Prioritizing a backlog with multi-attribute utility analysis (MAUA)

One of the most important things a product owner should do is to prioritize. What should we do now? Where should we put all our efforts, money and brains?. Inside a product team there are internal and external inputs that vary from opportunities, experiments and outcomes to risks, debts or defects. While different sub-teams can come up with a number of features or stories that come out of specific artefacts like dependency graphs, user story maps or metric optimization items, with enough amount of inputs the putting-all-together backlog can be hard to prioritize. If I have framed your situation please keep reading.

I learnt from Joseph Pelrine this technique based in estimating the impact that a particular action has in a specific set of objective prioritisation criteria. For the impatient here you can see an example spreadsheet

But before, how did you come to have such a backlog?

Before getting into the whole point of it a few words about rushing hard into having too much backlog items prematurely. If you have a number of stakeholders completely independent with conflicting goals then you will likely end up with this kind of backlog. This could be the case for a huge B2B customer base with some big fishes. But if you are launching a new product or a spin off it is worth a minute to think about the following questions

Are you writing shopping lists or do you copy the best of every competitor? How well do you know the customers whose needs you try to solve? How much have you focused on the outcomes? Any analysis in the user trips or funnels? Which personas do are you trying to satisfy immediately? Think about those or you could end up in a lot of product to build that does not apply to your domain or is out of control

Did you think what you want to impact from your own perspective? Is it McClure’s AARRRR metrics or lean’s Empathy, Stickiness, Virality, Revenue, and Scale growth stages that you are looking to improve? Doing too much experimental stuff without considering how to measure impact and gain validated learning? No “line in the sand” ? You already now that it is all about goals

Start with goals and their relative importance

The first step is to think about a short number of objective prioritisation criteria for your product / company, could be classic business drivers or something less abstract

Sometimes when companies grow they have to face challenges as the ones expressed in the following example

  • 1 – learn about the users
  • 2 – reduce our operations time
  • 3 – satisfy key accounts

Part of the problem with human minds is that we tend to count and divide, so if you have 3 prioritisation criteria it is likely that you think that they are 1/3 as important but trust me, it is never the case. So the next step is to put into perspective the relative relevance of your prioritisation criteria. Just do it by giving them points.

  • 1 – learn about the users – 8.5
  • 2 – reduce our operations time – 6
  • 3 – satisfy key accounts – 4

This is a classic example of comparison weighting: if criteria 1 has X points, how many points does criteria 2 have? and 3?

In general percentages work very well as metrics, so let’s also get the percentages or relevance

  • 1 – learn about the users – 45,95%
  • 2 – reduce our operations time – 32,43%
  • 3 – satisfy key accounts – 21,62%

How likely will the items you need to prioritize make an impact into these prioritisation criteria

Now it is time to check how will your decisions impact on your objective prioritisation criteria, without neglecting the probability, because we can sometimes just affect, not exercise complete control. This is how the table looks with the the probability of impacting in the prioritisation criteria in every cell.

user story 10,00,90,1
user story 20,80,20,7
user story 30,80,50,4
user story 40,60,10,5
user story 50,50,60,2
user story 60,30,50,4
user story 70,10,10,3
user story 80,00,70,1
user story 90,20,80,7
user story 100,10,10,3
user story 110,30,90,4
user story 120,10,10,2
user story 130,30,10,7
user story 140,80,20,8
user story 150,40,40,7
user story 160,00,80,0

It is worth another second to double-think: How do you know that prio? Did you disregard other things that could impact the criteria? Do you have a lot of things that have nothing to do with them? How well is everyone aligned here? because it could be that you have false prioritisation criteria

Finally: do the maths and order

Good news is that these maths are simple. Multiplying the probability and the relevant percentage for every prioritisation criteria and calculating the sum you will get a overall weight for the prioritizable item.

Order your items by the weight and done! this is how your prio table looks like!

weightprio order
user story 30,80,50,461,61
user story 140,80,20,860,52
user story 20,80,20,758,43
user story 110,30,90,451,64
user story 90,20,80,750,35
user story 50,50,60,246,86
user story 150,40,40,746,57
user story 40,60,10,541,68
user story 60,30,50,438,69
user story 130,30,10,732,210
user story 10,00,90,131,411
user story 160,00,80,025,912
user story 80,00,70,124,913
user story 70,10,10,314,314
user story 100,10,10,314,314
user story 120,10,10,212,216

You can find a sample spreadsheet available to download in ods format or
in excel format or grab the available google doc

 

Bytefilia