Let's find out if your organization has a data mess or a data mesh!
Rate the following statements from 0 to disagree completely to 10 to fully agree.
- When taking important decisions my colleagues back them up with data.
- When a new colleague joins and asks how an important KPI is calculated, you can just share a link to the definition.
- People around you understand that data is an asset that gets more valuable when it is used by more people.
- Data scientists at your company spend most of their time understanding business problems and doing analytics.
- Data is accompanied by meta data, making it easy to understand if it is fit for analytics.
Now, sum up your numbers!
If you score higher than 45, you can stop reading this article and go on to training AIs.
For the rest of you I have some good news:
Your data mess can be organized, so you end up with a data mesh [1] that enables your organization to get more value from data [2].
How do we get a data mesh that improves decision making?
- Use our software. It solves all the problems.
Unfortunately this is not true for any software. Problems as complex as organizing a companies information require not just technology, but also people and processes.
Let's start with the easy part. Technology.
Technology for the data mesh should
- be self-service by nature
- scale up and down with usage
- automate the boring tasks
- integrate seamlessly with existing systems
Here is an example for a technology stack that delivers this:
- Fivetran to easily onboard data
- Snowflake to scale workloads and quickly share data
- Sled to discover and trust data
- Tableau to visualize insights
- H20.ai to automate decisions
In the next post, we will discuss this stack in more detail, before we cover the more complex topics around people and processes.
You scored lower than 45 and are unhappy with the status quo of your data landscape?
Get in contact!
[1] https://martinfowler.com/articles/data-monolith-to-mesh.html
.jpg)
