Posted by Neno Duplan
The announcements by several EHS software vendors this Fall caught my attention. After offering their software on-premisses for over a decade, suddenly many are discovering and planning to introduce multi-tenant Software-as-a-Service (SaaS) while promising to continue to maintain their current on-premises or single tenant offerings. In essence, they are introducing multi-tenancy as if it were a new version of their software. This plan is not going to work! Let me explain why.
Most public announcements begin something like this: In the next several years we plan to expand our software offerings and offer our customers the option to move from their current on-premises solution to the cloud. However, is this even possible? What they consider the “cloud” may not be a true multi-tenant cloud. That train departed years ago, and most of the current EHS software vendors missed it. While multi-tenancy has been a game changer in the tech industry, many are uncertain of exactly what makes an application “multi-tenant” or why it matters.
There is a considerable degree of (intended) confusion in the EHS software space when it comes to the definition of a real cloud or better said, multi-tenancy. Companies that are considering SaaS solutions for EHS software hear all sorts of things from EHS software vendors hoping to tap into the momentum of cloud computing. Many go as far as saying “sure; we can do multi-tenant, single-tenant, whatever tenant you need!” –anything to win the job. These vendors do not understand the real cloud.
Multi-tenancy is a significant shift in computing and requires an all-new approach to the software architecture and the delivery model from the ground up. It is transformational, and customers who intend to buy the next generation of EHS software should spend the time to understand the differences. More importantly, multi-tenancy is a principle, not a software version or an upgrade. It is not an evolutionary step; instead, it is a revolution in the software delivery model and it matters in the long run for the customer.
Figure 1: The single-tenant model cannot easily be switched or “upgraded” to multi-tenant. The software architecture does not allow it for easy switch the same way as single family home cannot be “remodeled” to become a multi-tenant highrise. What differentiates multi-tenant application architecture is its effectiveness in achieving the same goal in a scalable and sustainable fashion.
Can anyone imagine companies like eBay, Salesforce, Google, Workday, or Amazon offering a “single-tenant” solution side by side to their multi-tenant clouds? I argue that any EHS software vendor who offers a single-tenant solution of any type, cannot be a serious contender in multi-tenant SaaS.
EHS software vendors with on-premise software applications or single-tenant web-enabled offerings are seduced by the seemingly low barriers to entry into the SaaS market with an architecture that leverages virtualization. This approach allows a software company to quickly offer subscription-based services of their legacy product to their initial customers. In the long run, however, this multi-instance approach just won’t scale economically. A recent wave of ownership change of EHS software companies is the best indicator that sold companies became victims of their initial success. A SaaS provider who leverages virtualization puts the long-term viability of the business at risk as more efficient SaaS competitors come to dominate the market.
Figure 2: Single-tenant requires many more vendor resources. The resource costs are eventually passed to customers. Each upgrade of the application will require each customer to upgrade independently and the ability to implement tenant management tools and tenant-specific customizations is significantly limited. The benefit of multi-tenancy is that instead of 100 copies of the OS, 100 copies of the database, and 100 copies of apps, it has 1 OS, 1 DB and 1 app on the server with significantly less vendor resources required to manage it. And it is those savings that are on a long term passed to customers.
Figure 3: Multi-tenant model requires less resources and easier (and rolling) upgrades (i.e. no version number necessary). Only one software instance and hardware stack for multiple tenants. All customers are always on the latest version of software. Locus Technologies figured this out in 1999. And they contribute their phenomenal success since then exactly to the multi-tenancy. They could scale up infinitely without adding proportional cost. Others cannot.
Figure 4: “Can’t we create a separate stack for just this one customer? I promise it’s just this one…” Even a single installation for one “special” customer, breaks the multi-tenant model. Don’t do it.
