Montag, 26. März 2018

Measuring Strategic Speed with PDE

“The essence of strategy is choosing what not to do”, Michael Porter has famously instructed us. Thus, a company needs to invest resources wisely in those few initiatives that ensure the vitality of a “unique mix of value” in the marketplace./1/ Strategic success depends in part on the speed that development projects implement the new products or whatever the benefits might be. This post explores how we can measure the speed and efficiency of achieving benefits at the strategic, company level.  This article is the last in a series on “project duration efficiency”. You’ll find the first two articles here and here.

Project duration efficiency (PDE) is a measure of how well we use the calendar time available to us. At the project level, we want to deliver the product as early as possible, so as to reap the benefits as early as possible. Any waiting time can be counted as a cost, i.e. “cost of delay”, which is, simply put, equivalent to the lost income due to later product delivery.

In the previous articles in this series, I showed how we can measure PDE by comparing the effort with the calendar time available. An ideal value would be 1, with every day of calendar time being spent fully on the single project, thus ensuring that it is finished as early as possible.  Such an ideal value is next to impossible in the real world, with values of 0.1 to 0.25 being typical for classically organized companies and values above 0.5 being possible for Scrum teams.

As we have seen in the previous posts, we can measure this value for all of an organization’s projects, thereby creating a company score. Here is where things get interesting at a strategic level: what we are measuring is how well a company prioritizes its resources. A perfect prioritization would rank all projects by their strategic contribution, and the organization would complete the projects in one after the other. Our highest ranked project would get all project resources and be completed quickly, thus maximizing its strategic contribution (and generating a high PDE score). And, the scenario repeats for each next most important project. By contrast, if a company performs many projects at once, the teams need to juggle the parallel projects. As a result, projects end up stealing resources from each other, with *all* projects being delivered later than they might have been (and thus achieving a low company PDE)./2/

Perversely enough, many managers will read this second scenario as a positive sign of well-utilized resources, even if the company’s PDE score is low. Such a manager, who may likely be serving on the project steering committee, might bristle at a low efficiency score. After all, his team is handling three projects at once, and the time recording system shows >95% time spent constructively on projects.

The CEO of the company will, however, ask “But, where are my benefits? We’re trying to get the product to market. If I’ve done my math right, we could have finished this project months ago, and we've missed thousands of potential sales.”

The manager responds, “But, we also created several other products in that time.”

And, the CEO asks further, “So, does that mean we’ve missed out on months’ of revenue from the other projects, too?!”

Oops! How is that possible?  Actually, it’s not that bad, but we’ll still have considerable lost revenue. To simplify, we assume that we have four projects, each requiring 35 days duration in an ideal world. The teams are working on all four projects in parallel. For the sake of easy presentation, let’s assume that each week is dedicated to only one project. The scenario might look like this:

Thus, after 28 weeks or 140 workdays, we will have rolled out all four products to the market. We further assume that each product earns a similar revenue per week. If we sum the weeks where we earn money, we see that we get a modest 6 weeks. The cost of delay is a massive 106 weeks.
Now, let’s compare that to running the projects in sequence, which is again an idealized scenario:

Here, we have an optimal 42 weeks of revenue and 42 weeks of cost of delay. “Ach”, you say, “James, you’re just full of theoretical nonsense.  That will never work in practice.” Maybe not, but if we aspire to it, we might end up with a project portfolio that looks like this:

Now we have 30 weeks in revenue (5 times our original scenario) and 54 weeks of cost of delay (about half the original scenario).

As Donald Reinertsen has said, measuring cost of delay is the single most important measurement for product development.  Since strategy is a by-in-large a matter of bringing new products and services to the market, CoD is a vital strategic measurement.  PDE allows a simple means of showing across all initiatives, how well we are reducing our CoD as an organization.

We might thus revise Porter’s dictum: “The essence of strategy is choosing which projects to do later”.

/1/ Porter, M. E. 1980. Competitive Strategy: Techniques for Analyzing Industries and Competitors. The Free Press, p. 64.
/2/ Please see the previous posts for the full example. If Projects A and B run in parallel and rely on the same resources, then at least one of them will finish later. Compare the following two scenarios, which show the exact same total resource consumption.

Parallel projects, with 50% of the resources going to each project:

Serial projects, with 100% of the resources going to a single project at a time:

The effect on the benefits should be obvious.

/3/ Donald G Reinertsen, The Principles of Product Development Flow: Second Generation Lean Product Development (Redondo Beach, Calif.: Celeritas, 2009).

Keine Kommentare:

Kommentar posten