Information technologists are fond of predictions in which the next big thing quickly and entirely renders the existing thing so completely obsolete that only troglodytes would cling to such outmoded technology. While this vision of IT progress may satisfy the egos of technologists, it rarely reflects reality. Mainframes didn’t disappear, for example. Although they long ago lost their dominant position, many remain key parts of corporate computing infrastructures.
Information technologists are fond of predictions in which the next big thing quickly and entirely renders the existing thing so completely obsolete that only troglodytes would cling to such outmoded technology. While this vision of IT progress may satisfy the egos of technologists, it rarely reflects reality. Mainframes didn’t disappear, for example. Although they long ago lost their dominant position, many remain key parts of corporate computing infrastructures. The IT landscape is a hybrid because technology users have varying requirements and constraints that can lengthen replacement cycles. Most business users of IT pay little attention to the religious wars of technologists because they take a pragmatic approach: They use technology to achieve business ends. This scenario is repeating itself in clamor about another corporate mainstay, the ERP system, which advocates claim will soon be redeployed en masse to cloud computing. That, too, won’t happen. I believe that ERP will increasingly become cloud-based, but it will be in hybrid cloud environments.
For ERP, vendors draw the rhetorical battle lines as on-premises vs. the cloud, but in the end it’s more likely to be a combination of the two. The term “hybrid cloud” refers to an environment where one or more clouds are connected to or combined with on-premises systems. These clouds may be private (that is, controlled by a single company for its sole use), public (a service offered to all comers) or a community cloud (available to a limited set of organizations or individuals). Clouds can be multitenant (where a single instance of some software serves multiple customers) or single tenant (where it serves just one).
Our benchmark research shows that more than half (55%) of 
We see three main reasons why companies have failed so far to embrace the cloud for ERP as fully as in other categories. The most important is that, as I’ve noted, multitenant ERP offers users only limited configurability, and this often is incompatible with what companies need to manage their business. The second reason has been data integration, which until recently could be complicated and difficult to manage. The third reason is that finance departments have been more conservative than most in embracing the Web, especially for ERP systems, because the information in them is sensitive and they fear security breaches. These last two factors are starting to diminish in importance. Data integration between cloud and on-premises systems has been facilitated by new tools and methods. Concerns over the risk of having company data in the cloud are also abating as cloud services vendors demonstrate reliability and security.
Many existing ERP systems already are hybrids. Companies often manage payroll, treasury and travel expenses in the cloud. Any aspect of business related to ERP systems that can benefit from mobile deployment (such as those related to inventory and manufacturing execution) is a candidate for cloud deployment. In the future, companies also will be using smaller (usually mobile) apps that interact with their ERP software for automating the execution of simple processes that require limited data input (such as giving approvals or filling out short forms) or for viewing data. As I noted, Microsoft’s Project Siena, in development now, is designed to enable business users with no programming background to construct such simple applications. Other such systems are likely to follow. Having these apps hosted on the Web may be the most attractive option, especially since most people will be using mobile devices to work with them.
There’s also a case to be made for midsize companies and second-tier ERP installations moving off premises into cloud-based, single-tenant deployments. Some midsize companies (those with 100 to 1,000 employees) may find the total cost of ownership of a cloud-based, single-tenant system lower than for an on-premises approach because of lower costs in implementing and maintaining the hardware and software. Even when those costs are higher, midsize companies that cannot accept the compromises that multitenant deployments entail may find it attractive not to employ an IT staff to support its software and outsource the maintenance of their ERP system to a software-as-a-service provider. For decades, large corporations that use enterprise software for their main business systems have used ERP tools designed for midsize companies to support smaller, stand-alone operations. Putting this “tier two” software in the cloud is likely to be less expensive than buying, implementing and maintaining it. Moreover, in most respects this form of software deployment is easier to control and audit because it has no physical presence and therefore facilitates separation-of-duties requirements in locations with limited personnel. Hybrid clouds also enable companies to maintain the applications in the cloud while conforming to local requirements that customer records be kept physically within their jurisdiction.
Until now, many companies have been reluctant to embrace the cloud for ERP. That is likely to change over the coming decade as security issues diminish and companies decide to take advantage of the cloud’s benefits. On-premises ERP software vendors and their partners can diminish their vulnerability to cloud-based software vendors by creating offerings that improve the economics and affordability of single-tenant ownership and facilitate using their software in a hybrid cloud environment.