How to Improve the Efficiency of the Board of Directors of CNC Gong

• 1
  Five things that move the needle first
• 2
  Why a CNC gong board meeting runs long and decides little
• 3
  Split the cnc gong board of directors into three working roles
• 4
  Design the agenda around decisions, not reports
• 5
  Give the board four numbers it can act on
• 6
  Five steps to run the next meeting differently
• 7
  When to bring a job to the board and when to keep it at plant level

Most board meetings in a machining business fail for one boring reason: the agenda is built from whatever arrived in the inbox that week. A gong shop quotes maybe 60 to 200 jobs a month, and each one touches programming, material, fixturing, and finishing. If every open quote reaches the board table, the meeting becomes a quoting session and nothing strategic gets decided.

The second cause is pack length. When the pre-read is 40 pages, directors read it in the room. That turns a 90-minute meeting into a 150-minute one and pushes the capex decision to next quarter. We keep our own pack at 12 pages and send it 72 hours before the meeting, which is enough time to read and flag questions but not enough time to rewrite the whole thing.

The third cause is missing data. A board cannot decide whether to buy a 5-axis center if nobody brings machine uptime and backlog hours. In our plants we track four numbers weekly: on-time delivery, first-pass yield, scrap value, and spindle uptime. Those four fit on one page and answer most capex questions before they are asked.

None of this is about the people in the room. It is about what reaches the room. Fix the input and the meeting shortens on its own.

• 1
  Quotes crowd out strategyRoute jobs under the value threshold to a plant-level approval, not the board.
• 2
  Long packs kill pre-reading12 pages, 72 hours ahead, questions submitted in writing.
• 3
  No data, no decisionFour weekly numbers on one page beat 20 slides of narrative.

A board of five to seven people does not need five generalists. It needs three functions covered: commercial, technical, and financial. In a gong shop the technical seat matters more than most owners expect, because tolerance capability and machine mix decide which orders you can accept. Whoever holds that seat should be able to read a drawing and a capability sheet.

Give each seat a written scope. The commercial seat owns pipeline and customer terms. The technical seat owns process capability, machine investment, and quality escapes. The financial seat owns cash, credit terms, and margin by job family. Write these on one page and review them once a year.

Overlap is where meetings go wrong. If two directors both feel responsible for quoting, both will bring quote issues and neither will own the outcome. Name one owner per decision type, in writing, and let the others advise rather than decide.

We also keep one seat for an outside view. It does not have to be a paid advisor. A retired production manager from another industry will ask better questions about cycle time than a room of owners will.

• 1
  One owner per decision typeWritten on one page, reviewed annually.
• 2
  Technical seat reads drawingsCapability and machine mix decide which orders you accept.
• 3
  Outside seat asks harder questionsCycle time and yield questions usually come from outside the industry.

Build the agenda backwards from the decisions you need. List every decision due this quarter, group them by owner, and give each group a timebox. A typical 50-minute board meeting in a gong shop looks like this: 5 minutes to confirm the previous log, 15 minutes on open commercial items, 20 minutes on quote and capex approvals, 5 minutes on risk, 5 minutes to close the log.

Reports go in the pre-read, not the meeting. If a director wants to discuss a report, they submit a written question 24 hours ahead and the answer goes into the pack. That single rule removed about 30 minutes from our own quarterly meetings.

Keep a standing quote-review item. Jobs above the value threshold come to the board with three numbers attached: quoted price, estimated cycle time, and material cost at the current index. Without those three, the board is guessing. With them, a go or no-go takes two minutes.

End every meeting by reading the decision log aloud. It takes 5 minutes and it catches the items that were discussed but never closed.

• 1
  Reports in the packWritten questions 24 hours ahead, answers added to the pack.
• 2
  Three numbers per quoteQuoted price, estimated cycle time, material cost at index.
• 3
  Read the log aloudFive minutes at the end, every meeting, no exceptions.

Directors cannot manage what arrives as a narrative. Four numbers, updated weekly, cover most of what a machining board needs. On-time delivery tells you whether the schedule is honest. First-pass yield tells you whether the process is stable. Scrap value in currency tells you whether the yield problem is expensive. Spindle uptime tells you whether you need machines or people.

Set a target band for each number, not a single target. For on-time delivery we work to a band rather than a point, because a shop that hits 100 percent by padding lead times is not actually efficient. A band of 95 to 98 percent with a stable backlog is healthier than 100 percent with a shrinking order book.

Review the trend, not the week. One bad week is noise. Three bad weeks is a process problem and belongs on the agenda as a decision item, not a report item.

If a number moves outside its band twice in a row, the responsible director brings a one-page cause and a proposed action. No slide, one page.

• 1
  On-time deliveryTarget band, not a single point. Watch the trend across three weeks.
• 2
  First-pass yieldStable process shows a flat line, not a rising one.
• 3
  Scrap valueCurrency, not percentage. It ranks problems by cost.
• 4
  Spindle uptimeSeparates a capacity problem from a staffing problem.