
Our team has been working on prioritization quite a bit lately and, perhaps by no coincidence I’ve seen a lot of great commentary on priorities. In my experience, there are at least 3 challenges with prioritization. And while there are no perfect ways to combat these challenges, by understanding them and keeping them in mind you should be able to make better decisions.
- No Clear Objective – This is possibly the biggest culprit of poor ability to prioritize features and build a great roadmap. And it’s generally not directly the fault of the product managers but comes from a lack of leadership – often starting all the way at the top. The leadership’s responsibility is to create an objective or said differently establish a problem to be solved and then come up with ways to measure whether that’s happening or not (and ideally, empower a team to then go after this objective). But absent that, a product manager is like a chef with a ton of ingredients and no idea what they’re trying to bake. One caveat here though is that the process of presenting problems to solve has been established. If there’s no discovery, then the leadership can’t create effective objectives and this situation snowballs from there. Which often leads to the creation of issue number 2.
2. You’ve Overfit the Product – In statistics, the goal is to build a predictive model from a random sample of data where there are a limited number of variables. But if the creator of a model puts in too many variables, some of which aren’t significant to the population at large, they create a model that is great at predicting the data included in the sample but terrible at generalizing thereby rendering it poor or even useless. Without discovery and OKR’s, stakeholders and sales people often bring singular client requests to the table. Jason Knight summed this up perfectly:
“For all the talk of ruthless prioritisation, one pattern I’ve seen before is that a company has Way Too Much Stuff to do, and all of it is somehow equally valid as the top priority. This can kill attempts to prioritise or trade-off. It’s all important!
Companies can end up in places like this because they went where the deals were (no well-defined ICP), they went fast, leaving them with a wide but shallow product that does just enough to cover the basics for many but not delight any.
They then end up with 15 different things that legitimately need to be improved (or, in worst-case scenarios, fixed) and enough revenue tied up in each to make it impossible to just ignore something for a year until you get to it.”
This is a sure-fire way to create feature teams constantly feeling behind, likely suffering in terms of quality and probably feeling like their product’s addressable market size – the number of people that benefit from the product – is very small.
3. Treating Models as Deterministic – Once a backlog becomes unwieldy, it’s tempting to apply modeling techniques to it and see them as the silver bullet to the prioritization problem. However, every model (that I’ve seen anyway) has at least one and usually many non-deterministic inputs to it. Behavioral researchers use these types of measurements in places where absolute measurements don’t exist (Likert Scales for instance – that ask you to strongly agree, agree, disagree etc.). To be fair these are better than nothing and can reveal insights. But these are things that can only be estimated or hypothesized but not demonstrated and they’re even subject to human biases. Intuitively (and ideally from listening to customers) a product manager can describe things like impact or potential. But translating that to a number and then apply to a formula doesn’t change the fact that the underlying driver is the Product Manager’s intuition. Don’t get me wrong, there’s nothing inherently wrong with using models and they do sometimes create thought and perspective just by going through them that may not have otherwise been realized. But be careful to not treat them as 100% gospel. They’re nuanced. And remember, even Einstein’s Theory of Relativity doesn’t hold up under certain conditions. Candidly, this is what’s both beautiful and difficult about product management – it’s not easy!
Hopefully understanding these challenges with prioritization will help or at least give you something’s to think about while grooming roadmaps and making decisions on next steps.