Product prioritization is simple, but not easy

At its heart, prioritizing product ideas is pretty simple — all you are doing is weighing up the benefits versus costs, which you can do in six straightforward steps:

1. Decide on 2–4 primary ‘Benefits’ that you want from each of your product initiatives. These should be driven from strategy, and might include things like Increase FY18 Revenue or Improve Customer Satisfaction.
2. Decide on 1–3 primary ‘Costs’ such as Time to Implement, Cost to Implement, and Risk.
3. Next you’ll most likely want to weigh your criteria, which allow you to place more importance on some Benefits and Costs. For example, if Improve Customer Satisfaction is twice as important as Increase FY18 Revenue, you’ll want to double the weighting of the satisfaction criteria. (At Trade Me, we use Product Plan to help with the scoring, but you can just as easily use a dirty old spreadsheet).
4. Do a preliminary scoring of each initiative against the Benefits and Costs using the scale of 1–5. (You can, of course, use any range you like, but I find 1–5 is about the right fidelity). These initial scores are your best guess and will change as you socialize your priorities. (Trust me.)
5. Calculate the totals (remember that costs are negative), and rank the initiatives from top to bottom. (BTW, you’ll very rarely be implementing them in the same order as the ranking. You might not have enough time before a brown-out to implement a feature. Or you may want to ensure you are balancing customer requests with addressing tech debt. Or a key person is away for 6 weeks.)
6. Now that you have a preliminary set of product priorities, you have something to talk about. Use this rough draft as a discussion tool with your stakeholders. Go to your key salespeople, marketers, developers, customer support, partners, customers, finance people and other product people, show them your preliminary priorities and ask for their help. How would THEY score some of these items? Which ones do they care about most? Which ones seem overrated to them? Go through them one at a time, as large groups will quickly devolve into arguments. Adjust the scores as you learn. Add additional Benefits or Cost criteria if you think you left something fundamental out. Keep socializing and tweaking until you have everyone nodding their heads or disagreeing on only a few key items.

See, easy! You can even plot it on a 4x4 quadrant:

It’s easy, but not simple.

There are two main issues I see with this approach.

Firstly, we don’t even do this much. Maybe we think a group exercise of dot voting on ideas is good enough. Or we feel like our strategy is set in stone and the features and improvements that come out of it aren’t changeable. This is lazy product management. We are oftentimes placing million dollar bets with these calls, and we need to have more rigour.

Secondly, we rely on pseudo-science. You can create a scoring guide to help you compare like-for-like. For example, saying that Increase FY18 revenue scores a 0 for $0, a 1 for $0-$100K, 2 for $101-$200K, etc. But even this is a guess. How do we know what revenue will come in? This is the hard bit.

But there are lots of different ways that we can try and more concretely and accurately quantify our scores. Use the Kano model or Buy a feature or Opportunity scoring to better understand what customers value. If you have a criteria relating to usage, then run some SQL queries to understand what features most people are using most of the time. If it’s adoption, then you might want to use choice modelling to understand what people would pay for the ‘parts’ of a product. The point is that there are a million different methods for removing subjectivity from the scoring. (Not that there isn’t a role for gut, but that’s a whole other post).

What do you think? How do you prioritize?