Prioritizing Product Backlogs

Posted on Categories Agile Process, Product
Matrix code
Share if you found this interesting ...Tweet about this on TwitterShare on FacebookShare on LinkedInShare on Google+Digg thisShare on Reddit

Matrix code

While there are many ways to prioritize a backlog, precious few manage to remain as objective as possible and many succumb to the whims of the team, the founder, the weather even! What better way to remain objective than quantifying business value through the eyes of your prospective customers?

I’ve previously talked about why quantifying business value might be your path to salvation so here I’ll just focus on one detailed example of how to do it. To keep it brief I’m assuming prior knowledge of Lean Startup and I’m assuming the scenario is one of trying to score customer feedback that you’ve diligently gathered with solution interviews while trying to conjure Silicon Valley’s next unicorn.

If you don’t know much about Lean Startup you’d only benefit from knowing more, start by consulting the book that started it all for a nice overview: Lean Startup by Eric Ries

It’s an easy read, but if it leaves you asking “Great! But how the hell to I do some of this in practise?!?” and if you actually plan to attempt conducting problem/solution interviews in reality you’ll want to move on to the de facto practical guide for this stuff: Running Lean by Ash Maurya

Once you’ve done your fair share of interviews and started tweaking your process, you’ll most likely want to know more about metrics / analytics to make sense of the data you are collecting. When I was digging for more specific examples I found Lean Analytics by Alistair Croll and Benjamin Yoskovitz to be quite helpful, but the authors’ blog posts were even more helpful, specifically here and also here which inspired what I’ll lay out below.

So with the context out of the way and assumptions all articulated I’ll dive in head first!

You have been, or will be, interviewing customers, you need to devise a way of assigning a weight to their overall opinion. You can make it as simple as arbitrarily assigning a value next to each customer’s name, but that is of course dangerous, rather decide on a set of criteria that will be used to measure the value of a particular customer’s opinion:

Customer name Fits our intended target market profile? (+2) Is actually paying us? (+4) Shouts loudly? (+5) Total opinion weight
Good fit cheapo’s Inc. Yes No Yes 7
Pushy misfits R’ Us No Yes Yes 9
Fence sitter Corp. Yes No No 2

As you can see the criteria need not be politically correct or in formal business language, as long as they are clear enough to discuss with your stakeholders. Next you need to express how much each customer truly needed each feature that makes up your solution(s), once again you can just assign a value based on your gut from the interviews, but defining a set of criteria that drives how you measure it is a more objective way to do it:

Reaction to solution Reaction weight
Hated it -1
Unknown 0
Paid attention 1
Foamed at the mouth 2
Customer / Feature name Payment gateway Makes Coffee Exports to PDF Platform independence
Good fit cheapo’s Inc. Foamed at the mouth Paid attention Unknown Hated it
Pushy misfits R’ Us Paid attention Hated it Foamed at the mouth Unknown
Fence sitter Corp. Paid attention Unknown Hated it Foamed at the mouth

Now you have the data needed to answer the question that often seems impossible to answer reliably: which feature(s) will benefit the majority / most valuable of my customers the most?

For each feature, multiply a customer’s score for the feature by that customer’s overall opinion weight.  Add up all the customer scores for a feature and voila! you have a list of all features that you can sort by their overall scores:

Feature name Feature Score
Payment gateway 25
Exports to PDF 16
Makes Coffee -2
Platform independence -3

Naturally you should not be a slave to this powerful tool either, if you start to doubt the criteria you’ve set is working well for you, change them!

If you know of prior commitments that have to be met or technical feasibility issues that would favour a feature with a somewhat lower score than those above it, prioritise it above them if you need to!

This is simply one lens to use, it is by no means the only one, where it really comes into its own is to counteract emotions, prejudices and preferences you or your other founders may have developed subconsciously over time, it keeps everyone pretty honest and helps you make decisions with better data.

You’d have to be some kind of masochist not to see the value in that!

Share if you found this interesting ...Tweet about this on TwitterShare on FacebookShare on LinkedInShare on Google+Digg thisShare on Reddit

Leave a Reply

Your email address will not be published. Required fields are marked *