You made the correction, returned to the evidence, and expected improvement didn’t appear. Field Note #13 shows how disciplined founders interpret verification as new evidence—investigating insufficient time, execution variance, insufficient leverage, and incomplete diagnosis—then deciding whether to extend, strengthen, or replace the correction without piling on unverified fixes.

When the same problems keep returning, reliability isn’t improving—it’s repeating. Field Note #12 introduces Operating Learning: using evidence to distinguish normal variation from recurring operating problems, make a focused correction, and verify whether reliability actually improves.