I would also add that single-tenant (hybrid) cloud applications are worse than on-premise installment. Why? Because they are fake clouds. In single tenancy, each customer has his or her independent database and instance of the software. These instances may reside on the same or different servers. In this model, a customer is, in fact, outsourcing maintenance of their application (software and hardware) to a vendor (or their consultant) that is not likely equipped to perform these tasks. No single vendor in the EHS software industry is large enough to undertake maintenance of the single-tenant infrastructure on behalf of their customers regardless of how inexpensive hardware or software virtualization may be. Even if they offer their hosting on Microsoft Azure Cloud or Amazon Web Services (AWS), they still cannot guarantee multi-tenancy as these solutions address only hardware challenges.
The Economist magazine in 2004 described it: “Those forerunners also promised a software revolution by hosting the software applications of companies. But they failed because they simply recreated each client’s complex and unwieldy datacentre in their own basements, and never overcame the old problems of installation and integration with other software. With each new customer, the old ASPs had, in effect, to build another datacenter; there were few economies of scale.”
To improve their position in a shifting marketplace, on-premises EHS vendors have found a way to market their solutions as “cloud-based” when they are not backed by the fundamental principle of what that means. Considering the large investment that is associated with the purchase or licensing of EHS software, it is critical for customers to be able to tell a true cloud product from a fake one. But how can you spot a fake?
Just ask the EHS software vendor these four questions:
- Do you support both single-tenant and multi-tenant deployments of your software?
- Does your software have version numbers?
- Do you charge for upgrades?
- Can we install your software on our infrastructure?
If the answer to any of these questions is yes, the vendor is not committed to only multi-tenant architecture, and you should not move to their “cloud.”
Multi-tenancy is the only proven SaaS delivery architecture that eliminates many of the problems created by the traditional software licensing and upgrade model where software is installed as a single-tenant application on a customer’s premises or at a customer’s or vendor’s data center. In contrast, in multi-tenancy, all customers access the same software on one or a set of linked servers.
Multi-tenancy requires a new architectural approach. Companies have to develop applications from the ground up for multi-tenancy. Once companies commit their limited financial resources to one architecture, it becomes nearly impossible for them to switch to the multi-tenancy model, no matter how many resources they have available. Moreover, for this reason, I am skeptical that many current vendors will be able to make a switch to multi-tenancy fast enough.
A vendor who is invested in on-premise, hosted, and hybrid models cannot commit to providing all the benefits of a true SaaS model due to conflicting revenue models. Their resources are going to be spread thin supporting multiple versions rather than driving innovation. Additionally, if the vendor makes the majority of their revenue selling on-premise software, it will be very difficult for them to fully commit to a true SaaS solution since the majority of their resources will be allocated to supporting the on-premise software.
And if they suddenly introduce a “multi-tenant” model (after selling an on-premises version for 10+ years) who in the world would want to migrate to that experimental cloud without putting the contract out to bid to explore a switch to well established and market-tested true multi-tenant providers? Even Google and Microsoft are playing a catch-up game with Amazon’s AWS when it comes to cloud hosting business. The first mover advantage when it comes to multi-tenancy is a huge advantage for any vendor.
In summary, an EHS software vendor can be either truly multi-tenant or not. If a vendor has installed their software on somebody’s else hardware and runs multiple instances of that software (even if the code base is the same) they are not and will never be true multi-tenant.
Figure 5: Where do you want your software to reside? In multi-tenant or single -tenant infrastructure? If multi-tenancy is attempted on old infrastructure or legacy application upgrade watch out. After vendor built the first few floors of that skyscraper, there is no easy way to replace the foundation. You will be lucky if they end up like the tower of Pisa or Millennium tower in San Francisco. To keep the tower alive they will have to do constant underpinning of the foundation and restrict access to the structure. And you, the customer, will pay for it. That is what many customers of single-tenant EHS vendors are facing today.
Therefore, when considering a SaaS solution, make sure that the vendor is a true SaaS vendor who is solely committed to the multi-tenant SaaS delivery model and has invested in a true multi-tenant platform. This is the only way to reap all the benefits that a SaaS model has to offer.