Further’s Post

Transcript

If we think of data quality, what's the definition of data quality? If data quality is data that is fit for purpose, then the data in the warehouse was fit for purpose, for it was good quality for analysis. It was terrible quality for the fraud department. And that's. It's not even an extreme example. It's a fairly typical example of having two contradictory use cases that demand different qualities of data. Now I said there's two things that you could take from this and one of them is they supported question that data has to be fit for purpose and that might not be the same purpose. So the idea that we have a single version of the truth isn't necessarily true. We have a version of the truth which is fit for it for for a particular purpose. That's really important. The second thing we can take away from that is that when it comes to the the work that we do with products of trying to understand use cases and understand, trying to understand users. We actually need to understand not only what are they trying to achieve with the data, what are they trying to, what are they trying to build, but how may that change over time. And in particular, and this is something which is a product designer, I've always had to bear in mind, it's not just what somebody tells you they want today that's important, but how the very fact that they are using your product and therefore becoming more analytical and more insightful means that in the future they have other questions to ask. And are you leading them in that direction so that you will be able to answer these future questions that they're able to ask precisely because they have developed a set of analytic skills and analytic insights from using. The the the the analytic ecosystem today, so it's complex to define and data quality and fitness for purpose, but this is also continually moving target.