Friday, I was visiting an enterprise client to review their architectural plan and discuss a business architecture initiative. During our conversation, we touched on the need for business users to simulate a proposed business change, without a big IT effort to provision the information and provide a front-end. In this scenario, a flexible front-end is crucial, because the business users are really experimenting with different models.
This got us to talking about Excel, and as luck would have it, I had just recorded a related podcast for Anne’s excellent Office 2.0 Podcast Jam. The client asked for a transcript of the podcast, and it occurred to me I should probably post it. Here it is:
For my contribution to the Office 2.0 Podcast Jam, I want share a simple idea that mixes “2.0” concepts with service-oriented architecture to solve a common and often contentious IT problem, that of end-user, excel based, computing.
First, a little about the problem. As we all know, one of the most common tools used inside the enterprise for planning, analysis and information exchange is Excel.
Spreadsheets that start out as personal tools, get passed around, augmented, and quickly evolve into important applications containing business critical data.
During this evolution, local databases begin to emerge, as do requests for extracts and updates to enterprise information sources.
As the spreadsheet evolves from a simple tool to a vital business application, a variety of risks emerge. Some examples:
• Data isn’t consistent with enterprise systems.
• There are compliance exposures due to the lack of security and audit.
• In some cases, the local infrastructure is outgrown, or expensive to manage.
Because of these risks, sooner or later, every IT manager is faced with a difficult decision to take-over or shutdown an end-user evolved application.
In a “take-over” (either re-platformed or IT supported) the application gains stability, but the business user, the application’s creator, loses control. The business user is no longer free to add new information sources, manipulate the presentation, or vary calculations. The creator’s freedom tends to vanish as important operational controls are added.
What’s needed, is a solution that promotes “do-it-yourself” while still implementing necessary controls.
“2.0 Constructs and Concepts”
Common across “2.0” technologies – are the concepts of networked access, everything (software, information, computing) as a service, collaboration, ease-of-use and the all important ability to compose (mix and match or mashup).
The two keys to composition are having resources to compose and tools to create the composition. Regarding composition, Andrew McAfee, in his insightful MIT Sloan Article: Enterprise 2.0: The Dawn of Emergent Collaboration, noted two enterprise 2.0 ground rules:
1. “Offerings should be easy to use”.
2. “Don’t impose preconceived notions about How work should be done… or How output should be categorized or structured. Build tools that let these aspects of knowledge work emerge.”
The second rule “build tools that lets the aspects of knowledge work emerge” is critical to the idea I want to share.
So here it is. I’m suggesting that IT provides a “2.0” based toolkit that gives business professionals using Excel easy access to information through data-oriented services from a service-oriented architecture.
The toolkit should be an Excel plug-in that allows business users to search for services, and then drag and drop relevant data fields into their Excel worksheet. The toolkit should be open enough to access services from external sources.
The services, besides delivering the data, should have enterprise controls for user access, auditing, sensitive data handling, and in the case of an update, data quality checks.
This combination gives business users the power of do-it-yourself, in a tool they’ve mastered, without risking data integrity, breaking infrastructure or violating compliance rules.
As added benefits, the business users are now working with the freshest enterprise information, and IT has a quick, visible win, in demonstrating the power of a service-oriented architecture.
For an example of an easy to use Excel Web Services plug-in, check out StrikeIron’s OnDemand for Excel. You might start with that, or create your own.
Whatever you do, don’t offer a programmer’s toolkit to your business users. Make it easy, and make it successful.
Anyway, that’s my, as promised, simple idea. I realize it isn’t exactly “Office 2.0” since I’m recommending a solution based on desktop resident Excel, but I think it is a good mix of “2.0” and SOA that offers business value today.
[Disclosure: StrikeIron is NOT a client of Elemental Links, Inc.]