've got total control over that. So you can go and create a key. We'll call it River Test. Grab that key. Which one was that? Okay, now I've got to copy this.
Doesn't matter if you copy it from a security perspective, I'll be deleting it after for those worried. We go into the playground, and now what you can see is at the top left here, you choose the AI model you would like to run, your own private model, that you control the data that's going into it and the compute that it runs on, so it cannot be taken. So let's choose Llama, for those of you familiar with that. We paste the API key in the top right, and you are all familiar with this. This is where we now write our question. "Write a launch email for our new inference endpoint." You click Enter, and it will start generating. So what it's doing is it's asking this question to a Llama agent, so think of an end. Look, it's already done. It's amazing.
Think of an enterprise that wants to control the use of all your data, and you want to upload that into a private instance that's sovereign, that you have total control over, and actually, no one can steal that data. All right. Steve's telling me that's enough. There you go. I think that's really cool. I just want to show you one last piece. Ignore that. I'm going to delete these servers because this is the important part. We're deleting the server. We copy and paste this back in here. And what has it done? It goes through a process of wiping that server and making it available back in the pool for someone else to use. 100% autonomous, totally delivered via code and delivered via a portal like that. Very cool. All right. You can see we're excited about it.
Now I've got to get back to the presentation. And we're back. All right. So people are saying, "Why is- Why did this make sense?" This is the next logical step for Megaport. We shared this in the presentation. The reality is... Oops, let me fix my laser pointer. We were the kings of network and are the kings of network. What if you look at IT infrastructure, they're made up of three pieces of the puzzle. In fact, every application you have ever used, excluding ChatGPT or AI, lives on network, storage, and compute. This, the compute element comes with Latitude.sh, and you can see that we're missing a piece of this puzzle, and you can probably imagine that that's something we're working hard to go and release as well to finish the trifecta of both IT of all the IT infrastructure.
I won't go through the pieces on the right. We shared the 100-day plan. We're a few months into the acquisition as we are today, progressing as expected, and we're really, really thrilled to have the team come on. It's actually probably progressing better, given what I've just shown you from an innovation standpoint. Two major innovation releases between when we acquired and when we landed. So big shout-out to Guy and Eduardo and the entire Latitude team for bringing that. I won't go through this. We'll talk through it if you need on the call. The last piece was we made an acquisition called Extreme into India, which is the Extreme acquisition. Why? Well, it's the fifth largest economy. It's the most populous country in the world.
The real reason is we had massive demand from all of our global enterprise customers, 3,000 different customers, asking for us to get into India. It's not easy to land. Regulatory purposes and actually run a network is difficult, which is why we've acquired to land. It comes with AUD 7 million in ARR. It's accretive to the business, 400 active customers, 40 data centers. Now, we're rolling out infrastructure to retrofit all those sites with Megaport-grade infrastructure to... And then we will offer all those services as soon as possible. We've shared this slide. The strategy has not changed. I won't go into too much detail, but you can see where we've landed. We're in the transform and reset phase. We've been rebuilding go-to-market. At the full year, we, we announced that. We've done incredibly well against that.
We're landing within what we've told the market with. From where Tish shared. You've seen the product. The net revenue retention hasn't stabilized, it's actually increased, so we're doing well there. Revenue's up. We'll land at the end of FY 2026 and then into this accelerate revenue phase, where we'll capitalize on that prior investment, continue to expand the TAM. We're going to grow the market share. We're going to continue to invest in revenue, with revenue growing faster than costs. That remains inside the business, and we're going to accelerate revenue through investment, constantly. And then, as we get to the future, it's going to be a fairly large. I mean, we'll be at significant scale at FY 2030 and beyond.
We'll be a global leader in infrastructure as a service powered by software with 20% sustainable growth, highly profitable, converting scale into sustained profitability and free cash flow. All right, guidance. Let's do it as quickly as we possibly can, even though it's a little bit complicated. What we've given you is two slides. This is the first one. As you see at the top right, Megaport network only. This is the Megaport network updated guidance versus the original. Remember, there's an FX component, and we've also had two acquisitions. We've excluded the two acquisitions, and we've kept constant currency, so you can see how the underlying business is performing. As we predicted, it's at AUD 0.65. We originally had revenue at AUD 260 million-AUD 270 million.
