Manhattan Associates Inc Citi’s 2026 Global TMT Conference

Yeah. Manhattan has been in the supply chain space for more than 30 years. Started out building software and services back in the custom bit days and in the on-prem days, and about a dozen years ago, recreated itself as a cloud company. Today, everything we sell is a cloud-based SaaS. We're the leader in that space. Gartner and Forrester rate us a leader for 18 years straight, leader in warehouse and also a leader in transportation and order management and point of sale. We recently launched our cloud supply chain planning product as well. Our focus has been on consistent growth in the cloud. We recently launched our AI products in Q1, and we've had very quick adoption of those products.

We announced by the end of Q2, already 10% of our install base were using our AI products, either in pilot or in subscription.

Yeah. So, like Eric said, we've been in the business for about 30 years. In the supply chain space for over the last 30 years, people have built and bought software in silos. When we get this opportunity 12 years back to reimagine, we not only looked at the technology and how should we do it, we also looked at the functional footprint and how should we really deliver the functionality. So we questioned, should software be built in these silos, right? What are the advantages you can get when you don't build in silos? So we imagined building a supply chain commerce platform versus a TMS, a WMS, an OMS, or POS. Right. I think the way to think of it is sometimes when you're building these silos, you get locally optimized solutions.

If you have to really, truly get a global optimization, it's very, very hard when you build software in those little silos of their own. So that's the biggest advantage we really see, is the ability that we can look across. If we think about supply chain, there are four or five dimensions: labor, space, inventory, transportation, customer sat. Typically, softwares have been built to solve one or two of those individual optimizations. To really solve the global optimization, you have to look at all five dimensions at the same time. The only way to really do this is to build a holistic platform which thinks about it in that space. So we build the software that way. People still buy in those silos to a large part, and we're trying to change that behavior to some extent.

When we get to the higher-up, they can look at these silos across and say, "Okay, I can get a better optimization." That's the biggest benefit we see. I'll give you one example. Concretely, in just TMS and WMS, if you think about it, you get an order down, and typically it would go to a TMS, which would do the planning, and it'll come down to the WMS, and WMS will reject 3%, 4% of it, saying it don't have inventory. So now your truck is leaving sub-optimized because it couldn't figure out what to do. When you have a unified solution, the same order will look at both inventory when we do the truck planning, and we know exactly what's there, so we can plan those things properly, and you have a lot better optimized trucks getting out. That's one example. Look at OMS, POS.

People have been talking about providing the single face to the customer and building it in two or three different pillars. You got your payments and taxes and pricing, everything duplicated, triplicated between those systems. A customer buys in POS, a call center cannot answer the question because it's a different system. You think of it as a unified, now you can truly provide customer in the single face. No matter which channel you buy from, there is one order, there is one payment, there is one pricing, there is one promotion. Those are the core benefits we see from a unified thing, and it really came back because we had an opportunity to redo this 12 years back.

Yeah. Maybe I'll start with our approach to AI in general, which is somewhat unique because with most cloud companies and SaaS companies that you talk to, they start the conversation around, all right, first you've got to move all your data to a data lake, and then you've got to do data indexing, and then we've got to worry about data security and latency and duplication, and where's the system of record? On and on with all these things. Then finally, maybe you get to a pilot in some number of months. Versus us, because when we built this platform, going back a dozen years ago, and we've stuck to this model of true microservices, all API driven, we are truly a versionless software.

We can run our AI agents directly on the platform, which means the AI agents are using the same APIs that a human user would use. You do not need a data warehouse. You do not have data latency or duplication or any of those other problems that come with so many of these AI deployments out there. That also allows us to move much faster. Again, versionless software. When a customer decides they want to do an AI pilot, we turn on the agents, and they are available to use the same day. Again, because we are versionless, the agents are actually effective that day. All of our base agents can handle the same routines for all of our customers. In the pilot, you also have access to our Manhattan Agent Foundry, which allows you to modify any of those base agents.

If you have done things in your system with extensions, or you have things that are unique, you can modify those base agents, or finally, you can build agents from scratch. All of this happens in a 90-day pilot, so at the end of that pilot, our forward-deployed engineering teams are helping build the business case and prove the value and the ROI of the agents so that they can justify the subscription. Based on that model, we have had 100% conversion from pilot to subscription. We have not had a single customer that has gone through a pilot and said, "Hey, it is interesting, thank you, but I am going to go in a different direction." All of them have continued with us.