Search This Blog

Tuesday, April 30, 2019

Agile Iterations also involve cost

I was recently facilitating a class on digital project management and bringing references to both agile and traditional frameworks. While the concepts of fixed scope within the time-boxing principles was readily understood, many people in the digital media class mentioned that agile projects don't involve cost or have principles of cost baseline! The students didn't resonate with risk management also in agile iterations. These misrepresentations of agile framework continue to surge and are not accurate representations of agile approaches to project management or product development.

For instance, consider that an agile team is made up of 9 team members. The product owner, agile coach or scrum master, and the development team spend time and are compensated by their performing organization. So, when these 9 members spend approximately 6 hours per day (at 75% commitment to the project), they spend 54 hours (9 members x 6 hours) per day. Considering a 10-day iteration, the team has spent 540 hours (54 x 10 days)  with one iteration. By allowing this iteration to continue with this agile approach to meet the strategic objective, the performing organization has, therefore, spent 540 hours; this time represents the opportunity cost of the organization choosing to allow these resources to work on this initiative compared to another initiative. So, how could an agile iteration not cost anything to the organization?

Let us drill a little more here. Often, people are unable to come up with the cost of an iteration. Now, this may be associated with the fact that agile doesn't favor big upfront planning. In traditional approaches to project management, we come with the WBS with the breakdown of activities worked on by specific resources and associate a resource-level pricing. It is therefore possible to come up with a cost. The similar logic can still be done in an iteration easier because agile is all about team level commitments. Instead of looking at individual resource level pricing (such as what's the hourly rate of product owner?), the agile team can work with the finance department to come up with a blended resource-loaded rate. The financial units within many organizations have such a blended rate and I have received immediate responses to my requests in my experience from the financial department for such a blended rate. Now, assuming the blended rate is $100, one can easily apply this blended rate to the 540 hours in our previous example and come up with a cost of $54,000 (540 hours x $100). So, the cost of an iteration can be found out easily.

Using other heuristics or analogous experience of delivering multiple iterations, one can also come up with a cost of a story point. Now, it must be mentioned that a few iterations should be done before we can come up with a reasonable and consistent velocity as the team matures. Assuming a median velocity of three iterations, if we hypothesize that the team has to deliver approximately 150 story points in every iteration, then, the baseline cost of $54,000 can be divided by this hypothesized 150 story points to come up with $360 ($54,000/150). When the team misses out on completing a 3-point story because of not proactively identifying and addressing the risks, managing stakeholder communication, or promoting the daily collaboration between business and technical users, then, the missed velocity in that iteration costs $1,080 ($360 x 3 story points) of non-delivered value.

By monitoring velocity delivered per iteration against this baseline projection, one can evaluate the CPI (Cost Performance Index). Again, extending the example above, the baseline projected velocity is 150 story points (with $54,000 or actual cost). When the team didn't deliver 3 user stories, the value delivered (Earned Value) is 147 story points (147 story points x $360 cost per point = $52,920). Since CPI is computed as a ratio of Earned Value/Actual Cost ($52,920/$54,000 = 0.98). This means the team is not delivering 2% of the committed value. Agile teams compute this as the Burn Rate, which is the reciprocal of CPI (1/0.98 = approximately 1.02). This burn rate represents how the agile team is meeting the projected budget and since it is >1, it indicates the team requires 2% more budget at a minimum for this iteration. Since agile uses time-boxing principles, as each iteration fails to deliver the minimum required velocity, each iteration costs more money towards the end in meeting the minimum viable product (MVP).

Hope these points are clear in explaining why and how agile projects do not exclude cost. It is critical to understand these concepts and subsequently have a plan to quantify value delivery using good return on investment principles.

Thoughts? Please share your comments.

1 comment: