Your First 30 Days of Daily Management: A Week-by-Week Plan
A week-by-week plan for how to start daily management in one cell, what to add each week, what good looks like at day 30, and where week two usually breaks.
The most common way to start daily management is to launch it in every cell on a Monday, with a laminated board template and a memo from the plant manager. The most common way it ends is three weeks later, with blank boards and a supervisor who says "we tried that." If you want to know how to start daily management so it is still running on day 30, the answer is smaller than you think: one cell, one board, one new thing per week.
Week one: one cell, one board, one huddle
Pick a cell with a supervisor who wants it and a problem everyone can see. Not your worst cell, because it has no slack. Not your best, because nothing will change and nobody will learn anything. A middle cell with a delivery gap people already argue about is ideal.
Spend two days getting a whiteboard, a marker in three colors, and agreement on four metrics: one each for safety, quality, delivery, and cost. Yesterday's number, green or red against a target. That is the whole board.
Then start. Five-minute huddle at the board, same time every day, same person running it. The supervisor fills the board before the huddle, not during it.
One rule from day one: any red gets a name written next to it before the huddle ends. Not a plan, not a form. A name. "Who owns it, and by when?"
Nothing else this week. No hour-by-hour, no 3C forms, no second cell. The goal for week one is five huddles in a row with the board filled before each one. If you try to do more, something gets skipped, and the thing that gets skipped is the habit.
Week two: add hour-by-hour
Add a column to the delivery metric: plan and actual per hour, running gap at the bottom. Pre-print the plan numbers so the lead operator only writes the actual and the gap. Top of the hour, under 30 seconds.
The supervisor checks the column at 10 AM and 1 PM. Not "how's it going." "Show me the board." A gap at 9 AM that gets seen at 10 AM is recoverable. The same gap seen at 3 PM is a loss.
This is the week that breaks most often. More on that below.
Week three: add 3C for reds
Any metric red two days in a row gets a 3C: concern, cause, countermeasure, owner, date. One sheet, taped next to the board. The printable 3C form has all five on it. Keep the writing under five minutes; if it needs more than a sheet, that is an A3, not a 3C, and it goes to the tier 2 board.
Cap it at three open 3Cs on the board. If there are more, the cell is collecting problems, not closing them. A closed 3C moves to a folder. Do not throw it out. In six weeks you will want to know whether the same cause came back.
Week four: first review, and the second-cell decision
Somewhere between day 26 and day 30, sit down for 30 minutes with the supervisor, the lead, and whoever sponsored this. Four questions:
Was the board filled every working day? Did every red get a name? How many 3Cs opened, and how many closed? Did hour-by-hour catch a gap before the end of a shift at least once?
Add a second cell only if all four answers are yes. If one is no, run the same cell another two weeks and fix that one thing. A 60-machine shop that now runs boards in 17 cells got there over about two years, one or two cells at a time, and the first cell took closer to 60 days than 30.
The second cell copies the first board exactly. It does not improve it. Improvements come after the second cell has held for a month too.
What good looks like at day 30
The board was filled 18 or more of 20 working days. Fourteen is not good, it is a habit that has already started to slip.
The huddle runs under 10 minutes and the supervisor runs it without the plant manager standing there.
Every red on the board has a name next to it. Two or three 3Cs were opened and at least one closed with a countermeasure that is still in place.
There is one story, told by the lead operator and not the supervisor, where hour-by-hour showed a gap by 9 AM and the shift recovered it.
What good does not look like: all green. If everything is green at day 30, the targets are too easy, and the board has nothing to say.
The most common way it fails in week two
Hour-by-hour dies. The lead skips one hour on a bad day, because the bad day is the thing they are dealing with. Nobody says anything. The next day two hours are blank. By Friday the column has more blanks than numbers, and a board with blanks stops being trusted, and a board nobody trusts stops being read.
Two things prevent it. First, the entry has to be a 10-second job, which is why the plan column is pre-printed. Second, a blank hour gets asked about at the next morning's huddle, not to blame anyone, but because the blank hour is usually the hour where the problem happened. "What was going on at 11?" is the most useful question in the first month.
The second way it fails in week two is quieter: the person who sponsored it stops showing up. Be at three of five huddles for all four weeks. The cell notices when you stop, and reads it as permission.
If you want to see what a cell's board looks like with a week of history behind it, one open 3C and two problems already moved up to an A3, the demo runs on sample data, no signup needed: Try the live demo →