Key Quotes
I think you can approach the problems in enabling ROI for data leaders through the lens of how we think about observability. You have detection solutions, you have resolution, and you have prevention. On the detection side you have automated machine learning driven monitors. You have ways to target your alerting to different teams to make sure you're managing that signal-to-noise ratio in terms of alerts. Then on resolution, you have tools where you can actually look upstream as an analyst, see the initial cause of the data incident that you're investigating, be able to resolve it, and talk to the right partner upstream. And then also for those data producers to be able to look downstream and see the full scope of an incident on their side, I think that's just a phenomenal innovation in this space.
We typically think of one of the issues of data quality being downtime: the erroneous, missing, incomplete, or delayed data that often plague data initiatives. The consequence of downtime can range from this almost trivial outcome where engineers or analysts respond, and the result is the hours lost to address the issue, to actually more existential, where you're losing trust, revenue, or even customers. And then, at the far end of the scale, you could actually be putting in danger the reputation of the business.
Key Takeaways
Before decentralizing a data team, it’s important that the data team is sufficiently mature to be able to handle decentralization efficiently and effectively.
Data teams should be focused on building data products that actually drive revenue in line with the organization’s goals.
It’s important to get the basics in place that free up your data team to do more expansive data roadmap work, such as self-service access, so stakeholders can get answers to basic questions without taking up team bandwidth.