top of page

The #1 Platform Engineering Problem You've Never Heard Of: Platform Decay (Webinar Recap)

Last week, Abby Bangser, Principal Engineer at Syntasso, presented her latest thoughts on "The #1 Platform Engineering Problem You've Never Heard Of: Platform Decay." This kicked off our new monthly webinar series.




Read on for a brief webinar summary and to learn why you need to care about platform decay!


Everyone is building a platform; do it intentionally

Abby opened the webinar with a reminder that if you are shipping software to end users, you have built a platform. You have a "platform" even if you've built this by "accident" or it has emerged through a natural evolution from continuous delivery (CD) or infrastructure as code (IaC) practices.


To counter against platform decay, she recommended stepping back and ensuring that your platform is designed intentionally. You must consider the three core goals of a platform, regardless of your organisation's mission or size: enabling developers to go faster, safer, and at scale:


  • Go faster: Platform teams need to provide “everything as a service” to help rapidly and sustainably deliver value to end-users

  • Decrease risk: Teams need to automate manual processes in reusable components

  • Increase efficiency: You need to manage and scale your digital platform and resources as a fleet


Abby stated that platform decay can be an issue regardless of organisational maturity. Entropy naturally increases over time within any software system, and a platform is no different.


She set the stage for the remaining discussion by introducing three signs of platform decay and outlined how these can be identified, explained (across the organisation and at varying levels), and fixed.



Lack of adoption: Navigating the trough of disillusionment

Launching into the first of the three platform decay topics covered in the webinar, Abby referenced a quote from our COO, Paula Kennedy, that was captured by Mark O'Neill, Gartner's Chief of Research for Software Engineering, at KubeCon EU 2024.


Gartner has recently cautioned that platform engineering is heading for the "trough of disillusionment" in their 2023 Hype Cycle. Using a "build it and they will come" approach can be dangerous.


Abby suggested it was essential to understand the developers' needs for a platform before building it. These needs must also be constantly validated throughout a platform's lifetime. Signals such as low adoption, high cost to onboard developers onto the platform, and poor developer survey results (devex NPS) can be lagging indicators of platform decay.


Applying a "platform as a product" approach and employing a dedicated "enabling team" (in Team Topologies parlance) can mitigate these challenges.


Low platform ROI: Avoiding the money pit

Next on the list of platform decay topics was a "low platform ROI", which began with Abby referencing an interesting Reddit thread asking, "Am I a bad platform engineer [...]"


If no platform roadmap is defined or there have been no deliverables for 6+ months, this can be a leading sign of a looming "money pit" platform antipattern.

Abby suggested this problem can be mitigated by identifying, scoping, and delivering a platform-based fix for a specific problem for a specific team. Unless you solve an immediate pain with your platform engineering efforts, your effort can easily become misdirected.


Never-ending platform migrations: Delivering value incrementally

Quoting Charity Majors, "if you are not migrating, you are stagnating", Abby cautioned that it was all too easy to have multiple (large-scale) ongoing platform migrations.


In this case, platform teams can often spend more than 50% of their time on migration-related work. If they are performing this kind of maintenance work, they are not adding direct value to the business.

Abby suggested, yet again, that the "platform as a product" mindset is essential for combatting this lagging indicator of platform decay.


Wrapping up

Concluding the webinar, Abby reminded attendees that everyone is building a platform. Being intentional about this and using a "platform as a product" approach is key to avoiding many issues that foreshadow a decaying platform.


Your platform should support the following goals:


  • Provide “everything as a service” to help rapidly and sustainably deliver value to end-users

  • Automate manual processes in reusable components

  • Manage and scale your digital platform and resources as a fleet


All platforms will naturally evolve and change over time, but a platform engineering team should constantly look for leading and lagging indicators of platform decay.


Learn more

You can learn more about building platforms as a product from these resources:




Any questions? Please get in contact!

Thanks to all attendees for asking great questions during the live event. We are always keen to answer more, and so please reach out to us via the Kratix OSS Slack or on social media!

185 views0 comments

Commentaires


bottom of page