Keeping Up With Data #101

5 minutes for 5 hours’ worth of reading

Adam Votava
4 min readSep 23, 2022

Measuring the value of data is becoming a standard agenda on many of my conversations.

It’s probably due to our focus on data due diligence and value creation that people often ask us about the value of data itself.

My answer is typically anchored in three points:

  1. Data has economic value, which one realises through better decision making, improving efficiency of operations, or creating new revenue streams. All of which has value — realised, probable, or potential (see the image above).
  2. Realisation of economic value of data starts with a business problem. A decision to be made, a metric to be increased or decreased. Without a clear business problem, the value of data is not even potential, it’s only theoretical or hypothetical.
  3. Focus should be on creating value, less on attributing the value to data. From the shareholder’s perspective the value is the primary, attribution secondary concern.

I do understand data leads wanting to put a big value on the data to make their agenda more prominent. But I strongly believe the spotlight shouldn’t be on data, but on the value it creates or can create.

Let’s not make the data conversation about ourselves, but about the business. In the end, data valuation is an economic exercise.

This week’s reading list looks at mindsets behind data project and data products, goal setting, and Fivetran.

  • Data Projects vs. Data Products — Why Mindset Matters: As data and analytics are increasing its impact on organisations performance, the data teams are moving from ad-hoc analyses to running machine learning models or BI dashboards in production. And this shift from data projects to data products is not easy. The data teams who were used to work on concrete outputs as specified in the project scope are now being challenged by taking an adaptive approach to abstract problem definition simply aiming to ‘maximise user value.’ To adapt requires changing a mindset (which is always hard). The article outlines the differences between project and product mindset when it comes to data. Something to keep in mind when your data teams are moving away from ticket system or pure R&D into data products. (Clayton Karges @ Medium)
  • TBM 41/51: Why Goal Cascades are Harmful (and What to Do Instead): Setting company’s goals often starts with the executives deciding on a few high-level goals and “department leaders then work to mimic that structure fractally.” The result is a beautiful, symmetrical, organised cascade. However, this is not how the real world works. The cascade assumes there are simple relationships between individual hierarchical goals and it ignores the fact that goals are often influenced by many factors and many other goals. A suggested alternative starts with “creating a model that maps the drivers of your business.” And such model looks very much like what I call a business data twin. Having a twin of the business is not only helpful to succeed with data, but it helps you optimise the processes, and (apparently) set (and measure) the right goals. (The Beautiful Mess)
  • How Fivetran fails: Having a tool extracting and loading data from SaaS apps and data warehouse is very handy. However, these “ELT tools often operate as an unofficial middleman, manually mapping themselves to the APIs of the services they source from. […] As customers, we pay that price somewhere.” (Benn Stancil); and How Fivetran + dbt actually fail: The cost comes by Fivetran deciding on the (highly normalised) data model and charging per row. We get all we need, very fast, but then we pay increasingly more in time. The mantra of cloud providers about costs of data storage and processing being practically zero doesn’t work when every data field is manipulated and moved dozens time. (Lauren Balik @ Medium)

Today is a new bike day, which as put by The Cycling Independent is “like a birthday that falls on Christmas, but in a leap year during a full moon when your tax refund arrives in the mail and your grandmother drops off a fresh batch of cookies.”

So, if you excuse me now I’m going to get my new bike!

Enjoy the weekend!

In case you missed the last week’s issue of Keeping up with data

Thanks for reading!

Please feel free to share your thoughts or reading tips in the comments.

Follow me on Medium, LinkedIn and Twitter.



Adam Votava

Data scientist | avid cyclist | amateur pianist (I'm sharing my personal opinion and experience, which should not to be considered professional advice)