One of the main things I hear when I am asked to assist an organization improving their portfolio operations is that everyone thinks the solution to their “problem” is to get more people. The feeling is that there is simply too much work that needs to be done. Organizations in this situation will decide that “something needs to be done” and so start looking to quantify how many more people would be needed to get all the currently approved work done. Organizations ask each project something like “given target completion date is X, and based on progress so far, how many more people do you need to hit the date?” Since so many projects are in trouble with will result in a “need” for 30% more people which is probably not possible given the current budget.

The reality is that given that, even if there was budget, onboarding is a lengthy process, we really need to understand whether we are getting the most effective outcome given the capacity we have. Further we need to step back from the problem and ask ourselves how we keep on getting into this situation? Generally, I have found that organizations are not making the best use of capacity for two reasons:

• Organizations approve too much portfolio level work which results in missing commitments, slow delivery, and a lack of focus on the value (outcome) required by the portfolio.
• Organizations do not focus on what it takes to deliver value, focusing instead on managing people rather than the system of delivery.

A couple of simple exercises can be used to provide visibility (data) into these areas. Firstly, we need to understand the demand we have on the system and compare it to the capacity available to deliver on that demand. The approach I’ve taken in the past is:

• To determine demand because of portfolio initiatives:
  • Generate a list of the current portfolio initiatives.
  • Have the senior technical or architects people size the remaining work of each initiative (ranges of person-weeks)
• To determine capacity available to work portfolio initiatives:
  • Determine number of people we have available to do work.
  • Determine how much of a person’s time is spent on initiatives in general. This is done by, for example, taking the total available capacity of a person for an iteration, subtracting off meetings, business-as-usual, production support etc. You will probably end up with 20-40% of a person’s time available for work on initiatives (note: this is often a surprise to all)
• Demand vs Capacity:
  • Compare the demand to capacity within some timeframe e.g. one-year look ahead. This will tell you roughly whether you have too much work “approved”. Gross numbers will show you how “bad” things are.

In many cases this will lead to a healthy discussion about how we focus our efforts on a few of the most important initiatives as well as a discussion of how much work we should kick-off at the portfolio level. I’ve seen situations where there is 5 years of work that has already been approved and there is still more work being approved! The resultant discussion typically results in a pure prioritized top-10 list to start to clear the backlog and rules for any subsequent approvals (e.g. cannot approve new initiative until another item is finished.)

A second aspect of this kind of analysis is to look at the rate initiatives are being finished. We can easily determine the throughput of Initiatives, which will provide further data on portfolio completion. To do this count the number of initiatives that have finished in a time period and apply that to the future. So, if we have 120 initiatives to complete, and we are completing 7 initiatives per month, then we can expect out backlog to be cleared (without any new initiatives being approved) in 17 months.

One place I worked showed that there was, according to the demand vs capacity discussion above, enough people to do the work but that based on people’s gut feel this was not correct. This analysis revealed there was a backlog of about 2.4 years. This lead the VP responsible for portfolio management to day:

So what you are telling us Hans is that we just suck at execution!

Additional analysis lead to a discussion about whether we were effectively using the people we had. Issues in this organization included:

• Alignment: There is a lack of alignment with respect to what is the most important thing to be working on. This results arbitrary and non-aligned decisions at all levels of the organization, slowing down everything.
• Flow Visualization: There is no common view of the how we deliver value through our system.
• Interruptions: There are a lot of interruptions. Interruptions come from both changing perception of priorities as well as “urgent” issues such as production support.
• Inflight Work: Related to the above, there is a lot of inflight work. This inflight work is another source of interruptions as people try to respond to current demands for progress.
• Dependencies: There are a lot of dependencies between organizations to deliver value, not just between teams implementing the value, but also support teams such as operations, procurement, security, HR, and so on.
• Reactive (rather than proactive): Related to the above, there was an inability to plan for items with significant lead times or resource constraints.
• People Churn: There is a lot of people churn in the organization, with many new people being brought on board both as associates and contractors, and many leaving. This typically results in impact on delivery of value as people need to be brought up to pace on the work (domain knowledge, technology knowledge, etc.).

Both tools - demand vs capacity and initiative throughput - are simple, lightweight tools to work the perceived need for “more people”. In most cases the “demand” side of the equation can be developed in a couple of 1-hour sessions, while the “capacity” side of the equation is the result of a series of questions to provide a useful range of figures. The throughput discussion is typically available by watching current initiatives generally generate from historical data. And while simple the data produced is hard to argue with. In all cases the data generated typically opens conversations and moves the discussion away from “how do we get more people” to “how do we ensure we are working on the most important things given the capacity we have”.  

• See Facilitation - Play, Pass, or Move for one approach to estimate the demand.
• Short version of this was published on LinkedIn and the FiveWhyz Blog