Proof of Resources

Method

Six steps to a resource audit.

An audit answers one question: does the planned work fit the people available, week by week? Each step produces something you can write down, and each has a stop rule that says when you are not ready to move on.

of 6

0 of 6 steps marked done.

  1. To do

    Fix the window.

    Do
    Choose the horizon in whole weeks and write down the first Monday. Pick the shortest horizon that covers the commitment under test: a launch, a quarter, a contract. Decide what a week means for this team, including any holiday that falls inside it.
    Output
    A start date and a number of weeks.
    Stop rule
    If you cannot name the date the plan ends, you are discussing a wish, not auditing a plan.
  2. To do

    Count the people, then their hours.

    Do
    One row per person, or one per team when individuals are interchangeable. For each row, write down the hours per week that are truly available for planned work: contracted hours minus fixed meetings, admin, support duty and recurring leave. For a team row, multiply headcount by hours each. Do the subtraction once, write the result, and leave it alone.
    Output
    Rows, each with a capacity in hours per week.
    Stop rule
    A capacity of 40 because “that is a week” is a guess. Replace it with a number you would defend out loud.
  3. To do

    Turn the work into weekly demand.

    Do
    List the work in hours and spread it across the weeks it will really occupy. If you must estimate in ranges, plan committed work at the upper figure and keep hoped-for work on a separate list. Sum each week into one demand figure.
    Output
    A demand figure in hours for every week.
    Stop rule
    Work with no estimate is not zero. Give it an hour figure, however rough, or exclude it in writing.
  4. To do

    Allocate, person by person, week by week.

    Do
    For every person and week, enter the share of capacity the plan assigns. A cell is everything: all projects plus the non-project work that will happen anyway. Use 0 for a deliberate week off, and leave a cell unset only while you have not decided. Unset counts as zero hours and prints as a dash so it cannot pass as a decision.
    Output
    A filled grid with no unexplained blanks.
    Stop rule
    A dash in a week that has demand is a hole in the plan. Decide it, or flag it.
  5. To do

    Read the heat.

    Do
    Read in a fixed order. Red cells first: each one is a person asked to do more than their week holds. Then the Gap row: negative means booked hours cannot cover the demand. Then Team load: a team at 95% in any week has no slack for anything unplanned. Last, the shape across weeks. One hot week is usually movable; a block of hot weeks is structural.
    Output
    A list of red cells and short weeks, with their week numbers.
    Stop rule
    Do not smooth numbers to earn a pass. If it does not fit, the result of the audit is “does not fit”.
  6. To do

    Decide, record, and set the next look.

    Do
    Choose one lever for each problem (see below), change the grid, and read it again. When the verdict is stable, export the CSV and file it with the date and the decision. Write who owns each change and which events trigger a re-run: a new project, a departure, a slipped date, a change in anyone’s hours.
    Output
    A dated CSV and a written decision.
    Stop rule
    An audit nobody re-runs is a snapshot. Put the next review in a calendar before you close the grid.

When it does not fit

Three levers. Use them in this order.

Every fix to an overbooked plan is one of three moves. The order matters because it runs from cheapest to slowest.

Move

Shift work in time, or hand it to someone with free hours. Cheapest and fastest. Check that you have not just relocated the red cell: after every move, read the whole grid again, not just the cell you touched.

Cut

Reduce scope or drop work. It is the only lever that lowers demand. It needs someone with the authority to say no, so name that person in step 6, not afterwards.

Add

More people or more hours. Slowest, because new capacity needs time to become useful. Enter new capacity from the week it is genuinely productive, not the week it is signed.

Worked reading

Illustrative example

One row, read properly.

These figures are an illustrative example, invented to show the arithmetic. They are not data about anyone.

Suppose a row has 30 hours per week of real capacity. In one week its cell reads 120%. That is 30 × 1.2 = 36 hours of planned work in a 30-hour week: six hours that do not exist. The cell turns red and prints OVER. Nothing about the person’s skill or effort changes that; the week is simply too small.

Now suppose the demand row for that week says 100 hours and the team has booked 94. The Gap row prints −6 and the word short. Two separate failures, one cell and one column, found by the same reading. Fixing the first by moving six hours to a colleague with free hours may also fix the second, or may not. Move it, then read again.

Failure modes

Where audits go wrong.

  • Planning every week at 100%. A plan with no slack is a plan that fails the first time anything unplanned arrives. Decide your own ceiling, and use it consistently.
  • Counting calendar hours. A 40-hour contract is not 40 hours of project time. Subtract first, then enter.
  • Forgetting the work that is not a project. Support, hiring, reviews, training and admin all occupy a week. They belong in the cell.
  • Averaging. An average over a quarter can read comfortable while three consecutive weeks are impossible. Read columns, not totals.
  • Treating part-time as full-time with a smaller number. A part-time row has its own capacity. Give it one rather than reducing every cell by hand.
  • Auditing once. Capacity and demand both move. The audit is only as good as the date on it.

Run it in the plannerPick a template