The biggest question to ask yourself when you’re deciding on a model is “What are the consequences if we’re wrong?” The lower the stakes, the less work you put into scoring. The higher the stakes, the more work you need to do. But, there’s should always be a healthy tension between effort vs reward so that you don’t spend unnecessary time scoring.

Sidebar: There are other related concepts to consider that may unknowingly make you spend more time scoring than you could get away with otherwise. The biggest one is not breaking down work into small enough pieces of value. Often times a large piece of work contains many smaller items of variable value, robbing you of the ability to prioritize them separately. Also, having lots of big work means work takes longer and when work takes longer, there’s more escalation of things in the wings – that escalation causes more overhead. Fixing the underlying cause of “too big work” can reduce unnecessary prescriptions of extra scoring overhead. So, think about any other aspects that are causing a need to score to see if there are any that can be fixed closer to the source. There are so many aspects that intertwine into any other aspect in complex systems and its fun to do a little sleuthing.

This PDF, by Troy Magennis at FocusedObjective, guides you to different methods via answering questions like how big are the consequences.

What makes up value?

The first question to answer is “what makes up value?” You can order via any method but first, you have to understand how you’d make the ordering decision in the first place. There are common considerations on both the positive and negative side:

plus Revenue protection or generation

This is by far the best thing to learn how to consider. It takes practice and a lot of assumptions, but, over time you hone your assumptions and begin to get much better at forecasting actual economic value. Normally you think ROI (return on investment), but that’s just one part of the picture. ROI is a snapshot in time, but you may not make the same ROI if you launched something AFTER a competitor than if you had launched it before. This is why Cost of Delay is the most valuable economic method. It requires that you understand ROI but you anticipate how any delay will impact that ROI. I definitely suggest that this be part of your scoring.

plus Cost protection or reduction

This is similar to revenue in that we’re dealing with dollars but this time we’re keeping our costs low or reducing them.

plus Capabilities

Will delivering this feature give you any capability that you didn’t possess before: allow you to deliver new things, allow you to deliver things faster or with higher quality. This is where “technical debt” can become valued highly.

minus Total Operating Cost

The sunk cost (actual $ spent, payroll, opportunity cost, etc.) to deliver the work should be part of the value discussion. Don’t get into a situation where you build something that you’ll make $50K on, but it costs you $60K to deliver. When considering cost, don’t just think about up-front cost, think about the total operating cost.

plus minus Risk

This one goes both ways – it can be a net positive or a net negative. Does doing this increase or reduce risk? How does that risk manifest? Is it in reduced revenue/market share? Is it in capabilities? You may incorporate this into the above categories instead of calling it out on its own, but definitely proactively consider risk.

The simplest approach to prioritizing based on value.

You can use what’s called “reference class forecasting” to map items based on relative value. Take one item that has an average value and set it down. Then take another item and assess if it has more or less value than the previous item. Repeat over and over with new items and you get a stack-ranked ordering by value.

The mantra goes, use the simplest thing that works. If this will work for you, use this. If it won’t then use something more complicated.

Cost of Delay (CoD). It combines urgency and value. This is where the activities we did in our workshop come into play.

There’s an easy level of CoD called Qualitative Cost of Delay that you can start with. It results in “buckets” of items with similar values –

There’s a more valuable level of CoD called Quantitative Cost of Delay. Here you get right to using dollar figures. It doesn’t have to be as hard as we think because it doesn’t need to be as precise as we think. Some info is better than the zero info we have now.

There’s a chance you should forget everything I wrote above…

We aren’t always great at estimating – value or otherwise. That’s why those methods are helpful. It is very possible to spend a lot of time calculating value, perfecting the process to do so, and, yet, still end up getting it wrong and making bad choices.

Another option is just don’t. If you don’t absolutely have to prioritize, don’t. Sound crazy? Well, prioritization becomes more and more important the longer it takes us to deliver value. In this way, it is failure demand. If we become able to deliver value quickly, then the need to prioritize might significantly lessen or disappear altogether. If you can get to the point where people don’t have to wait long for anything then it can be a better choice to just do the work instead of taking time to prioritize… do items first in, first out.

Unrealistic? Maybe, maybe not. At the very least, we can consider it a perfection challenge. We can continually improve in the hopes that one day we might reach this land of nirvana. Until then, I hope the methods linked to in this post provide some help!