Purists, Pragmatists and others re: cloud

Does the cloud vs. on-premise dialogue have to be so polarized? Can't we see something other than stark one-way decisions? Here are some items two IT research analysts have come to terms with.
Written by Brian Sommer, Contributor

Tom Ryan is an ex-Gartner Group analyst and someone I try to have lunch with once a week. We've teamed together on some consulting projects and his supply chain/manufacturing focus often helps me see perspectives in the finance, HR and advanced technologies space I might otherwise miss.  Between bites, we discuss/debate a number of things and lately the topics involve topics like:

  • Cloud vs. on-premise
  • Single-tenant cloud vs. multi-tenant cloud apps
  • Practical considerations where cloud may/may not work
  • How cloud solution uptake in corporations will evolve

Tom recaps a number of these in a blog of his re: SaaS and the Warehouse Management software space.

His take and mine have a lot of similarities but you'll have to read both posts to get it all.

1)      Cloud or on-premise solutions will not be all or nothing decisions for most firms. To hear some analysts and marketing mavens, a company must commit, fully, to the religion of cloud (or on-premise) or be branded a non-believer. There are many churches/philosophies of cloud thought out there and each has their zealous advocates. The reality is that firms are and have been experimenting with cloud solutions and all but a few organizations will be totally cloud or totally on-premise. We need to see cloud adoption as following a bell curve distribution with few firms on the far edges/tails of the distribution. In a survey of CIOs I recently did, every one of them had something in a cloud or was seriously considering getting something cloud-like into their firm. They spoke of cloud -based office automation suites, cloud HR products, cloud CRM, etc. What these CIOs lamented was that they couldn't get enough vertical business applications in a cloud format.  Some functions, they will tell you, may remain on-premise for some time and may never move to the cloud. Bottom line: most organizations will have some cloud apps and cloud adoption will continue to rise. Cloud adoption will not reach 100% and on-premise may never fully go away.

2)      Single vs. multi-tenant software - Lots of vendors see the move to cloud as something they must support at least in Marketing name only. As a result, there are tons of single-tenant, hosted (or as Tom refers to them as ASP-style) apps. Multi-tenant apps have some distinct economic advantages for the user as the vendor is providing the application maintenance, patching and upgrading. The vendor is also able to make the upgrade once and be done. In the single-tenant worlds, the vendor or the customer must do the maintenance for each customer individually. The latter is an expensive model and one that could fail in time. Yes, some collection of buyers will want the ability to choose which upgrades they want and when they want them but this may become a minority of the market.

3)      Not all business environments may be great cloud candidates. Tom writes how certain manufacturing or logistics processes require almost instantaneous switching/routing of parts/parcels. The latency in sending an electronic signal to a cloud system and returning a command in a mere thousandth or two of a second may give the nod to a dedicated on-premise solution. I can't argue with that. I'm not sure I'd want a continuously running facility, like an oil refinery, using a cloud solution to run its operations without an awful lot of safeguards in place. What would happen if a lot of other users from other firms suddenly swamped the cloud provider's solution at a critical moment? Cloud belongs where it makes good business sense.

4)      Cloud uptake will not be abrupt. No, cloud uptake will be more gradual. Few firms can take the risk that they'll convert all of their systems to a new platform overnight. The instances of big bang conversions today are rarer and rarer. Business people will want to carve off chunks of their business for new solutions and do the moves over time. As a result, you will see companies that are mostly on-premise or mostly cloud. Companies will be trending from one deployment mode to another. So, all the folks getting hysterical that "The whole world is going cloud!" can take a breath. The whole world won't go and those that do will likely phase it in.

5)      Some software buyers are more concerned about achieving a key business benefit not just reducing the IT administrative workload. While IT maintenance savings via multi-tenant cloud applications can be significant, some software buyers, especially SMB buyers, are more worried about improving their cash flow, reducing inventory levels, reducing the number of stockouts, reducing scrap, etc. These buyers will choose whatever solution delivers the best mix of operational and IT improvements (not just one over the other).

If you're in Chicago and would like to join Tom and I for lunch sometime, give us a call and be prepared to get your shots in quickly. We'd love to try out our newest ideas on some new folks.

Editorial standards