And How You Can Avoid This “Mistake”
I take pride in delivering robust data models and efficient reports that meet or exceed a client’s expectations by delivering a finished product early or providing more functionality than originally requested. I would never have guessed that delivering a product that fulfills 100% of the specification would fall into the failure category. Let me explain.
A building materials company commissioned the development of a complex solution that analyzes historical purchase order information and determine the average spend for a given material and plant. The value would be the basis for comparisons with forecast data to negotiate better prices.
The solution would also save the company 60 man-hours each month and allowed the business to analyze a large number of products quickly.
The project scope was large and encompassed 20+ source sytem tables and 15+ BW infoproviders. The business logic that connected the various components were complex and the project duration was 6 months.
While wrapping up development and quality assurance testing early in 5th month, my business counterpart and I received unexpected news. The internal departments were realigned and another team would be the consumer of our report. Naturally, they asked for changes in the requirements.
“You delivered exactly what was asked for. But we need something else.”
Not only did we get a change 6 weeks before the scheduled go-live date, the scope of the request was massive. The new approach was 180 degrees in the opposite direction. The average spend, a key component from which all calculations were to be derived, should be calculated from the General Ledger postings along with any variances that may impact the final price.
Somehow, we managed to salvage the existing work and turn in an acceptable solution in 8 weeks. The work was frantic and I vowed to never repeat this “mistake”.
The lesson?
- The business landscape can change quickly.
- We need tools to deliver even quicker.
Since a pursuit of efficiency and optimization is a switch that I find hard to turn off, I’ve been actively reviewing new tools to reduce the time and effort to quickly deliver a project in 8-16 weeks.
More to come in a future post.