Upstream thinking
A short read:
Imagine a mountaineer gets a deep wound while climbing. He has a long way to go till he reaches to the top. So he decides to just patch his wound with a band-aid. The bleeding stops for a bit and he continues to climb, within a few hours he sees the band-aid is ripped and the wound is bleeding. He repeats the process because the move to the top lures him. He does this a couple of times until he manages to reach a plateau. At this point the mountaineer is a bit dizzy which halts him from moving forward. Luckily there are a few folks out there, of which one is a doctor. He mentions that the mountaineer has lost a lot of blood and must get immediate help, any further delay will cause amputation. To not make this post barbarous, the mountaineer was saved as he got the help in time.
Now you must be wondering why is this relevant.Think of the numerous times when you have solved for symptoms and didn’t take the time to dig deeper into what may be the underlying root cause. The mountaineer should have realised that it was probably a bigger problem when he had to apply the band-aid the second and the third time.
Our minds are by default lazy. It jumps to conclusions, has it’s biases and doesn’t like thinking too much. This is what is called System 1 thinking which is reactive in nature. It creates a sense of urgency and requires immediate resolution. So when solving problems our first response is to keep patching for those symptoms (downstream thinking) instead of identifying the root cause and nipping it in the bud.
As a product manager we do this by looking at lagging metrics like revenue and churn and not focusing on leading indicators, because tracking them is hard. Take an example of a subscription business, that has a high churn rate. Most organizations take a downstream thinking approach and start sending emails, deals, offers to these customers in hopes that they may renew. But if this problem was approached using “upstream thinking”, the product manager would have tried to get to the heart of why this was happening i.e. instead of customer and revenue churn, they would be looking at “activity churn”. Probably the users never came back after the first month, maybe their week over week retention was declining or their product usage was low. You can identify these patterns by identifying the ways in which users engage with your product and then assigning a score to each user based on their activities. It is then easy to sniff out activity churn and combat customer and revenue churn before it happens by acting on these leading indicators. That said you don’t need to send a “we miss you” email to every user who doesn’t use your product for a week.
To summarise it is paramount to identify if you are solving for the mere symptoms or for the underlying problem.
