At an Oracle Modern Finance Experience event this past week Dr. Michael Mandel, senior fellow at the Mack Institute of Innovation Management at the Wharton School, forecasted that a total of $2 trillion will be added to U.S. Gross Domestic Product over the next ten years as a result of the productivity gains enabled by cloud services.
The core thesis of a report commissioned by Oracle is that cloud will make many more innovative technologies such as blockchain, artificial intelligence, cognitive computing, machine learning, and intelligent automation available more broadly and faster than traditional approaches to on-premises IT.
That’s a reasonable set of assumptions. But where application providers such as Oracle try to take that argument next approaches the self-serving. A core argument Oracle made last week many times over is that customizing applications is bad for business. The basic contention is that most businesses employ the same basic processes and customizing applications which results in higher costs. While that’s true in many cases involving a non-differentiated business function, not customizing applications to meet the specific needs of a business makes it difficult to create any differentiated value from software investments. In fact, a decision to not customize applications flies in the face of the accepted conventional wisdom that every company is now a software company. If every software company is employing the same set of software in the same way, it becomes next to impossible for any company to establish a competitive edge based on the intellectual capital it acquires around a business process. Arguably, the whole point of becoming a so-called digital business is to establish unique value; not merely reduce costs.
Cost of customized cloud applications is starting to drop
The real issue, of course, is that building and maintaining custom applications has been historically an expensive proposition. The good news is the cost of customizing application is starting to drop. Rather than simply using application as it, it’s becoming easier using, for example, low-code tools to bend the software to the needs of the business versus requiring the business to bend to the way the software works. In fact, many customizations no longer require the skills of a professional developer at all. Instead of limiting how any application can be employed; the real focus should be on reducing the cost and complexity associated with tailoring to the specific needs of the business.
Oracle is not alone when it comes to making a case against customization. Just about every provider of cloud applications makes the same case, including Oracle’s arch-rival SAP. The primary reason they make this argument is that Software-as-a-Service (SaaS) applications become expensive for the provider to support should an organization decide they want to customize some aspect of the software. It’s not in the economic interest of the SaaS application provider to encourage customization.
Many organizations for this reason are eschewing SaaS applications when it comes to the processes they value most. They may happily use a SaaS application to manage travel. But when it comes to ERP applications there’s a lot a benefit to be had from hosting application software on a public cloud to reduce costs without sacrificing customization. Of course, how much can be saved by hosting an ERP application on a public cloud comes down to how heavily it’s used. The more an application is used the more expensive it becomes in the cloud. Conversely, the more an application deployed on premises is used the less expensive it tends to become.
Vendors tend to favor overly broad generalizations based on their own agendas. Managed service providers (MSPs) that owe their living to keeping the best interests of their customers paramount would be well advised to put the overly broad statements being made by cloud application vendors in proper perspective for their end customers.