One of the areas that I think companies find hardest to let go of is the concept of locked-down technology stacks. There's no doubt a number of good historical reasons why it was once important to maintain rigid controls over the technologies your staff use, but with the plethora of advanced, secure tools and systems now available, times have changed.
And yet I still encounter teams in companies that are unnecessarily restricted in the technologies they can use and at a time when innovation, prototyping, tesing, collaboration, efficient communication and project management are so critical this becomes a potentially significant problem.
The answer of-course is about balance. To have sufficient controls in place to enable effective technology stack management, remove unecessary compatibility or security issues whilst still empowering teams to do their jobs better, use tools appropriate for the job at hand, and enable healthy levels of autonomy.
It's rarely perfect and there are always challenges but utilising broad lists of approved technologies and tools that then enable choice and flexibility by individuals and teams is a sound approach (and the guiding principles of the GDS technology transformation programme is a good place to start).
So I liked what Robert Rees had to say over on The Guardian's developer blog about the technology stack they use and how instead of being a fixed stack, they allow the teams to choose the technology that solves their problems whilst also having a 'whitelist' of approved tech:
'A lot of organisations standardise technologies to help them move faster. Our whitelist attempts to do this for us. If you believe in autonomy, though, then you have to let teams choose the technology that solves their problems most effectively.'
Like most things it's about balance but it does seem to be one that many companies struggle to move forwards with.