We've raised the lower end of revenue by AUD 4 million, based upon the success we're seeing inside the business. We've held EBITDA margin unchanged, we've held CapEx unchanged. Why would we up the bottom end? Well, we've seen outstanding performance in ARR, up 24% in North America. We're seeing net retention up 3 percentage points, and we've got sustained growth when we look at new logos. Pretty obvious, but that's why we've tightened the bottom end of the range, and we're confident there. As we said, there's two material changes, being the AUD to USD and the two acquisitions. So then you're thinking: "Well, Michael, what does that mean for us?" Great news, we've answered that test too. We've gone, all right, let's give the combined group updated FY 2026 guidance.
We've got it at 0.70 AUD to USD. We're not saying that that's what it will be for the entire half. We don't know what it will be, but we're certainly telling you what it looks like as we see it right now. And we've also given sensitivities on FX. We've also given great detail, if you want a breakdown, because we've added two companies and three companies coming together in the appendix. So 302 million to 317 million in revenue, 21%-24% of EBITDA of revenue, AUD 90 million-AUD 100 million of CapEx. That includes the tightening of the range of the bottom end of Megaport. Unchanged for Compute to Latitude from what we shared in November, so no surprises there.
We've added in about AUD 3 million-AUD 4 million for the internet exchange business we acquired in India. We've also included the CapEx to go and roll out that hardware. Worth highlighting that the maintenance CapEx on the combined group is actually less than 2%. If you look at it, pretty much all the CapEx we're deploying into these is growth CapEx. And lastly, just to point this out, the sensitivities, just to give you perspective, there's detail further on, is we're at AUD 0.70. If you look at a AUD 0.05 movement, that is an AUD 9 million. So when the US dollar weakens, that is an AUD 9 million reduction in revenue for the business. All right, we made it. We're up to questions. Let's do it.
We'll now begin the Q&A session. As a reminder, if you'd like to ask a question, please select the Q&A button located at the bottom of your Zoom screen. Alternatively, only research analysts can use the Raise Hand feature to ask verbal questions. We ask that you keep your question to one at a time and be brief. You are welcome to rejoin the queue after your question. Given the time constraints today, unanswered questions will be responded to post-event. Our first question comes from Nick Harris. Nick, you may unmute.
Okay, hopefully you've got me now.
Got you.
Awesome. Thanks for taking my call. Congrats, great to see the core business, in particular, flying. I'm pretty excited to see that NRR continuing to lift. Just trying to unpack that, could you help us maybe understand, what's happening behind the hood or under the hood there? Specifically, you know, are you lapping a really poor, quarter 12 months ago, so it's just the mathematical average getting better? Or is the front of the, you know, like the Q2 FY 2026 pulling the average up, as in, is the core business-
Average what?
In the... It's a trailing twelve-month NRR, right? So-
Oh, NRR. Sorry, I missed that.
You got the key word, NRR.
You're saying... Yeah, yeah, no, I was thinking which metric? It's good. Maybe you said it, and I missed it. Okay, so NRR, so back, back, I think we've shared it a few times, but the, the net retention inside the business is lifting for a number of reasons. One, the U.S. really is on fire, so we've seen massive expansion inside to the, the United States, in particular, customers. Globally, we're strong, but the U.S. really is quite impressive. I'd say again, back to that AI tailwind, we're selling services that are very different. So if you remember the, the journey we've been on, you know, we used to be this, tell this story of, like, we're just a cloud connectivity company, you know, and if you think of selling a connection, to a cloud, that could be a $100 connection.
We've seen million-dollar single VXC connections. Think about that. You can either sell a single connection to a cloud at 1 connection for $100, or you could sell a single connection, subsea long haul, Global WAN across, the subsea, basically long haul at 100 gig, and you're looking at $1 million in NRR. I'm just giving you a perspective that by changing and adding these products, you will get your net retention expanding because you're selling much higher value services into the existing customer base. And the other piece is there's a tailwind around what companies are doing. So everyone is trying to innovate their businesses and move forward, and so people are making moves to spend things, but they're doing it at such massive scale, and that's why we've said, like, data center internet is a really good example.
Such a simple product, but so important and actually helping us drive that. So what I would say, it's two things: It's the fact that we've now got a serious go-to-market team deployed and mature globally. We went to—I mean, we had an off-site in North America for the sales kickoff. There was 150 people that turned up. When I first came, there was 4 people and 1 customer success person. It's a really big change, and so when you've got that customer success team and new products and a market that makes sense, you get massive expansion. Europe's doing well for us as well. Asia Pac, I think, has been a little slower just because I think they're trying to figure out what's going on with AI, but we've still got growth in those areas.
So if you sort of throw that through your lens, then you get the outcome. The U.S., for us, will always be the largest growth and the largest opportunity, to be clear, which is why we're investing strong there.
That's great. Thanks. I've got some more, but I'll jump out of the queue and share.
Thanks.
Thanks, Nick.
Our next question comes from Eric Choi. Eric, you may unmute.
Morning. Thanks very much. Would it be possible, Agree, to ask two number questions that are kind of related to Tish? Sorry, Michael. Just one on FY 2026 and FY 2027. Just on FY 2026, just trying to unpick what's changed in EBITDA guidance. So can I just confirm, if you took your comments in November, and if you added seven months of Latitude, we should have been getting about $73 million of EBITDA, and today you're guiding to about 70, so it's really only a $3 million difference. And then of that, I think you can work out about $2 million is FX, and then there's another $1 million, which is just extra Latitude investment, but Latitude revenues are in line.
Yeah.
That's the first one. And the next one-
Yeah
sure.
Eric, that's a good question. So we highlighted when we acquired Latitude around the 50% EBITDA margin. Now, again, in terms of they didn't have a particularly large go-to-market team, and so part of that is investing across both to make sure that we are able and set up ready to sell the compute. That's really critical, so we've built in that. But, but don't... The target is to, you know, particularly into FY 2027, is to try to get to that back to... Well, we will get back to that, that margin. But in the meantime, we've got to invest to be able to grow. So same similar story to Megaport.
Good stuff, Tish. Just on 2027, thank you, they're really helpful. Just if we extrapolate current trends for the core business, 111 NRR, you're doing about 8% land, and then if you take the earn-outs that you've got for Latitude, I just want to confirm that kind of suggests a AUD 450 revenue number, which would be above consensus. And then, like, in that scenario where you're getting earn-outs, should we assume you keep reinvesting, therefore, we shouldn't assume much margin expansion for each of the divisions into 2027, besides Latitude?
I think just make sure you're not just taking a top end there, kind of to make sure you're taking a midpoint. You know, we want to, we will continue to drive forward to earn those, to ensure that the Latitude founders earn out that, that because we earn that revenue. But just in terms of modeling, I never like to go too, don't get too ahead of yourself just yet.
Okay. Thanks, Tish. Thanks, Michael.
Cool. Thanks, Eric.
Our next question comes from Andrew Gillis. Andrew, you may unmute.
Thanks, guys. Can you hear me?
Yes. Yep.
Thanks, guys. Really solid result. Great job. Just a quick one on Latitude, probably for you, Michael. Like, that transaction's been presented ex synergies, but presumably all of your network customers buy compute, particularly in North America, where you've had a really strong underlying revenue growth or ARR growth number. You know, you've got those existing relationships with the telco service distributors. Like, I appreciate you not disclosing a number, but, you know, are there some low-friction sales opportunities there that could help expand NRR? And then can you maybe touch on sort of Equinix Metal? I've noticed you've got a new section on the Latitude.sh website. There could be a direct opportunity there as well. Thanks.
Yeah, great questions. We didn't model the business with synergies. I spent six years at Cisco acquiring a bunch of companies, and synergies is a dangerous thing to just add into this. You know, it's sort of like you solve the equation with miraculous synergies. The way the acquisition with Latitude went down was actually, they weren't looking for an acquisition, they were just looking for some investment into the company so that they could actually get some CapEx expenditure. And so they had a business plan that was related to them not being acquired and just running as a standalone company.
And that business plan is what we've basically acquired them against, without any synergies built in, and giving them a stretch ability if they go and succeed beyond a certain point, to earn additional components beyond what their original planning was, and then allowing them to get to that component. That is excluding synergies. So the point was, well, you, you need to build that business as it is today and continue to expand. The reality is, it's up to us to go and take those synergies, ideally, and then leverage that across the business. So it should, in theory, be cream on top, if you think of it from that perspective. But what's important to call out is it takes time. So I'll give you an example, and we've been on this journey. We always say it's an 18 months.
You start hiring, enterprise sellers today, you start building the platform to deliver more enterprise-style services, and then what you start selling that, it's a 6-8-month ramp, I don't know, could be a 9-month ramp in terms of compute. We'll be finding that out. And then you end up revenuing it. So if you think about it, it's revenue that's in their targets, not ARR, and so it can be a delayed approach. That said, we've already hired, I think it's 5 or 6 frontline sellers. And if you sort of follow us on LinkedIn, you'll see that these are high-end, Equinix bare metal sellers in North America and in Asia Pacific, including solution architects, customer success managers, and frontline sellers. We landed those folks in December, which is astounding.
So the teams worked incredibly fast. When we announced the platform, the acquisition, we actually had a huge amount of inbound. It turns out that bare metal and compute sellers are passionate about this space. A lot of them reached out and said, "Well, actually," to the Equinix bare metal turn, when they've turned that off, they love the platform and could see how valuable it was, and could see that Megaport was serious about it. And so we've actually got these folks coming in. So, that said, they become an overlay sales force to the existing sellers at Megaport, and your point is totally true. The conversation is that Megaport connects folks that are connecting compute between data centers and the cloud. That's what we do today. So the next obvious question is: "Hey, did you know we had a compute platform?
Would you like to explore opportunities to either save money or get better performance?" Or whatever it is. That becomes a sales motion, you push that through the machine. So that is the synergies that we would expect, but hard to model that in the short term. You need to—we need to build that out, and we're already building. I mean, we move fast, if you haven't figured that out. So we're... I'm bullish on that space.
Mm. Perfect. Thanks, guys. I'll jump back on you.
Yeah, the Equinix Metal piece, I think they've sunset that platform, but it's great. It's proving to be a great opportunity for us. So it's a good outcome, particularly with hiring and with our customer opportunities.
Awesome. Thank you.
Our next question comes from Siraj Ahmed.
Hey, Siraj.
Hi, Siraj.
Hey, Mike. Can you hear me okay?
Yes.
Just give me one second. Yeah. All right, just, I mean, maybe actually a follow-up to Andrew's question right now. Just the Latitude momentum, right? The ARR, as of December, looks like they, looks like they only added AUD 2 million in that quarter, so a bit slow, but you are guiding to revenue of AUD 25 million-AUD 30 million for the half. Just, just, yeah, keen to hear, yeah, what gives you confidence, or is it a bit delayed, right? Maybe it's the Equinix piece, et cetera. Would, would love to think of the drivers for the, for Latitude. Thanks.
The most important driver for Latitude is having access to compute that is available for customers to consume. That is the key. Latitude was, if you looked at them, very much a start-up style business, capital constrained, went to market in, I think, want to say March, looking for capital. Obviously, we went through a transaction, and it takes a long time to get that through. So I would say it delayed their ability to procure infrastructure throughout that time, so it wasn't available during that quarter. So if you don't have product availability, you can't sell anything. So that is a prerequisite sort of an obvious but critical statement there. Since that, we've been ordering infrastructure to build out across all the sites, and we'll continue to expand sites.
That was easily to see and to understand why, purely because they hadn't deployed compute in that time. Hopefully, that makes sense.
So that's super helpful, Mike. But I think they got, what, 1,000 servers last month, right? So maybe you're already starting to see that come through. Is that fair? And should we be thinking maybe this half, it's more towards the lower end, and then it ramps into next year? Thanks.
Yes. So, there's a few factors, and we've shared this. There's a ramping period of time from when the servers are received, installed, and then the software layer is added, so they're actually published into the platform. And so just 'cause you've ordered something, doesn't mean it's available for a customer to use, and just because it's available for a customer to use, doesn't mean it's revenue, and you need to obviously ramp that. And so that's why I think we shared there's a ramping profile, from when infrastructure lands to when it gets deployed, you know, when it comes in for a customer to utilize it. And the tricky part is lining up when infrastructure is gonna land is always tricky.
