Most agile teams use a form of relative estimation to size their user stories.  The term relative estimation comes from the root word relate.  This means that, when teams decide how much effort a given user story requires, they are not really thinking in terms of hours or days but rather about the effort involved in relation to other stories they have done in the past, or maybe even stories they are considering right now.

There are several relative estimation scales that teams use and none is better than another.  It comes down to choosing a scale that makes sense to the team.  Probably the most popular scale is t-shirt sizes: XS, S, M, L, XL.  Other teams prefer numeric scales.  The two most popular are the Fibonacci sequence and powers of two.

If you don’t remember the Fibonacci sequence from math class, don’t worry.  It’s a very simple number pattern.  The series starts with zero, then one, one again and then every following number is calculated by adding together the two previous numbers.  When used for estimating, teams drop the zero and first one, so the scale looks like so: 1, 2, 3, 5, 8, 13…” and so on.  Powers of two might be easier to remember: 1, 2, 4, 8, 16, 32…” and so forth.

Other teams don’t opt for any of the above and invent their own scale.  I have a client that uses mattress sizes to estimate stories.  All stories are either twin, full, queen, king or California king.  Perfectly fine.  It makes sense to them, and it also makes it fun. 

But when teams use relative estimation with numeric scales, a little extra caution is required.  This is because the number series that those scales are based on go on forever.  For example, powers of two looks like this: 1, 2, 4, 8, 16, 32, 64, 128, 256…  Don’t give your team all those choices because it will push them into choice overload.

A close relative of decision fatigue (making too many decisions leads to overwhelm and therefore poor decisions) choice overload occurs when we have too many choices from which to pick.  We can become paralyzed and overwhelmed.

Here is an example of choice overload from everyday life: the cereal aisle in the grocery store.  Let’s say you’re at the store and a member of your family has asked you to get Cheerios.  No problem, right?  But then you get to the Cheerios section and see…there are regular Cheerios, Honey Nut Cheerios, Cheerios Oat Crunch, Cheerios MultiGrain, Cheerios Protein (and that one comes in multiple flavors, compounding the problem.)  You try to call your family member to see which one they want, they aren’t picking up their phone…and you finally say to yourself with an exhausted sigh “…we’re not getting Cheerios today.”

That is choice overload.

So, whatever relative estimation scale the team uses, encourage them to limit their size choices to no more than five (ex: 1, 2, 3, 5, 8) and recognize that, for many teams and especially those new to relative estimation, three options will be a better choice.  For beginner teams, an ideal scale is t-shirt sizes but fewer: small, medium, large. 

That’s it. 

Limiting size choices when doing estimation will help your team avoid the paralysis of choice overload and achieve the goal of using a relative estimation scale: to be able to estimate quickly and intuitively.