Many business-side decision makers’ marketing, sales are 10 months to five years behind customers in their marketing, sales and services strategies, still clinging to threadbare IT paradigms like client/server data centers and monolithic legacy applications, according to Peter Coffee, Salesforce.com vice president and head of platform research. It’s up to IT to recognize their mandate to build business-IT alignment and move focus from cost centers to customer behavior and using cloud APIs rather than buying systems.
“The customer is way ahead of business now,” said Coffee. That lag has existed historically, but getting rid of it is critical today because of the speed of technology and application development consumer adoption. “Today, business and IT have to watch the consumer…like hawks.”
“They spontaneously realized that a replication strategy of always having to think about synchronization of content across devices is just a complete waste of time,” Coffee said. “And I didn’t have to help them figure that out; it was just completely obvious.”
With this example of consumers quick acceptance of using cloud and new devices, Coffee launched into a discussion of business and IT alignment, the right and wrong ways to pitch and use cloud computing, the increased importance of cloud APIs (application program interfaces)) and more.
Some business decision makers believe that cloud computing’s cost reduction promises are hokum. What’s true and not true about that belief?
Peter Coffee: This usually stems from the initial pitch for cloud. I warn IT people to not get so enthusiastic when talking about the economies of cloud computing. If you sell it as a cost-reduction strategy, you’ll be digging the grave for high-ROI strategic projects. Execs will cry, ‘But you told us that cloud is about reducing IT spending.’ Your explanation of the project’s 30% ROI and it paying for itself won’t hold water, because your cost-reduction pitch was equated to ‘cut spending on IT’ by business.
Don’t create that cost-reduction headwind that you will have to sail into. Sell cloud on the value. Sell it on the ROI. Sell it as a new world of IT, where deployment happens quickly and it works the first time. Sell it on saying goodbye to the cost and competitive disadvantages of failed development projects.
Do many businesses feel too locked into big enterprise systems to migrate to cloud?
Coffee: I don’t use the migrate word. It’s not about migration. It’s about addition. Migration implies that you’re going to take the complexity you have that you don’t like very much and move it somewhere else and declare victory.
That stuff in your on-premise data center? Think of it as Chernobyl. You put a concrete cube around it. You pull data out of it. But you don’t go into that toxic cube to do your innovation. You do your innovation where it’s safe and fast and inexpensive and deliverable. That’s SaaS cloud. You add your mobile and social uses and applications using SaaS. You treat that concrete block of poison in the basement as what it is. It was expensive, so you don’t want to replace it. You don’t want to spend a lot of time inside it either.
What are the tangible benefits of pairing legacy applications with SaaS?
Coffee: SaaS takes legacy apps to new levels. For example, in the cloud, enterprise resource planning (ERP) becomes an actionable tool instead of a bulky, cumbersome archival and reporting system. Now it’s a sales tool, a support tool and a competitive edge that it never could be before.
I call it “liberating the latent value” of your past IT, because cloud essentially provides an historical archiving function for the headquarters staff. Consider the uses, such as legacy apps properly, selectively and securely deployed to handhelds enabling actionable real-time information for field sales and support. That one use case supports the customer, for instance, by providing real-time looks at inventory. Think about the implications if every time your customer goes to place an order, they already know it’s available. Think about what that means for your cost of sales.
Businesses have spent so much money on ERP over the past 10 to 20 years and haven’t even begun to tap that extensibility.
Private cloud or cloud hybrids proponents say that those options can offer more device support flexibility for large legacy applications, too.
Coffee: You remember that line in the first Men in Black movie? ‘Now I have to buy the White Album again.’ Well, that’s what systems vendors want to sell. They want to sell you the Beatles’ White Album again. They sold you a mainframe, and then they sold you some minis. Next came client/server systems and middleware, and maybe they sold you SOA enterprise service bus. Now they want to sell you a ‘private’ cloud. My response is, stop buying technology.
Businesses want to lease services, not buy technologies, to get function. They want to make their SAP or Oracle or whatever infrastructure more of a business resource.
So, would you not consider APIs a technology SaaS sells?
Coffee: Fully disclosed cloud APIs are the new resource that we bring to market. We don’t deliver any hardware or software. We only expose APIs and get paid for the value that people get by invoking those APIs. Obviously, the better job we do of making semantically-complete, rigorous, functional APIs, the more value we have to sell.
Let’s go back to the legacy apps discussion and look at API-level integration with SAP. Some businesses find it easier to use Salesforce APIs to add value to an SAP installation than to wait a year for an SAP update that doesn’t work quite well and costs twice as much.
Isn’t it a difficult mind shift for business and IT people to suddenly have to consider the cloud API so important?
Coffee: People are accustomed to the idea that integration is done by writing code. They also know that when you do those integrations, you wind up polluting the code layer of what was previously an off-the-shelf product. The result is an application that’s no longer readily upgradeable without expensive and difficult regression testing. APIs provide a continuum and not a dead end.