Well, it's become more tricky at the moment just because of supply chain, so it's not clear to get exact dates, so to predict an exact date when you're gonna get something and then predict the exact ramp of utilization is tricky. So what we have is a range around that, depending upon those factors. Some of those factors, you know, are obviously out of our control. The one thing we can control is what we order, when, and order as soon as possible to get that built out, particularly given supply chain. Hopefully, that just gives you a perspective on it, which is why we have a range.
Super helpful. Thanks.
Our next question comes from Tim Lumb. Tim, you may unmute.
Hi, guys. My question's just around the go-to-market hiring process. Mike, can you give us a bit of a sense in terms of how far through it you are? And then once those guys are up and running in a mature level, is there a way for us to think about, you know, average ARR contribution per salesperson?
So can you say the first bit? Data market? Sorry.
Go-to-market.
Oh, go-to-market.
The go-to-market.
Sorry.
Like, how, how are you progressing-
No, got it.
... in terms of the hiring?
Oh, you said the data market. I was thinking, I don't know what that is. Okay, so we're, as I mentioned just before, if you follow us on LinkedIn, you'll see that we've just hired, I think, five frontline sellers or maybe more. Pretty strong, very experienced sellers, I think most of them from Equinix Metal, so very, very strong in this space. Understand the products, understand the market, and so forth. And we've also got solution architects who are incredibly strong in that space. Now, what's different about building a standalone business is you would need significantly more folks to take a business like that to market if you didn't have the existing frontline sellers inside Megaport. So what is unique is you have, like, a, what you've created is an overlay sales force.
Think about it, our existing sellers will be paid on anything that gets sold as compute, but they're gonna only know so much about it, and they only need to know enough to be dangerous. The conversation with the customer is literally: "Hey, have you considered... Did you know we had this product?" And they'll say: "Oh, I don't look after that," or, "Maybe it's this person." You say, "Do you mind introducing me to that person?"... and then they send in this specialist. What's so different about those specialists is they're not out there punching the pavement, sort of outbound hunting. We have a machine to do that for them, and they become the specialist in that space.
So they will probably, over time, have a much higher dollar figure that they would bring per head because they have a machine beneath them, if that makes sense. I wouldn't compare them to a traditional frontline seller. They are more of an overlay sales function. They are paid totally on compute, just to be clear, so it's not like some free kick, but it's gonna be a lot easier to get into the 3,000 enterprise customers that we have versus cold calling. So you would... We haven't landed on the exact productivity per head expectation from them. Remember, the tricky part here is that a vast majority of the existing business came from product-led growth, and as I said, we aren't disrupting that, and so this is really the cream on top to help build and scale the business.
As I said, that's probably an 18-month journey before that machine is actually working or at least showing through the revenues. It will, we will see it work much sooner, just like with Megaport, but in terms of the revenue contribution, it comes later. So it's gonna be the core business that against the business plan that we acquired them against, and they've got great opportunity to go and be successful on that, and there's, it's, it's a low risk in that if it's not successful, it's not paid. If it is successful, it's paid. If they're really successful, we pay more, and every which way, it's good for shareholders. Does that make sense?
Tim, I think it's fundamentally applying the same principles as Megaport to the compute and then overlaying that with the India acquisition as a collective group. So that's kind of how we're starting to think about it into FY 2027.
Yeah. I mean, sorry, sorry, Michael. Thank you for that answer. My question was more around the core part of the business. Like, you guys have flagged a material uplift in reinvestment back into the business. So, so how far through you are, how far through are you in terms of finding the headcount that you need for the core part of the business?
Oh! You should have stopped me.
I'm sorry.
I'm carrying on about Latitude. You're allowed to interrupt.
Yeah, you can, Tim.
All right. Very simply, we hired all of those, I think before we... I think we, we'd shared that. Most of that-
Absolutely
that sales force was in place, which is why you've seen, like, our expenditure where it is. Continue to add, but the vast majority were added at the start of the year. You're also seeing probably the impacts of those already come through with significant new logo and all the net retention, all the expansion, and, you know. I think you're seeing the success of that already, frankly.
