# What and how many discrete priority values should I use to describe importance of announcement inside web application

I develop CRM for my company. I created module to post announcements for rest of the crew. I would like to create scale of importance to choose from like 4 values: 1,3,5,7 or descriptive values.

Question might seem stupid but for instance on pivotal tracker there were no priority values like 1,2,3,4,5 but like 1, 4, 6, 10 etc. so I think a psychologist or someone had a deeper thought on this one. Should I associate those values with colors(seems obvious)?

So the question: What is the best way to describe level of urgency/importance of announcement/task inside application?

• Your question is confusing, can you flesh out with some visuals on what exactly you are asking? Commented Sep 17, 2014 at 19:56
• If I understand right, you're asking about how using normal numbers (1,2,3,4,5) vs other systems like odd numbers (1,3,5,7), Fibonacci numbers (1,2,3,5,8), or powers of two (1,2,4,8,16) impact how accurately people can predict task complexity (in Agile story point estimation). Is that correct? Commented Sep 17, 2014 at 20:23
• @3nafish Yes, that's corret.
– Yoda
Commented Sep 17, 2014 at 23:48

# Just don't do it

Priority rating is one of the most misused practices about.

# Absolute vs relative

The height of something can be expressed in absolute terms (1.6 metres), so is the volume of a cube.

If you really think about it, priority is a relative intensity measure.

The priority of a single item, if there will never be other items in its group, is nonsense.

With two items, one is more important than the other.

With more than two items, it is easier to sort them than assign a fixed number to each - if you have 2 items with priority 4, which should you do first? Or what if you have an item marked 3 and an item marked 4 and a new item you add is clearly between the two (3.5)?

# The (lack of) meaning of numbers and words

Scale numbers mean very little to people and what each number represents is largely subjective (if you've done a fair share of UX, you must have noticed that sometimes two users or experts, despite seemingly having the same impression, will tick scales rather differently). Even when using words (priority: low, medium, high) these words have no absolute or global meaning.

I've spent dozens of nights doing UX research on task and product management, and it was not until I've reached the realisation above that things started falling into place (and yes, I know nearly all task management applications have a priority, importance, or urgency scales).

If you need a scientific evidence for it, just search for issues with scales in quantitative research. Funnily enough, one remedy (if one wish to stick to scales) is to introduce relativity - a question that is simply either an anchor, used as a factor, or has some more global acceptance ('how would you rate the Gmail interface', then 'how would you rate our interface' rather than just the latter).

# The solution

So to conclude, the best way deal with urgency/importance is not to let users assign a 'level' to items, but rather allow them to sort the items. By this, you free the brain from having to assign an illusive meaning to a number, and leave it with the fairly simple task of 'any-two comparison' - is A more important than B. Much easier to do!