,
8 tweets,
2 min read
Read on Twitter
When people complain about tech debt, and business disregard for it, they don't usually *actually* mean a limited set of specific issues.
What they're complaining about is a lack of "slack" to take on un-prioritized, but important work.
Slack is important for any healthy team.
What they're complaining about is a lack of "slack" to take on un-prioritized, but important work.
Slack is important for any healthy team.
Setting up explicit slack in your roadmap (of 10%-20% of your time) can actually speed up product development and make your roadmap more consistent.
How?
How?
(1) People on the ground have a lot of visibility about what "sucks" for them; what improvements to tools and technical construction can make their flow more effective.
Give them time to work on those improvements.
Give them time to work on those improvements.
(2) people on the ground have visibility into what are of the code is a rat's nest of potential issues. Bugs and SEVs always congregate (a sign we didn't understand the problem well enough).
Let people fix high-risk issues; before they are a roadmap disrupting emergencies.
Let people fix high-risk issues; before they are a roadmap disrupting emergencies.
(3) Engineers often have intuition about what parts of the system are okay for now -- but will fall down later / under load.
Give them time to test their guesses; they will construct load tests, and figure out projects for the roadmap instead of waiting for emergencies.
Give them time to test their guesses; they will construct load tests, and figure out projects for the roadmap instead of waiting for emergencies.
(4) architectural experimentation! Let people test AWS tech and see if it's reasonable to replace your entire stack for something at 20% of the cost. Let people have "hack weeks" that take long slow thinking and results in brand new products.
Empower natural creativity.
Empower natural creativity.
As @mattbarcomb points out, all of the above is related to what may be considered "short-term" debt.
Large projects like AWS migration, or re-arch. for long term growth are a different beast. They're eng-driven roadmap items; which also need to exist, but often don't get prio.
Large projects like AWS migration, or re-arch. for long term growth are a different beast. They're eng-driven roadmap items; which also need to exist, but often don't get prio.
To wrap up for the day:
