As I wrote last year, this "web as platform" gospel is starting to spread quickly from consumer to thoughts on the enterprise. In my mind, what has enabled this enterprise web phenonomenon has been two thoughts – lightweight and simple. Of course, lightweight and simple equals cheap and fast to implement. It is quite easy now for sophisticated users to find and download new software and run it themselves, for them to take simple scripts and tie together various web apps. We are quickly moving to a world where the end user on the edge can and has taken matters into his own hands rather than wait for IT to get something done for them. I call this the age of DIY (do it yourself) in the Enterprise. Why go through centralized IT and their processes when I can get something done with my own departmental budget? Linux, Jboss, and many of the open source opportunities started at the edges first before being brought into the centralized IT organization. As we all know, many new technologies are typically adopted by consumers and then pulled into the enterprise, not pushed. Amazon and other web apps started exposing their APIs and existed long before Salesforce.com.
Jeff Nolan points to an interesting post from John Hagel which highlights this changing enterprise world. What has been deemed as the agile enterprise driven by SOA has actually turned into anything but. The enterprise version of "lightweight" called SOA stands in stark contrast to the next generation web perspective of lightweight. As John correctly points out, enterprise lightweight in the form of SOA means plumbing, it means expensive, it means complex, it means lots of consultants, and it means lots of dollars. In contrast, next generation web technologies are easy, incremental, and driven by the edge and focused on people, not plumbing. While some of these next generation apps may not scale, there is clearly something that centralized IT can learn from the edge, their frustrated internal customer, that things can get done more quickly and more cheaply. As these two philosophies become more tightly coupled we will have some interesting opportunities to invest and make money. While not directly related to this SOA/web mashup discussion, one of the companies I have always found interesting is Splunk which is bringing a Google-like approach to network management. It is downloaded, driven by the edge user, and then pulled into the corporation from the bottom-up rather than the top-down. It stands in stark contrast to EMC’s (Smarts) and IBM’s (Micromuse) way of selling and using their respective products. There will be many more opportunities like this in the enterprise as enterpreneurs leverage user interfaces and technology from the consumer world in the enterprise. Of course, this means a whole new way of reaching customers (frictionless sales), selling to them, and supporting them but this is saved for another future post. The good news is that this new age of DIY in the Enterprise is not going away and is only getting stronger everyday. This also means the creation of many more disruptive enterprise software opportunities in the next 5 years. I agree with Jeff that this is an interesting area to watch and is beyond web mashups-rather, it is a philosophy enabled by all of this new technology, the philosophy of DIY in the Enterprise.
But here is the problem. You write:
“Why go through centralized IT and their processes when I can get something done with my own departmental budget?”
Most (if not all) semi-large sized companies (where I have sent my whole career, including my current 30k person fin serv firm) have poicies preventing this — and indeed have totally centralized IT policies (for many reasons, good and bad. SO something like this is just not possible.
It depends on how you interpret this and what it is. Buying new security software that has to be managed in a data center is not going to happen without IT approval. However, there are plenty of companies that I have funded that sell to department level teams or groups that can sell product without going through IT. Part of the great thing about software as a service, as an example, is that departments can buy on their operating budget and not on a capital expenditure perspective. There were many groups in large companies using Salesforce.com without IT approval. Greenplum just sold some data warehousing software to a department of a large organization which wanted to slice and dice its own data without having to go through the centralized data warehouse. And the initial price was just right for this to happen under the radar. Does this mean a whole organization will standardize on this stuff? Hell no. But as most people know innovation happens at the bottom and only becomes standardized with top-down support. So the adoption of any new technology takes time but it seems that vendors and customers are increasingly trying to understand how to leverage this web phenomenon which means they are paying attention to the edge.
In a simpler response to John’s comment- the IT department can block the installation of local files, but they’re going to have a harder time taking away people’s web browsers. Subscribing to SaaS is one way of dealing with slow/unreponsive CIO orgs/borgs.
My second comment is that a lot of enterprises I have worked with would like to take the applications that are being run in SaaS fashion and bring them behind inside the firewall. This is particularly an issue with depending on startups- you don’t want to wake up and find they’ve run out of cash and you can’t run your business.
Finally, I like the idea of corporate/government users using SaaS in a non-central-committee fashion, but the SaaS vendors then making a behind-the-firewall product available so that it can be transitioned to a controlled environment.
“It is quite easy now for sophisticated users to find and download new software and run it themselves, for them to take simple scripts and tie together various web apps.”
While DIY may get edge users up and running quickly, in the long run this leads to IT sprawl. IT diversity quickly morphs into complexity. We only have to look back at the Mainframe to Distributed competing transition to understand the exponential “management” complextiy web 2.0 in enterprise is going to introduce.
Also, remember enterprise business processes tend to be more complex than consumer web experience and this is reflected in the undelying IT achitectures.
Lastly, it is one thing to download a new piece of software – quite another thing to manage it over its lifecycle.
What Web 2.0 DOES provide is the building block approach to IT – that i think is going to be useful in taming IT complexity
Rajesh-I agree with you. It can lead to IT sprawl and get more complex. That being said, there is lots to learn for IT to understand how their users increasinlgy do not have to wait in line to get something done. And as I mention in my post, as the IT version of complex plumbing merges with the web culture of simple and lightweight, we will have some interesting opportunities. This will not happen overnight but centralized IT is starting to see what the next generation web can do for them. This process is just starting and would not happen were it not for the edge pushing the boundaries and actually being able to DIY.
As I said in the following post, I think that integration is the key factor – and by integration I mean the gamut of services, not just the raw ability to communicate.
I envision something like this:
“I need some IT services and… hey look! Startups X, Y, and Z provide what I need, I just have to configure them to get their inputs/give their outputs to each other the way I want!…But I don’t trust these guys to say alive, so – here’s company B that does backups, and can back up X, Y, and Z’s data in a way that will survive even if the companies don’t! Oh, but what about security? Built into the system! Just have to set up access rights. And what about integrating the whole thing into Corporate IT? Same as before! What worked for X, Y, Z, and B, will work for IT – just slightly – okay, greatly, more complex. …And, by the way, none of this involves programming!”