Salesforce.com made another announcement this week. This time they have announced a new integration facility that is based in part on the announced but as yet undelivered Apex programming language.
Not to quibble, salesforce has had a good year, but the company is slowly beginning to get into a futures game. The Apex language as well as the integration facility, ApexConnect, which was announced this week, are both scheduled to be introduced in the first quarter of 2007. Since the company has been very good at making its delivery dates on pre-announced products and since we’re into that part of the year when much of what many companies do is to plan the next year and try to close the books, there is little exposure for the company or its customers. Nevertheless, with two announced but undelivered products on the books I think Salesforce is starting to get a little ahead of itself.
It reminds me of the 1980’s when database companies would routinely announce features and functions months ahead of time. It’s a smart tactic that sometimes freezes buying decisions and drives competitors mad. The logic is something like, “Let’s not buy anything until we see the newest whiz-bang” even if that means waiting many months.
But wait a minute. Who is the competition? There isn’t much. Competition in the on-demand development, deployment, and integration space is rather limited. It might make sense to make an announcement simply to keep the brand in the public eye but announcing a product like this just to gain some publicity seems a stretch. Perhaps the announcement was aimed at some systems integrators and their customers who need to make decisions.
There will be plenty of need to make noise in the market in the new year and plenty of opportunities too, and NDA briefings go on all the time between vendors and prospects in the sales cycle so I am not sure why there was such a hurry to announce ApexConnect so early.
Still a good idea though
That said, the idea is solid and, I think, necessary in the grand scheme of things. There are already products on the market that do some of what this announced product will do. For example, Cast Iron Systems offers an integration appliance that provides very rapid connection to products like SAP for Salesforce users and AboveAll, offers a software solution that does similar things, just to name two.
Nevertheless, as a vendor of a development and hosting solution, it was incumbent on Salesforce to provide functionality for a broader spectrum of needs. For instance, in addition to connecting to popular packages like SAP and Oracle ERP, Salesforce also had to provide integration for its more than 400 AppExchange partners, middleware partners, other third party application developers, and the systems integrators that are building practices around all things Salesforce. That’s a lot and rolling it all into a single strategy makes sense.
According to the company, this new approach will enable a variety of vendors to pre-integrate solutions and thus reduce the cost of integration, which brings us to another reason for Salesforce to act—economics. According to Gartner, integration costs account for 35% of the cost of design, development and maintenance regardless of the size of the company or the applications.
That’s like saying integration is a green field for a company like Salesforce that has made a living out of doing applications better, faster, and most of all cheaper. If ApexConnect can help bring that cost down for Salesforce customers it will be another reason to buy Salesforce over traditional solutions.
In this announcement I see a subtle change taking place in the market. Historically, it was the responsibility of the customer to buy the necessary equipment and hire the systems integrators but increasingly that responsibility is being assumed by vendors. It makes sense too. IT seems to be the odd duck when it comes to creating basic infrastructure. Think of any utility—electricity, telephone, water, gas, even cable—all these infrastructures were produced by direct investment by the utilities themselves. If the trend is for the SaaS vendor to increasingly take on the integration and thus infrastructure responsibilities that trend is simply returning to historic norms.
More than simply making the connection, though, this announcement, therefore, is about something bigger. It seems to me that we have always taken a rather simplistic and clinical view of integration. In the past, we integrated systems to make it easier to get data back to the headquarters where the accountants could count the money. Today, though, integration is likely to be more about supporting the customer and the end-to-end business process. Also, integration is to be accomplished to support data flow in both directions—Salesforce might update SAP but the potential exists that SAP will need to update Salesforce too.
With ApexConnect Salesforce has announced the ability to update backend systems but other parts of the company’s integration strategy already make it possible to do the reverse. The net result gets us to the round trip for the business process and that should go a long way to helping Salesforce and its partners carve a bigger piece of the enterprise pie.