Social Glass

In a previous post I described an Enterprise 2.0 implementation roadmap. The last component in the roadmap is Enterprise Mashups. In a utopian world, enterprise mashups give the knowledge worker spreadsheet-like flexibility to dynamically create composite applications that suit his specific business needs. Companies like Kapow, Serena Software and IBM are building enterprise mashup engines, but none have reached this utopian state.

Dion Hinchliffe wrote a great post about challenges facing enterprise mashups. He identifies 10:

1. No Commonly Accepted Assembly Model
2. An Immature Services Landscape
3. The Splintering of Widgets
4. Management and Support of End-User Mashup Apps
5. Deep Support for Security and Identity
6. Data Quality and Accuracy
7. Version Management
8. Awareness and Realization of the Potential of Mashups by the Business Community
9. Low Levels of Support by Major Software Firms
10. Few Killer Demo Apps

In my professional life I’ve encountered obstacles around items 2 and 4 when discussing enterprise mashups with potential clients. We can show impressive demos, like IBM’s QEDWiki, but we’d need a seriously mature SOA to build the “widgets” a knowledge worker would use to create such a mashup (issue 2). And Dion asks who’s going to support these things once they break? “The IT department? The business units? Using what tools” (issue 4). Good question.

At the end of the day an enterprise must have a strong SOA initiative and support from IT before the value proposition behind enterprise mashups can be fully realized.