Warehouse management systems were relatively late in coming to the cloud. Dan Gilmore, chief marketing officer with Softeon, puts the issue into context and lays out the various options for cloud deployment of WMS.
SCB: What should people know about WMS in the cloud?
Gilmore: As prominent as this has been in the last few years, there's still a lot of kind of confusion out there. If you think about WMS in the cloud, you have three variables or attributes. First is the deployment model. Is the software going to be deployed on premise, as it's typically called, which means either within the warehouse itself or the four walls of the enterprise? Or is it going to be deployed in the cloud in some fashion? That's variable one. Variable two is the pricing model. Is it going to be traditional license pricing, upfront pricing, or more of a subscription or transaction model? That's the second dimension. So you've got a deployment model that’s on-premise or in the cloud. You've got pricing model that’s based on subscription versus an upfront license.
That gives you four choices. The traditional model is on-premise deployment with upfront license pricing. The emergent cloud model, of course, is cloud deployment with transaction subscription pricing. But you can have hybrids of that. You could have traditional onsite deployment, but still use a transactional pricing model. Conversely, you could deploy it in the cloud and still have an upfront license model. Those are the things that people can look at in terms of deciding what they really want to do and why. And then the third dimension is the management model — who’s actually going to be responsible for managing the application over time. That can either be your in-house staff, or the vendor itself.
SCB: Does that level of complexity begin to hint at why WMS was a little late to the cloud party?
Gilmore: There's a variety of factors. Number one is that some of the larger WMS players were indeed a little late to the party, and held things back. And then some of the original cloud-based vendors had somewhat limited functionality. But the biggest reason was the fear that with material-handling systems, response times were just going to be too slow. They require very timely, real-time-type communications and messaging. The good news is that this objection has really kind of gone away.
SCB: What about issues of reliability, continuation of service and the like? That was a concern with the cloud with all types of software.
Gilmore: The big thing about WMS is if it goes down, you're not shipping product. And if you're not shipping product, you're not billing revenue, so that's a little different than a forecasting application or something along those lines. Clearly, there were some concerns about uptime. I think now the bigger concern is about cybersecurity. But there are a number of security protocols and standards that leading vendors are putting in place to minimize the risk of that as well.
SCB: Of all these different options that you have outlined, what are the trends now in actual WMS deployment in the cloud?
Gilmore: There’s no question that we’ve seen a pronounced movement over the last two years to cloud deployments. That's in the neighborhood of about two-thirds of companies that are now going for cloud versus traditional on-premise. Having said that, you can see some hybrid models. Say, for example, that a large 3PL has a dozen sites. They may have nine or ten in the cloud. They have two for perceived bandwidth reasons, or other kind of factors, that have led them to go on-premise. So it's a hybrid model. In some cases then you can migrate that on-premise deployment into a cloud deployment if you've got the right architecture to make it seamless for the company.
SCB: So it's not so much that WMSs are necessarily site-specific. They can link multiple warehouse facilities with the same WMS, all being stored in the cloud in the same place?
Gilmore: Yes. For a given company you could have one cloud instance that would be able to support multiple facilities as well.
SCB: What are some of the lessons that we've learned about WMS and the cloud?
Gilmore: First off, the benefits are very real. Deployment time is reduced by 50 percent. Having said that, people sometimes are a little confused that there are a lot of the things you have to do regardless of the deployment model — testing, training and other factors. That's a constant. So, while there are real benefits in terms of time to value and speed of implementation, you still have to do the basics and the 101 stuff you need to make a successful deployment.
Second, we're finding that the market, at least in North America, is more interested in what's called a private cloud deployment versus a multi-tenant SaaS. In other words, you are in the cloud, but you get your own cloud instance, if you will. That may run across multiple facilities, but it's not the same exact system for everybody. You've got all these things that are unique to each scenario. A cookie cutter-type thing works for something like Salesforce.com, but not in the WMS market. Maybe it will someday, but right now, that doesn't seem to be the way customers want to go. However, we are finding internationally, for example in Latin America, that maybe a Saas model does make sense, because implementation is simpler.
The last important point is that the concerns about response time that we talked about earlier are turning out not to be a factor. There are a couple of ways you can address that. First off, even if you have a cloud deployment, you can put local software agents inside the distribution center that handle real-time communication with things like voice or material-handling systems. That’s a way to mitigate concerns about bandwidth. But frankly, we're finding that most of the time you don't even need that approach. The bandwidth that's available in the marketplace is sufficient to run a high-volume WMS, and give you the kind of millisecond response time you need to be successful.
