Most job-search tools are document tools: a CV builder, a cover-letter template, a profile headline. Building a peer community for people between jobs kept surfacing a different failure. Searches did not usually collapse at the document stage. They collapsed in week four or five, when the days lost their edges — every hour felt like job search, almost nothing felt finished, and waiting on an inbox became the whole job.

So this part of the build was not another template. It was a working week: defined blocks for applications, a small separate slot for networking, deliberately protected non-search time, and a simple pipeline record — applied, replied, interviewing, closed — that turns waiting into information.

Designing the week

Three choices shaped this part of the build.

First, applications live in a bounded block, with a number settled in advance. This many tailored applications, then the block ends. An open-ended "apply all day" rule produced volume on day one and avoidance by day ten.

Second, networking gets its own small slot — one message, one introduction asked for, one former colleague checked in with. It is the activity most likely to be postponed forever when it has no slot at all, precisely because it has no deadline attached.

Third, the rest of the week is deliberately not job search: a course module, a project, exercise, family logistics. Those hours are not a reward for progress. They are what makes the search blocks sustainable at week six instead of only week one.

What worked

The pipeline record did more than the motivational copy ever did. A simple applied / replied / interviewing / closed list turns a closed line from a verdict into data: it says where effort actually went and which kinds of roles answer. Members who kept the record could adjust — fewer applications to roles that never replied, more time on the categories that did. Searches run on memory felt worse than they were; searches run on a record could be changed, and a changeable search keeps its owner's confidence intact longer.

Bounded blocks also made "done" possible. In an all-day vigil there is no point at which the day's work is finished, so rest always feels like quitting. With a settled number and an end time, finishing the block is a real event, and people reported protecting the non-search hours instead of apologising for them.

The networking slot worked because it was embarrassingly small. One message is a task someone will actually do on a bad day. "Do networking this week" is not.

What broke, or nearly did

The first version tracked too much. Stages multiplied — researched, drafted, submitted, viewed, ghosted — until updating the record took longer than some applications. A tracker that feels like a second job gets abandoned, and an abandoned tracker is worse than a simple one, because the gap in it reads as failure. The record was cut back to four stages and stayed there.

Streaks broke next. A daily-application streak looked motivating in the design doc and behaved badly in real use: the first missed day — an interview, a family emergency, a genuinely bad day — broke the streak and took the whole routine down with it. Blocks replaced streaks. A missed block is just a missed block; the week still has its shape.

The hardest limit is still only partly solved: structure helps the searcher, it does not change the market. A tidy pipeline does not make a slow employer reply faster, and pretending otherwise would be dishonest. What the structure changes is where the waiting sits — inside a week with edges, instead of spread across every waking hour.

Lessons for builders

Find where the process actually collapses, not where the templates are. For this community the collapse point was week four, in the calendar, not week one, in the documents.

Make "done" reachable every day. A routine with no finishing point will be experienced as failure even when it is working.

Keep the record smaller than feels thorough. Four honest stages that people maintain beat nine precise stages they abandon.

Shrink the avoided task until it is undeniable. The activities with no deadline need the smallest, most concrete slot, not the most inspiring copy.

The project behind these lessons is WUUUH, a peer community for rebuilding after a layoff, where people compare routines, share which structures hold up, and report the small wins a pipeline makes visible.