Why This Course: Feeling Faster Is Not the Same as Being Faster
The sentence you're about to hear
Twelve weeks from now, someone on your team is going to say "this automation push is really working." This course exists so you have a real number to check that sentence against, instead of just trusting it.
The trial that should worry you
In a randomized controlled trial run by METR, experienced developers used AI tools on real issues from their own repositories. They took 19% longer to finish. Afterward, they believed the tools had made them about 20% faster. Wrong in the flattering direction, by close to forty points, in people who knew their own codebases cold and had a clock running the whole time. If a single task with a clear start and end can fool experienced developers that badly, a twelve-week program with no fixed comparison point is the harder case, not the easier one.
The gap this course closes
Twelve weeks in, a program can feel completely transformed and be running flat, for the same reason developers in that trial felt 20% faster while running 19% slower: a felt sense of progress is not evidence of progress.
If you're running (or reporting on) an automation push and your only read on how it's going is how the weeks have felt, you're in exactly the position that trial warns about, except with less structure than a single measured task ever had. This course fixes that with one number that doesn't move: a Week 1 baseline, logged once and never re-measured, that every later week gets checked against.
What you'll be able to do
By the end of this course you can install a real, tested Claude Code skill (not a hypothetical, working software you clone and run) that logs five metrics once a week against that fixed baseline, with revenue stored only as a ratio so the dashboard is safe to share without exposing your books. You'll be able to catch a week where the numbers disagree with how it felt, and trust the number. You'll point the plain JSONL log at whatever charting tool your own project already uses, correct a real logged mistake the sanctioned way without rewriting history, and read the graduation gate correctly, including a real inconsistency in the original program that this course names and fixes rather than hides.
Twelve weeks. Five metrics. One number that never moves.
The full case for measuring over feeling, and the fixed-baseline fix this course runs on.
Reply here and it goes straight to Rod. Same as replying to one of his emails.
