In a previous piece I wrote, I described a situation where a national restaurant chain wanted to increase sales and reduce costs by better matching its food inventory with local sales demand. I asked readers how they would use IT agility to enable the company to accomplish that goal.
Some readers were dubious about the wisdom of using agile IT methods to tackle such a complex business problem. With millions of dollars of inventory at stake and hundreds of variables and business rules to take into account, is it wise or responsible (even if it is possible) to act quickly and deliver only a partial solution?
I keep talking about IT agility because it works. There is real business value in quickly getting a system into production that works but covers only 80 percent of what is needed.
Here's what the team I assembled did for the client.
We used joint application design to pool ideas from business and technical people and agree on the functionality needed in the 80% system. Then we used process mapping to draw out existing workflows and design more efficient business processes.
Based on the process maps, we used data modeling to define the logical data model for the system database. Once we knew the new business processes and the data model, we where able to create a prototype of the user interface and the technical architecture for the system.
Inventory planners who would use the new system reviewed and tweaked the user interface design. Company IT people reviewed and tweaked the technical architecture so it conformed to company standards.
Then we applied object-oriented design and programming to configure the system from a set of reusable IT components: Web pages, a relational database, Web services for data transport, and spreadsheets and personal databases for data analysis and reporting. All these components are tied together with small chunks of program code that were installed and tested in a few weeks.
Version 1.0 of this system was in production within 30 days. We had decided on a 30-day schedule at the start of the project and then shaped the work to fit within this tight time frame. We couldn't do everything, so we focused on doing what would make the biggest impact in the time available.
The new system eliminates faxes and e-mails that inventory planners used to type into personal databases. It provides Web data entry so manufacturers and distributors in the supply chain can enter their own inventory numbers. Those numbers are then imported into a central database for the whole system. Inventory planners can download the data they need into their personal databases.
We provided a set of stored procedures to produce standard reports and graphs. Inventory planners can do their own ad hoc queries and load data into preformatted spreadsheets to conduct further analysis. They now spend most of their time doing high-value analysis and decision-making instead of data entry.
You may be thinking, "Well, gee, that's awfully simple." Yes, it is. That's why it's so effective. The system has been in use for several weeks now, and we are working on the next round of enhancements. We'll add automated data collection for those suppliers that want it. We'll provide data encryption and additional error checking. There will be enhanced performance-monitoring features too. The inventory planners have also put together a list of other features they want.
Again, we'll scope the work in this next round to fit into a tight time frame. The system practically grows before our eyes. And as it grows, it becomes more tailored to fit the exact needs of the people using it, even though it is being built from standard IT components.
This is agile IT. It delivers major business benefits at a fraction of the cost that would otherwise be incurred if the company were to go the conventional route of buying and installing a large, complex, all-in-one package.
Michael H. Hugos is a CIO at large and speaker. He is a member of the 2006 Premier 100 IT Leaders class. He can be reached at firstname.lastname@example.org.