What is an “edge case” when building products ?
‘Edge case’, A term frequently used by product teams to decide if a specific user need in question is large enough to be built or not. The trouble is that these discussions on edge user needs or behaviours are usually subjective. Moreover, objectively confirming if a need is felt by the majority of the user base requires analytical and/or qualitative deep dvies, which might not be worth the effort.
Let’s take an example of a fictional user need,
In Pinterest, consider a case of pinners who use Pinterest to save articles (instead of images) as a private ‘Save for later’ repository (instead of dedicated services like Pocket). These pinners are insecure that the source of the article might shut down one day (common with blogs and publishers) and need a way to save to the cloud for later access.
In such situations, I find it helpful to ask a simple question that helps the team effectively discuss such cases for a short and focussed period of time and then move on to more pressing concerns if needed. Here is the question :
How many things need to be true at the same time for users to have this specific need?
Or more simply put, the more coincidences that need to happen for a problem to manifest for users, lower the priority of the need in question should be. As you visualise the situation, count the number of ‘Ands’ and ‘Ifs’ you use. Doing this will make the priority of the specific need jump out, often dramatically simplifying the discussion. Let’s try it for the Pinterest example above:
In the example, the need to save articles by pinning will manifest if the following happens:
‘If the user is a pinner (and not just a browser) and the user is already using Pinterest to save articles and the user tried revisiting one of the previously pinned articles and the publisher of the article has shut down and this situation has repeated at-least twice (to help the user realize the pain and appreciate the solution when it comes by) and the user has enough articles pinned in Pinterest that she is locked-in and does not want to move to a competing product and the pinner actually discovers the new functionality when it launches’
We used 8 ‘Ands’ and ‘Ifs’ indicating the conditions that need to be true for this need to be felt by users. If we started of with 100 users and drop (a conservative) 15% of users with each ‘And’ and ‘If’, then you are already at ~27% of the user base. This means the solution you build will probably not get used by more than 70% of the users. Now it’s up to you to decide to spend the effort to find a good solution, build and iterate on it or not, based on your current goals and the metrics you want to move.
Of course, there are different ways of looking at the customer journey, but the idea is to have a quick mental model to help guide such discussions and quickly find a path forward as a team. Your team’s time and focus is expensive so help them invest it in the right things without analysis paralysis.
Share some of your own thoughts below or send me a tweet at @karanperi
