It's not just cloud vs. on-premises anymore
In technology, everything evolves, mutates and, eventually, goes away after being replaced by something altogether better, newer, cooler and more powerful. Sometimes, even new technologies experience rapid evolution. Cloud computing is definitely in the mutating and evolving phase of its existence.
Let’s look at how cloud computing has changed in the application software space so far.
Early cloud pioneers, think salesforce.com and Plex, were launched in the late 1990s, offered a multi-tenant SaaS (software as a service) solution. Their chief selling points were economic in nature. There was no software to install, no on-site hardware required, etc. Many early cloud SaaS solutions even came priced on a monthly basis. These solutions could scale up or down based on business need.
More cloud pioneers emerged - particularly in the HR and Talent Management spaces. Workday, SuccessFactors, Taleo and many more emerged. These firms, like their predecessors, not only had multi-tenant solutions, they also provided a very important economic benefit: the vendors performed all of the application upgrades and maintenance. This one benefit turned out to be approximately 40-60% of the total cost of software ownership (based on reviews we did in 2010).
After a time, another crop of SaaS solution emerged: SaaS-querades. For the most part, these were older, on-premises solutions that were now being hosted in a predominately single-tenant fashion. The product functionality was often antiquated, the user-interface was obsolete and the economic savings behind these products was seriously comprised. Since customers, not the vendors, still had to provide the time and labor to upgrade and maintain these applications, the business case for these products was lacking.
While all this was going on, a number of old-school vendors were continuing to churn out FUD (fear, uncertainty and doubt). They, unsuccessfully, argued that cloud solutions weren’t safe, weren’t mature, etc. In time, as cloud applications matured and as users have grown accustomed to other cloud applications, the uptake of cloud solutions is in full swing. FUD is not working anymore.
PaaS (platform as a service) become a reality a few years back. Salesforce.com introduced Force.com, its application software exchange and other tools to enable many other firms to develop applications on their cloud platform. NetSuite introduced its NS-BOS platform, too. Apple even has a vibrant ecosystem with a separate application store for commercial-ready software products. Platforms have given birth to all-new cloud application solutions like FinancialForce.com, Kenandy, Rootstock, etc. This evolutionary step to PaaS has been successful in some ways but needs more support/vibrancy to continue.
Cloud evolution, at the end of 2012, resulted in the best solutions offering a number of cloud applications in a multi-tenant environment. The better firms possessed a strong channel ecosystem that contained top-flight integrators and software developers. PaaS technology and application stores make these vendors’ ecosystems more desirable. By the end of 2012, the best solutions offered in-memory database technology, powerful analytics (also in the cloud) and a robust set of PaaS development tools. However, even the best solutions rarely offered all of the components mentioned to date in one solution.
Cloud vendors (and their competitors) have continued to innovate and their newer adaptations are interesting. Some of these changes may be significant and market changing. Let’s look at a few of these:
- Elasticity – One of the beefs larger firms have about cloud software solutions is that upgrades may be too frequent and require movement to a new release on a calendar that doesn’t match that of the customer’s needs. Vendors have solved the frequency problem with most adopting a schedule of 3-4 releases per year. This is an improvement from one vendor’s 160 releases/year. What these customers also want is the ability to have more control over the timing of when an upgrade occurs. Granted, many cloud solutions give users a multi-week window as to when to upgrade. But, some firms want the option to wait months (or longer). They may want to sit on the current release and delay upgrading for a couple of releases. This is key for firms who have many moving parts (e.g., frequent mergers/ acquisitions/ divestitures, numerous legal entities, a huge, volatile IT portfolio, etc.). These customers want some elasticity in the new upgrades/updates. They want what Agresso and NorthgateArinso offer: the ability to stay up to a couple of releases back with no adverse effects.
- Forgiveness - Look into your cloud solution provider’s built-in ability to help you recover from some accidental deletion of data. Let’s say you inadvertently deleted a chunk of your customer database within your cloud CRM solution. You might learn that the vendor’s restart/recovery capabilities are there for the vendor’s mistakes or disasters – not yours. Quickly recovering from your mistakes (or, in some cases, an employee’s malfeasance) can be really difficult, expensive and time consuming. Cloud application software users want something akin to what insurance firms offer: ‘accident forgiveness’. They want a way to rapidly, cheaply and painlessly recover from a goof. Few vendors automatically provide this at this time.
- Completeness –Many of the original SaaS innovators had only one initial application (e.g., performance evaluation). Older, on-premises vendors often possessed large enterprise suites comprised of many applications. As a result, on-premises vendors argued that cloud solution buyers would have to stitch together a bunch of best of breed applications to get the same functional coverage they’d get from their firms. The reality was different. These cloud buyers found that integrating one cloud application to another wasn’t a big deal. Vendors like Dell/Boomi, IBM/CastIron and Informatica made the integration of cloud applications to other cloud and on-premises solutions quite easy. Moreover, cloud vendors have continued to add to their product lines. Some cloud vendors, by virtue of their prior architecture decisions, were able to move most of their on-premises functionality to the cloud. For those vendors that took an inelegant SaaS-querade approach to moving to the cloud, they only achieved the movement of an older product line to a single-tenant hosted existence. The best vendors moved an entire product line, with all of its functionality, to a multi-tenant world. So, what’s happening is that large suites are now available in the cloud either via all new innovation or movement of older products to the cloud. These larger suites of cloud applications are popping up everywhere. The breadth and depth of these are impressive and growing. Larger suites mean trouble for the on-premises vendors. If they weren’t concerned over cloud vendors picking off customers who wanted a peripheral application or two, now they have to be worried when a cloud vendor takes the lion’s share of a customer’s application needs.
(If you’d like more information on any of these three topics, I’ve developed three different white papers on each of these. Agresso has licensed these and they are available for download here. The papers are titled:
The Elastic ERP Solution: A less rigid approach to multi-tenancy
Agresso Accidental ERP Disaster Forgiveness: A new type of cloud deployment
Changing Rules for Selecting ERP Software (the Completeness issue))
What’s next in the evolution?
Market watchers and software buyers will want to watch the following:
- User Experience/User Interface – Software vendors, particularly those with older offerings, will have to upgrade their dated user interfaces. The smart firms will move away from forms-based approaches and move to more user-defined, process driven flows. The very best will use analytics to predict what information users need and intelligently guide the user (via process management technology) through the completion of the task.
- Containers - Oracle has this capability in its 12c database. At first blush, containers could offer a way for Oracle to offer a multi-tenant experience for its largest customers while physically isolating each customers’ data. I am, though, a big believer in in-memory processing (over traditional databases) and would want to understand how this container technology would be superior to in-memory storage. How this plays out will be interesting to observe.
- Integrated and expanded big data and analytics – Workday has created a way to permit a tight coupling between its in-memory transaction data and external big data. It brings an ability to change what we know about businesses, competition and strategy. It will absolutely change how we look at reporting, analysis and decision making.