We'll continue hiring into the second half, though, Tim, kind of in a more steady cadence, but you'll see that's kind of why we provided guidance on the standalone.
Yeah, if you think about it, you've got to move fast for a year. The faster you move, the... Because obviously, if you hire in the last half, you're not actually making that much impact. So there's almost no impact, because by the time you get them ramped, they make no impact to the business. So we're very, very fast there. As we've been incredibly fast with the Latitude folks as well. That's what we do. We hire wicked talent, and it's actually... There's a lot of people that love to be at Megaport, and it's not hard to hire because we've got such a big machine of folks that recommend wicked talent to us. It's very rare.
By the way, every single go-to-market hire in the end, and pretty much everyone that's coming through is coming from a recommendation from someone inside the business. Makes it so much easier.
Understood. Thank you.
Our next question comes from Bob Chen. Bob, you may unmute.
Hey, Michael. Hey, Tish. Just a question, follow-up to your comment earlier around the biggest constraint being compute for the Latitude business. I mean, there's been a lot of noise around shortages as well as price increases for DRAM and servers. Like, how easy is it to pass on the price increases to your customers? And then, what are you guys doing on the supply chain side to try and mitigate the shortages?
Yeah. The good news is we've got choice in terms of the SKUs that we can procure and the vendors we can procure from. There's definitely been. I mean, for those who don't know, there is a pretty significant memory challenge globally at the moment. I think memory price is up 300%-400% in the last few months alone. I think it was OpenAI, I think, took out a big line. I think Western Digital's even said that they've not even taken orders. So there is, there's definitely. That's flowing through the entire industry. I think the scale at which hyperscalers are procuring this style of infrastructure is pretty large.
Even though, I guess, we're, you know, it's a sizable CapEx for Megaport, it's still a very small component on the global scheme of things, and so there's plenty of different providers that we can leverage. I think there's five or six different paths we can procure, and we've been changing, depending upon pricing and opportunity and who's willing to work with us, and so forth. We also, you have to spend time escalating. Unfortunately, when you end up in a position where everyone... What happens is everyone tries to jump the queue, and kind of the noisy wheel gets the oil, so to speak.
And we've had that with some of our vendors, where we've jumped in, they've sort of pushed out and delayed delivery, and then we've jumped back in, and they've actually pushed us back forward. So at the moment, we're in a very good position. A lot of ordering and infrastructure was prior to price rise as well, which is helpful. We will see how the pricing plays out in the market, but it's, it's likely that ultimately someone—prices will typically get passed on to the consumers over time. You can't withhold those prices, and that'll happen across the board. We don't need to do that for the short period b—where we're at, but at some point we will, and, and then all of that will get washed through the business. So it's, it's a very...
You're constantly monitoring pricing in that, in that game and making sure that you're competitive. It's returning a great margin, and so forth. So what's probable is that you'll probably see... It could. These are the funny things. They, they, they're hard, and sometimes they play into your favor because the rest of the world can't get access to something that you've got access to, and so you could, you could see things change, but let's just... Let's just see how it plays over the next six months. This space changes almost daily.
Great, thanks for that. And I guess it doesn't change the underlying business case on Latitude at this point in time, at least?
No. And if you think about it, it, it's when we went through this, it's kind of discretionary growth. If you, if you grew too fast and you started to burn too much CapEx, you just slow down the CapEx you deploy, if you ever got to that position. So it's, it's kind of a discretionary business, where you can control the inputs in terms of the CapEx that you're deploying, and then you can monitor it with pricing in terms of, like, and to manage the utilization place. You, you've got levers, if that makes sense. So it doesn't run away from you, unless you choose for it to do that.
Great. Thanks, Michael.
Our next question comes from Roger Samuel. Roger, you may unmute.
Yep. Oh, hi, morning all. Just a quick one on your EBITDA margin. So you mentioned that the exit run rate was 21% for the half. Now, given that you reported 26%, is there any possibility that you might go below the 21% to 24% range in the second half? Yeah, or is that 21% is the floor for the second half?
