If you’re in the IT analyst business (either the “industry” or “Wall St.” variety) software-as-a –service (SaaS) as a major or even dominant delivery mechanism is a foregone conclusion by now. In that context, this blog may appear pathetically late. However, if you spend any time talking with folks that don’t make a living out of IT you find that most are just warming up to SaaS as a mainstream choice. They may just be catching up with the digitally aware, but catch up they certainly will over the next several years as the key barriers to SaaS adoption continue to diminish.
I don’t see SaaS displacing on premises deployments in the coming decade. I do believe that the hybrid model will be very common as on-premises vendors structure their offerings to allow companies to add portions of the capabilities in the cloud. This SaaS portion of “the stack” will be especially attractive in cases where those participating in a business process operate outside of the firewall (such as the sales force or field service) or where a portion of the value of the service comes from delivering a volatile external data set (tax rates, product pricing) or where the software does something less than strategic (such as travel and entertainment expense).
There have been three generic barriers to adoption that have held down the market for enterprise SaaS applications: security, functionality and data integration. To be sure, there are several areas where in practice where these barriers have been low or a non-issue (sales force automation or T&E, for example) but in others (ERP, for instance) they remain a sticking point. Several years ago there was a fourth –SaaS is just another fad – but this was certainly demolished when Salesforce.com‘s revenues crossed the US $1-billion mark.
Security concerns dogged the SaaS model from the start, especially where business sensitive data is concerned. Technology improvements and external certification of vendors’ should have addressed these concerns but there continues to be a reluctance to “put our data out there.” However, I think this perspective will be replaced by a growing recognition that a company’s data may be more secure if it’s not on the server down the hall in the closet. A company may back up servers nightly but in the event of a water or fire disaster they won’t be of much use. Relatively inexpensive 32 gigabyte thumb drives have made large-scale data theft a simple reality. And let’s not even talk about disgruntled employees. I’m not sure when the tipping point of the mindset will take place – the point where enough people perceive equal or greater security off premises – but it will happen.
A decade ago, when hosting ERP systems designed for large corporations was initially considered, proponents thought midsize companies would be willing to take a plain vanilla version. This proved to be a delusion then and remains the case today. To be sure, some companies will compromise when it comes to replicating existing business processes in new software, but some capabilities are too critical. Functionality has not been an issue for some types of categories (T&E, for instance) or has diminished considerably as vendors have built out the capabilities of their offering, but (for example) ERP continues to have significant enough limitations for some companies. Nonetheless, over the next five years I expect most of these functionality barriers will evaporate. Limitations of the Web interface has also been an impediment in some cases because of technological limitations and here again, emerging richer Web clients will alleviate this issue. It’s also not too hard to imagine a day when a SaaS-delivered application in practice will prove to be functionally richer because vendors will facilitate and promote companies using more of the capabilities of their software. Our benchmark research finds that a majority of corporations underutilize their ERP system’s core features, including those that would enhance their company’s effectiveness.
If all your data is “in the cloud” then integration of your data with a SaaS-based application is going to be less of an issue. However for more than few companies and applications (such as call centers in larger corporations) the need to tightly integrate data from multiple data stores into processes with short response requirements makes data integration a show-stopper. Solutions to this challenge here are emerging from, for example, Informatica.
The growing importance of SaaS as a delivery option will drive software buying in the coming decade, making a wider range of software-driven capabilities available to corporations, especially midsize companies, which can now run their own call centers. Cloud computing can enhance a company’s ability to collaborate and must be part of a CIO technology and information strategy moving forward as my colleague pointed out. For software companies, having a SaaS option for all or parts of their offering suite can provide a stream of incremental revenues and increase their “share of wallet” when it comes to an organization’s IT spend.
SaaS adoption barriers are falling and SaaS spending is rising. SaaS is taking longer to ramp up than proponents forecast at the start of the decade but I think that’s the way these things always work and, besides, it’s taken longer for many SaaS vendors to deliver software that serves as a real substitute for the on-premises option. SaaS will not replace on-premises anytime soon but its share of the IT budget will accelerate in the coming decade and resistance to embrace it is really futile.
Let me know your thoughtsor come and collaborate with me on Facebook,LinkedInand Twitter.
Robert D. Kugel - SVP Research