Share Unreflected data insights
Share to email
Share to Facebook
Share to X
By Timo Dechau - Deepskydata
The podcast currently has 15 episodes available.
How to define a data product?
Not that easy I have to say. And also I am not so good with definition. So I give more an explanation how I see data products and what makes them different compared to for example dashboards.
My definition right now:
Data products help to solve problems with data.
I will talk a lot about the problem part of this definition.
So, here is the recap:
- Why writing about building was much better than expected but still can better
- What went wrong with the dev setup for Datatasks
- What kind of learnings I take with me for the next months
Startups know this phenomena as well. In the very early days when there are 4 people sitting in one room the whole time churning away code, ideas and tasks. That changes drastically when the company then grows. The lost in translation problem starts to occur. Context gets missed out, people don't communicate, processes need to be established.
I have my own lost in translation issue with Datatasks. I decided to split up frontend and backend and let the development been done by different persons (for some time including me). For Mind the docs I decided differently and we develop everything with Django. So one dev is responsible for frontend and backend at the same time. Works 100% better. So learning for me, focus resources so no lost in translation happens in the first place. That something for a later stage.
It's not that I did not made some progress. But in the main area I want to make progress - the datatasks app. I did some progress in the backend. But in the frontend nada. It's the initial phase where we need to align frontend and backend to finalize the configuration. And this is right now pretty asynchronous back and forth. This needs to change.
For my "Mind the docs" product I planned to do the full development. I don't have the budget to have two dev people working on my two products at the moment.
And with this decision I was again tinkering how I should approach the dev setup for Mind the docs. And this was a good indicator for me to investigate a bit deeper what kind of tasks work good for me (where I can excel) and which ones block me.
Spoiler: For development the frontend part is definitely a blocker - so outsourcing at least HTML & CSS is a must for me if I don't want to end up in a demotivated state.ย
And you will also hear why I will go back to my Django roots.
I am announcing my second product: mind the docs - https://www.mindthedocs.app/
It's basically coming from datatasks where I was initially planning it as a feature but recognized that it does not really fit there. So I already had the idea to make it a stand alone product.
After spending some time on concepts and early wireframes I decided to launch it as my second product (even when the first one is not live yet).
The good thing is that it shares a bunch of code with datatasks and both work neatly with each other.
Here is the official announcement I wrote on Linked In:
๐ค So what is next for data & analytics documentation?
I did a small announcement on Monday and let me do it bigger today ๐ .
Presenting:
Mind the docs - a documentation platform for analytics and data setups
I am trying to build a bridge between great documentation and few time to maintain it - by using assistants that prepare and update parts of the documentation for you automatically.
Registration is open for early access:
https://www.mindthedocs.app/
Because..
Everyone likes great documentations but no one likes to create and maintain them.
I do data consulting as my major job and currently building datatasks is done during my free-time. Currently most of the work is done by external devs because I am quite occupied by my consulting work.ย
So datatasks is making progress even when I don't have time. But it also occupies me from recording updates.ย
Now I'm back with an update.
I was thinking a lot how I should handle the different checks in Datatasks. Simply having a list of different checks for the different analytics platforms?
It didn't click for me. After a lot of thoughts I had the idea to create something like an app store. So basically a checks store.ย
Currently I am planning to bundle different the types of checks into a package. So the "Google Analytics Quality" Package. So datatasks users can simply take a package and activate it for their accounts.
The podcast currently has 15 episodes available.