Stop Doing It and Wait for the Reaction

Sarah was out sick the week of the fourth, so the pack did not go.

Nobody said anything. Not that week, not the week after when it resumed, not since. It takes her three and a half hours to build and it goes to eleven people, and the one time it failed to arrive, the silence was total.

She has been thinking about it ever since and has not told anybody.

That accident is the only real test anybody has ever run on that pack, and nobody meant to run it.

Thirty Days, and Nobody's Permission

So run it on purpose.

Stop producing it. Do not announce it, do not ask, do not send a note explaining that you are pausing pending review. Just stop, and wait.

A weekly pack that takes three and a half hours to build means four missed editions. Then one of two things happens.

Somebody Notices, and That Is the Good Outcome

If somebody comes looking, you have learned something an audit could not have told you.

You now know who. Not the distribution list, which is eleven names and means nothing. The one person who actually opened it, actually used it, and noticed its absence inside a month.

Ask that person what they do with it. Almost always the answer is a fragment. One table, one number, one comparison. The rest of the pack is scaffolding around the part somebody needs, and it has been rebuilt every week for years because nobody asked which part.

That is not a failed kill test. That is a scoped requirement, arrived at for free.

Nobody Notices, Which Is Louder

Four weeks of silence from a distribution list of eleven is not ambiguity. It is an answer.

The honest move at that point is to stop permanently and say so, once, in writing. Not as an announcement, as a record. The pack has not been produced since the fourth; if you need it, tell me what for.

You will get nothing back. That is the point, and it is also the evidence you were missing.

What the Silence Does Not Tell You

One limit, and it matters more than the test.

Silence tells you nobody needed the artifact. It does not tell you nobody needs the question underneath it.

Somebody once asked something real. The pack was the answer, and the answer got scheduled. Killing the pack kills the answering, and if the question is still live you have removed the only thing watching it.

So before you stop producing, work out which of two things you are holding.

Looking Back, or Watching

Every standing artifact came from one of two intents.

• Looking back. Something happened and somebody needed to know what. That question has a finish line. Once it is answered it stays answered, and the artifact has nothing left to do.

• Watching. Somebody wanted to know if it happens again. That question does not close, and it is a legitimate thing to want.

The silence reads differently depending on which one you have. If the intent was looking back, four quiet weeks confirm what you already suspected.

If the intent was watching, four quiet weeks tell you only that nothing crossed the line during those four weeks, which is not the same as nobody needing to know.

Two Hundred Pages to Read the Total

The watching case has a better answer than a weekly pack, and it is about forty years old.

Somebody used to run two hundred pages onto green bar paper, carry the stack to a desk, and read the total on the last page. Everything above that total got printed because printing was cheaper than deciding what mattered.

Management by exception is the response to that. Set a threshold, get told when something crosses it, and stop reading the rest.

A pack with four thousand two hundred rows in it, where about twenty-five cross any line worth crossing, is green bar with better fonts. The work is not producing the pack. The work is knowing the line.

So the Fork Is Not Keep It or Kill It

Run the test and you land in one of two places.

If the intent was looking back, the question closed and the artifact did not. Stop permanently. Nothing replaces it and nobody has to buy anything for you to be right.

If the intent was watching, the need is real and the instrument is forty years old. Something does replace it, and now you are asking for a purchase, which is a different conversation in a different room.

Same thirty days. Two very different next meetings.

Start With the One You Are Least Afraid Of

Pick the artifact you produce that you are most confident nobody reads.

Stop for thirty days. Do not explain, do not soften it, do not build a one-page replacement to prove you are being reasonable. The whole value of the test is that it costs nothing and asks nobody.

Thirty days of silence will tell you more than any review you could schedule, and the review would have needed somebody's permission.

-BZ