Why are you building what you're building?
Every data professional has worked on a dashboard and wanted to cram all sorts of fancy charts/data points in it. Every data professional knows that no one is really going to use the dashboard properly.
They’ll still want to make the dashboard as awesome (from their POV) as possible.
Just so y'all know:
— Max Prilutskiy (@MaxPrilutskiy) March 27, 2024
Every time you push code, you appear on this real-time wall at the @github HQ. pic.twitter.com/rqfpDPDDjB
How are these metrics being consumed to create insights or action points?
— Darshan Pania (@i_m_Pania) March 27, 2024
No doubt this is one of the coolest dashboards I’ve ever seen, and no shade towards Github, this might even have some actionability (that I don’t see at all).
But this issue goes beyond just dashboards and extends to practically any data related feature.
Knowing your customer is the first step to working on any feature.
For a marketing automation company, the customers are marketers who usually don’t have any real analytics experience, and are under pressure to hit their KPIs. Can one really expect them to be concerned about the MAP and AUC of a predictive model?
Too many data science features in the market are just vanity and have ~0 adoption, because they were built by people who don’t understand their customers. These features just end up as marketing fuel.
🔥Gen-AI powered
🦾Advanced machine learning methods
AI-first…🤖
we basically used all our resources to build that... which our competitors were trying much simpler and more scalable solutions (e.g. segment by universities, make everyone friend one account) that were "good enough" and in fact preferred by users!
— Joyce Park (@troutgirl) March 26, 2024
The biggest gift a startup
It is so easy to fall into this trap. Everybody thinks their approach is unique and game-changing.
None of your novel algorithms and techniques matter if the intended user doesn’t enjoy using your product.
Wrote this to remind myself to not fall into this trap.