Keeping Up With Data #95

5 minutes for 5 hours’ worth of reading

Adam Votava
3 min readAug 12, 2022

Many years ago, our team was working on a model predicting churn of a bank’s clients. Some colleagues in the marketing and product departments were discarding the importance of such model saying that by that point it’s already too late and one needs to focus on truly preventative measures — mainly designing better products and services.

While the bank had thousands of employees and working on many initiatives wasn’t a problem, people often invested more effort in preventing activities (of others) they perceived as inefficient than actually working on activities they considered efficient. But I guess that’s the nature of large organisations. Analysis paralysis…

As the image above shows, there are many measures aiming to prevent churn. And an organisation should (at least) consider all of them. It’s their combination that is the most powerful.

No matter how good your products and services are, there will always be customers who are more likely to churn than others. And that’s when the churn model comes into play. I wrote about my experience with churn prediction model a year ago.

In this week’s reading list, there is an article by BearingPoint’s consultants echoes many of my points.

  • How Airlines Make Billions From Monetizing Frequent Flyer Programs: I remember reading Doug Laney’s article about United and American Airlines using their loyalty programs as a collateral. The value put on the programs was huge and wildly exceeding the market capitalisation. Why was that? This article offers a peek into the workings of the loyalty programs, associated cashflows, profit margins, and importance of bank partnerships. As the article puts it: “Over the past few decades, airline mileage programs have transitioned from a way to generate a bit more loyalty to a massive profit center for airlines.” And massive profit centres typically have massive value. Yet again, data valuation as an economic exercise. (Forbes)
  • Data & Analytics Recession Playbook: Make Your Data Work Harder…And Smarter! Bill Schmarzo opens by saying: “With a potential recession lurking on the horizon, 99% of companies will make the same old ‘safe’ mistakes: hunker down, let people go, shrink, and hope to hold on for dear life. However, growth-oriented organizations will see this as a business opportunity — an opportunity to leverage their data to ‘do more with less’. You see, data is an asset that all companies already possess.” He then offers six plays for organisations to use the data (and analytics) to their advantage during the tough times. (Data Science Central)
  • Four tips for predicting and preventing churn: Churn prediction is often one of the first projects where companies employ machine learning. Knowing which customers are likely to leave gives an opportunity to change their mind and not needing to win them back or acquire new ones. Consultants from BearingPoint offer a few tips to make the churn models successful — clearly define the churn, use individual churn drivers to find the right action, and find model’s optimal intervention level. (BearingPoint Data, Analytics & AI @ Medium)

Because we live abroad, we get a lot of visitors over the summer. This weekend we’ll be saying goodbye to one of them and welcoming the next four. It’s nice to see the friends and families and show them the beauties of our new home country.

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)