Aka “Secret Santa”.

The idea of “relative sizing” can be applied in a lot of situations. Most first learn of the concept in agile when we do estimates of the size of work. When agile Teams require an estimate of the work, they often use Story Points, a measure of relative size, as an estimate. We use Story Points because it is faster than estimation of absolute amounts. And the data that results is often more accurate as well. The idea that we can gauge something that is bigger or smaller than something we have, that we can do it quickly, and that we can turn these the result into numbers if required (through affinity mapping) is not limited to job size. We can apply the concept to estimates of relative business size, relative risk, relative impact, relative whatever.

Relative sizing also does not have to be done in a single dimension. For example, at the same time that we estimate relative risk, we can also estimate relative impact, if we set up both a horizontal and vertical axis. The technique can be applied as both a single and dual dimensions.

When this approach is used you will see:

1. It's a pretty fast process which means we have the data we need for decisions and so can be responsive to requests for information. For example, a portfolio level prioritization of business value and time criticality of more than 100 items where no-one had seen all the “epics” took less than 2 hours.
2. It gets the criteria out in public and so can be used to help defend prioritization decisions through a more non-biased thinking approach.
3. It helps break down the silos of prioritization - my stuff is the most important! - by having a more objective set of criteria and so helps with but-in
4. As you do future estimating you are can refer back to the criteria you have already reducing the need to revisit every single estimate when you do a new session
5. It helps people align on what each of the items under discussion mean thus improving clarity of the work and,
6. Most importantly, you get to alignment with the group of people of what is meant by “valuable” and what the most valuable items are.

For relatively comparing the size of a single axis (for example, effort, value, criticality, risk, impact, opportunity, etc.):

• Start with a list of items written on individual sticky notes
• On a wall place a large sticky on the left that say “Low” and one on the right saying “High” (or whatever the “scale” is that you want to use)
• Participants form a line
• First person in line takes the first sticky note, reads out the item, and places it on the wall between Low & High based on his or her judgment, returns to the end of the line, discussing the thinking as they do this.
• Next player can choose to:
  • Play by picking up the next note & placing it on the wall relative the positions of other notes on the wall, discussing why their thinking as they go
  • Pass placing a note on the wall, or
  • Move one existing note on the wall to another location, again discussing why this item needs to be moved
• Once the group places all the notes on the wall, they’ll continues to circulate through the line with Pass or Move options only
• Once everyone in the line has passed on adjusting the items the first part of the activity is over.

Often you want to assign numbers to the low-ness and high-ness of the items. To do this simply look at the groupings of items you have. We typically use a modified fibonacci scale for number assignment - 1, 2, 3, 5, 8, 13, 20. Got to the lowest group, and magically assign the number “1” to all those items in the group. Similarly with the upper most group. Magically give it the number “20”. Then make decisions for in-between groups for the rest of the number sequence. Collect up the results, for example, by marking each sticky with the appropriate number.

If you want to do two variables, just create a horizontal and vertical axis. People place items of the wall commenting on both of these axis.

Here is an example laying out probability of risk and impact:

The above assumes we have some understanding of the value we are comparing and also that there are often many different perspectives on value. To me when you do “play, pass, move” type approach you have to establish alignment on what value means to this group of people. For example value might be “revenue” or it might be “cost savings” or it might be a relationship between those two things. I typically start by holding up two items and asking “which of these is more valuable” and people agree “this one”. I then ask “why is this more valuable” and capture the reasoning as a series of “Value characteristics”. I might repeat this a couple more times.

Then, as they talk about positioning the rest of the items out there (by doing “play, pass, move”) you will sometimes hear something that sounds like an additional characteristic of value. Stop the estimating, discuss whether it means there is additional characteristics, capture if required, and continue with “play, pass, move”.

Some ideas:

• Have some fun with this - treat it as a conga line
• If you have no numbers that are commonly known by the group, don't use numbers on the axis. Low / high, big / small etc are better labels and helps people focus relative sizing as opposed to trying to guess an absolute number.
• You can create values between low and high using numbers in the Fibonacci sequence: 1,2,3,5,8,13, etc. This is done by affinity mapping clusters of items.
• You can use circles to map out, for example, high risk or impact, and so show where you should focus first (see picture above)

• What is the Basis of Our Estimating Process? for an understanding of relative sizing