Roger, I guess for us, it's more highlighting that that's 26%. You've got one month of Latitude in there, which is a higher margin business, heavier CapEx, higher margin business. The reason I've highlighted that from a Megaport underlying EBITDA exit margin is because we will continue to add costs in the second half. It'll be a mix of recurring and non-recurring spend, and so yes, that's why we've provided guidance to the 18%-20% mark-
Right
- for the Megaport.
Okay. No, yeah, I'm just wondering if you may go below the 21% figure in the second half, given that you give a range of 21%-24%. Yeah.
No.
Yeah. Okay. Thank you.
I think, yeah. Oh, you mean on the group? Yes. No, that's why we've given the-
Yep
range on collective. Yes. Sorry, just ensuring I understand the question.
Even for the second half as well.
Our next question comes from Suraj Ahmed.
He's back!
Suraj, what happened?
Welcome back, mate.
Yeah.
We missed you.
I'm back. I think it's a good follow-up to Roger's question, actually. Can you hear me okay?
Yeah.
We can. Yep.
Yeah, Tish, one for you probably. I mean, you did add, what? AUD 10 million costs, half on half in the core business, right? The guidance, especially given exit rate's only 21%, but to get to 18%-20%, it actually has to go dip to sort of 15% in the second half, right? That's a big step up in cost in the second half, but Michael just said that you did most of the hiring at the beginning of the half.
But they don't-
A bit confused, right?
Don't forget-
Yeah
don't forget, they don't start on 1 July, and so you-
Yep
add those costs throughout. So it's, it's kind of, it's exactly the same thing that we talked about last year, I think almost this time last year.
Yep.
And so it is that collective ramp, and again, that's why I kind of highlight that there is recurring and non-recurring spend within the business, and that is, that is part of... You think about Megaport standalone, you've added two acquisitions. We do need to make sure that we've got collective marketing and activities built into the second half. And so that's why I really want to focus from an EBITDA standpoint on the collective group margin, because it is-
Okay
the sum of components. So that, that's kind of why we've tried to provide some guidance on that at the collective group, 'cause I think that's really important.
Yeah, that's super helpful. So just to clarify again, given your comment on recurring and non-recurring, so maybe we shouldn't be using the second half margins as the... or for full year 2027, because that's what I'm getting a lot of questions on. Because that sort of implies, you know, run rate in next year is much lower, right? But you're saying there's non-recurring spend as well, so all of that doesn't carry into FY 2027.
It's a mix. It's a mix.
Yeah.
And so, yeah—
Got it
I think you've got two very different margin businesses coming together, so, and the India expansion. So there's just a sum of all the components. So we'll provide guidance for FY 2027 at the full year. So, but yeah, you're kind of on the right track at least.
Okay. Yeah, and can I just follow up on one thing? The synergies that you made-
Of course
the synergy question, which Michael addressed, was revenue. But on the cost side, I think what Latitude is now going to use your backbone or your network, right? So does that mean there's a bit of cost synergies for the business as well? I know it's not material, but at least Latitude benefits from that.
There is lots of benefits on that. So there's we talked about synergies in terms of just, like, go to market. There's synergies in terms of the platforms as well. Like, if you think about it, we're in all these data centers, we've got these relationships in space, much easier to expand and scale. We can actually do some really cool stuff where we land smaller sites with leveraging the Megaport network, so that there's a significant reduction, and well, not material, but enough to be a reduction in cost to land faster in more locations. So you'll probably see us expand the number of locations a lot easier 'cause we can land a lot, a lot simpler, and then we can scale from there. That's because of the Megaport backbone.
But a lot of the Megaport components cross in because you've got an ability to say, "Well, we could put a 100-gig VXC," for example, "straight into the Latitude business," and we have a choice as to how we can charge that to a customer or not, or offer a differentiated service. So a lot of it comes through. There's a lot of differentiation that comes out of it. There's a lot of innovation that we're going to bring from an enterprise perspective. Think VPC, which is a basically enterprise-style networking components into the compute stack, very similar to what you'd see with a cloud provider. We'll build that and integrate that into the platform and offer it. So, the two businesses are pretty tightly aligned in many ways, and benefit each other on both sides.
We've got one more question. Thank you.
Our last question comes from Paul Mason. Paul, you may unmute.
