A story of data and tools
We all need some key data to run our project, our company but we have tons of them, spread over multiple tools and generated every time we interact with them.
We choose at Meta API to create a dashboard to make our key data available to everyone who bring value to our company: employees, investors and advisors because we believe these data help people to see "the big picture" to be invested by common objectives and, more important, to see their impact in their day-to-day work.
So we create Polaris, the "North Star" for our company to display and track these KPI and we want to build them from the most relevant data.
Polaris : our homemade tool to track our KPI. Of course, there is some data instead of emoji 😉
We figured out two things :
We need to get data from multiple tools, multiple SaaS platform (like Stripe, Matomo, Google Spreadsheets) and our own data, the most valuable
We need to process these data: filter, aggregate and transform to make them relevant and easy to understand for everyone.
A journey through an ocean of data
Our data are nowadays spread on three different tools:
In the future, we want also to plug in:
How we build Polaris with Meta-API
We would like to create a quick tool, in less than two days of work to display these data.
We split our work between Marcel and myself :
Here a back-stage look on the Spells we created :
Polaris' Spell for processing Stripe data
Here some highlights about this Spell:
Normally, we only need two connectors : one for getting the initial page of invoice and the second one (the same one) to get more invoices if we hit the pagination limit. We double it because we retrieve, at the same time, the previous period to see the evolution for month-to-month for example
We all to personalize using request's params the period we want to retrieve : 1 week, 2 weeks, 1 month, 3 months, etc.
So in less than 130 lines of code, we create and abstracted API able to:
As we created this Spell, all the monitoring is done automatically and if Stripe is changing it's API, we will be notified.
Our monitoring interface to check status and performances of our Spells
We created 4 different Spells like this one :
One for Stripe's data
One connected to our endpoint to retrieve global statistics
One connected to a specific endpoint to follow our user's onboarding process
One for Matomo's statistics
And for our upcoming tools and usages, we'll just need to add more Spells or to improve the existing ones to extend their capabilities. All these endpoints are secured, monitored and can be used on API client we need.
What Polaris brings to everyone and why we'll make it our only "North Star" to follow
Making Polaris available to our team brings 3 major changes to our organization:
Anyone can look on these data which create stimulate discussions and ideas on how we can improve them. Everyone feels both concerned and powerful over this metrics.
We know we can add more KPI in a snap, from any tools and without spending days maintaining them.
We have the power to correlate data from different tools without depending on their built-in integration, making us closer to our users and customers.
Data is power when you need to take enlightened decisions for your company: so take them and share them to your teams, with your own tools and your own style. It's the only one that matters.