- All requirements would be solidified during the first three weeks of the project. After that period, new requirements would necessitate a change order.
- The prospect would accept the out-of-the-box look and feel provided by the BI platform -- that is, we would not "skin" the application to dramatically change its appearance.
- The data sourced from the transaction systems had to be reasonably clean. A big data purification effort would result in a change order.
After a few rounds of negotiations, the prospect's COO decided that a fixed bid might, in fact, be too risky for his company and proposed instead a tightly controlled time and materials engagement with weekly status meetings and limits of 40 billable hours per week for consultants. This turned out to be a very prudent decision, saving the project from what we later learned would have been change order hell.
There's a saying in the BI world that the project doesn't start until the key user first sees a prototype with real data. That was certainly the case here. Not only did the company decide it wanted a snazzier, nonstandard look and feel that would have to be further specified and programmed, the prototype brought out many new business requirements and highlighted less than stellar data quality.
The key user? None other than the CEO. With a tight time and materials contract in place, we were able to continue without interruption as new needs surfaced rather than slowing down with an arduous -- and expensive -- change order process. The customer's consulting manager did a great job holding our feet to the fire throughout -- annoyingly so, I thought at the time. Overall, the engagement was extended a month, and the customer, though not thrilled with an additional 15 percent expenditure, was quite satisfied with the application. The customer also understood that what was delivered was a lot different than what was initially contracted. Both sides now enthusiastically reference each other as partners.
It's been our experience that fixed-bid contracts are generally not an effective safeguard against vague requirements and, thus, not an effective risk management device for customers. Companies who use fixed bid in this way usually don't get what they want and risk change order nightmares. Instead of a win-win, a lose-lose relationship ensues.
Unwise practice No. 2: Evaluating prospective consultants by their hourly rates
Rather than judging suitors on the total cost of consulting services to deliver a given basket of functionality, consumers are often boxed in by limits on the hourly rates of the consultants they engage. Low rates fly under the radar, while those above a threshold are often non grata, regardless of quality and productivity.
The fact that it might take a $50-per-hour consultant three times as long to complete a task as a $100-per-hour candidate is irrelevant with this thinking. Many times, the low-rate consultants end up collaborating for long periods with inefficient, combined customer/consultant project teams, resulting in a bigger management investment from customers.
What's needed is an evaluation process that measures the total cost of delivery, including opportunity costs, for specific functionality -- and a recognition that "burn rate" does not equal "value generation rate." Our advice to customers: Always look to minimize cost for demonstrated quality rather than risk an unknown quality for minimal cost. In short, understand your risk-reward equation and make a rational decision.