Hey, team. I just wanted to ask a bit about the NRR and services numbers. Like, the services numbers have shut the lights out, right? A much bigger set of additions than previous halves. Just wondering if you could give any color on the contribution from the second half 25 cohort to that. Like, is, is what's driving this big step up, like, the, the relatively new customers that, you know, you had a big surge last half? Or is, like, the, you know, the upselling more coming from just, like, the broad base at this point? And I suppose where that's leading is because you had a second half in a row of really good customer adds, is this then gonna re-accelerate those numbers, like, even more significantly because it's all coming from, you know, a much bigger set of new customers?
Well, you know, we've been on this journey for a while, explaining the impact and drivers around net retention, and also the impact for what's important for the company. The answer is both, of course, and that's why we keep sharing that, A, net retention, so we're selling more stuff to the existing customer base, and that's why they're taking up more services. They're much bigger services, there's more revenue associated to them. But also, when you have more products, you land more logos because you can meet them at a different sales cycle for each component that you have. So it...
This mission that we've been on, I don't know when it was, two years ago, and we said, "Hey, these are all the levers to fix the net retention." It wasn't just fixing net retention, but it was like adding new products will help you expand more, earns you the right to sell more. But what's really interesting, and we kept sharing that, is there was a period of time before, sort of, I can't remember what year it was, like 2022-2023, where new logos had declined. And new logos on the back of no sales force and lots of things, but we won't go back down that path. But when you have really, really high lands of new logos, you typically follow. That gives a positive tailwind to net retention as well, because they expand faster than an old logo.
So when you have... It's like the perfect storm. We've got all those elements working at the moment, and then you've got a great market in the United States, specifically demanding all of these services, and we have the platform that delivers it. So there's not really anyone else that can do what we do. So we're in that really, really great position, which is why you're seeing that come through in all the metrics. Like, this is an outstanding performance from the underlying Megaport business, and on top of that, we've acquired some really cool acquisitions. One, in particular, is gonna take us to a really different company in the future. So, yeah, I'm glad you noticed it.
If it's all right, if I can sneak in one other quick one. I was just wondering, you know, one of the things that you introduced when you joined as CEO was, like, this sort of solution selling, and, and Global WAN was, like, something we hadn't heard out of Megaport before you arrived. Have you got any ideas on, like, something like that, that would bridge Megaport and Latitude, like, actually a, you know, a cross-platform solution, sort of, that could be a new go-to-market motion for you guys yet, or-
Yes.
Something like that?
Yeah, absolutely. These, while they are different technologies, they serve the same function. Like, if you think about it, I mean, you know this, but for folks on the call, if you look at IT infrastructure, there are only three things: it's network, compute, and storage. They either live in a data center, they live in a cloud, or they live on someone's on-premises, literally. And so by stitching these three things together, so I say three because we will look to build out a storage business as well, when you stitch those three together, you solve the customer problem. The outcome is you want to run an application, and the application wants to be served up in a very high-performance, low-cost, predictable manner, and that is what this solution delivers.
And then you want to be able to make that resilient across multiple locations, time zones, countries, you name it, and the network, and all of these things stick together. So it really is a beautiful combination of what will be these three businesses moving forward, at least the compute now and the network, and then we'll add the storage element to it. That is, couldn't be more like a solution selling discussion. But the cool part is, you can always land a customer at any point. You could just land on a GPU that a customer wants to use, and then you have the right to cross-sell into the network element, add compute, potentially add storage. We've got this 3,000 enterprise customers on the network side.
They've got sort of close to 2,000 on the compute side, and we have this ability to sort of take both to both sides. So it's really exciting. It will take time to build that, but that is what the, that is what the opportunity is ahead of us. It's really cool.
Okay, great. Thank you.
All right.
That brings our Q&A session to a close. I will now hand back to Michael for closing remarks.
I think we're done. That was a little longer. Thanks for those who hung in with us. It was longer because of the two acquisitions. This time next year, we'll keep it a lot tighter. We wanted to make sure everyone had an opportunity to understand the businesses, understand the changes around the guidance components, the strategy of why we acquired, what it looks like from a product perspective, give you some insight on some really cool innovation that we're already launching inside both businesses. Thank you for your support. It's an incredible opportunity for us, and we're just going after it. Yeah, look forward to catching up on the roadshow.
Thank you.