Transformation at scale isn't easy

Digital transformation is hard and we don’t talk about that enough.

It’s one thing to craft a vision, but it’s something different to turn that vision into reality. This work requires going deep, getting teams and departments to align and making hard decisions about what’s in and what's out.

This week, we’re sharing case studies from Spotify, Netflix and AirBnB to hear their transformation stories and learn from the hurdles they’ve encountered.

In 2017, Spotify already knew they had issues with their deep links and attribution systems, but no single group owned this functionality. Given their size and scale, it took five years to complete prioritizing, exploring, testing, building and migrating to a solution. In the end, the team's takeaways had nothing to do with technology but rather ideas on how to improve strategic alignment, collaboration, and cross-team communication.

AirBnB wants your search results page to be completely personalized. Getting this level of personalization is easy if you have the data, but it’s too cumbersome for users to enter all that data. For hosts, it would involve adding 800+ potential attributes to their listings, and guests often have a nuance to their needs that isn’t apparent (Is the pool shared or private? What specific kind of crib do you have?) In this piece, the AirBnB team outlines how they’re using machine learning to observe user activity and automatically extract these data points instead of having folks enter them manually.

There's no single movie poster for Netflix titles. The in-house team creates multiple versions for each film but you only see a single image based on your profile and preferences. In this article, the internal team at Netflix goes in-depth about how these images are produced and selected by combining the best of what humans do with the best of what machines do to ensure the end result performs well with users.

Futures & Foresights is a weekly newsletter of curated links from our Director of Technology, Will Turnage .

Subscribe to receive a notification when each week's edition is published.