Our practices for producing analysis content have shifted meaningfully, and the specific changes are worth naming.
We got more disciplined about starting from evidence rather than conclusions
Early analysis pieces sometimes started from a working thesis that shaped which evidence got emphasized. We now try harder to let the evidence genuinely drive the conclusion, even when that means the piece ends up somewhere we didn't initially expect.
We started killing pieces more readily when evidence didn't support the thesis
We've become more willing to abandon an analysis piece mid-process when further reporting undermines its original premise, rather than publishing a weakened version of the original argument.
We got more explicit about distinguishing analysis from opinion within pieces
We try harder now to make clear, within the piece itself, when we're following evidence toward a conclusion versus when we're stating an interpretive judgment.
Why this evolution matters
We think analysis only earns its name when it genuinely follows evidence rather than confirming a predetermined thesis, and we take that discipline seriously even when it costs us already-invested work.