The Metric Showdown
“If you torture the data long enough it will confess.” Ronald Coase
This week in product management we discussed metrics and planning for them. If you think about the right metrics to track the right things early enough then you hopefully get to avoid the gruesome quote I began with. Instead you have a nice dashboard confessing everything in a slick manner.
Team ToDone talked about a few things: the conclusions we can draw from the data we have, the data we’ll need operationally, the data we’ll need revenue wise and data investors like to see.
Let’s start with what looking at our data taught us. First due to lack of use (one total phone call) we’ve opted to kill AlexDoes Assist. This was a product where one of our team members would help you create your todo list. We also looked at the outcome of our last Start Up Sprint, AlexDoes Social a site where you can post your todo items and help other Cornell Tech students check off their items. This had use during launch, but has since gone stagnant indicating it shouldn’t where we focus development for the rest of the semester. That will be on directory integration with current todo list platforms.
Next we talked about what kind of metrics we would need for different aspects of the business. One of the class instructors was making strong points for paying more attention to active usage, not simply things like downloads. Here are a few of the important operational metrics we need to track:
1) Todo list upload attempts (and success rate)
2) Integration pulls with todo list vendors
3) Ratio of tasks that can be completed to total (both by list and aggregate)
4) User behavior metrics - how often do they revisit after upload, time on site, acceptance rate, etc
For sales and business model purposes we identified different metrics. Ones that not only we need to track for revenue purposes, but ones that could disprove our current ideas of what the revenue model should look like.
1) Cost per task and the revenue breakdown from different types of tasks
2) Average tasks per order
3) Spend by user each month (perhaps people set aside a certain budgeted amount and a subscription model would fit their needs? We don’t know yet)
We’ve got a great start for what we need to track and will find more metrics as we move along.
Cheers,
Kate