Managing for Daily Improvement

MDI stands for Managing for Daily Improvement. Most people reduce it to “daily standup meetings.” It is not. MDI is a management system with three interlocking components:

1. Visual management systems — boards, charts, and indicators that make performance visible to everyone, not just managers. If you have to ask someone how the line is running, your visual management is broken.
2. Daily accountability — structured tier meetings where yesterday’s performance is reviewed, today’s issues are surfaced, and action items are assigned with bilateral accountability. Not a status report. A commitment mechanism.
3. Leadership standard work — structured daily, weekly, and monthly tasks that ensure leaders are in the gemba, verifying that the system is functioning and strategic direction is reaching the floor.

Remove any one of these three and MDI does not work. Visual management without accountability produces pretty boards nobody acts on. Accountability without visual management produces meetings without data. Leadership standard work without the other two produces management activity without connection to reality.

Why MDI is the backbone

Hoshin kanri sets the strategy. Kaizen executes improvement. 5S shapes the environment. But none of these sustain themselves without a daily management system that connects them. MDI is the connective tissue. It is the system that ensures:

• Strategic objectives translate to daily actions
• Problems surface before they compound
• Resources flow to where they are needed
• Improvements, once made, do not silently decay

Without MDI, improvement is episodic. You run a kaizen event, get excited about the results, and watch them erode over the following months because nobody is tracking daily performance against the new standard. MDI prevents that erosion.

MDI operates through a tiered structure. Each tier has a specific scope, cadence, and purpose. Information flows upward as escalation and downward as resource deployment. Both directions are equally important.

What each tier resolves

• Tier 1 resolves problems within the team’s authority and capability. Operator reassignment, immediate quality containment, safety near-miss response. If the team lead can fix it today with available resources, it stays at Tier 1.
• Tier 2 resolves problems that cross team boundaries or require resources the team lead cannot authorize. Maintenance requests, cross-training needs, material supply issues, shift-to-shift handoff problems. Area managers coordinate across teams.
• Tier 3 resolves system-level issues: capital expenditure requests, supplier changes, cross-value-stream conflicts, hoshin project barriers. Issues that reach Tier 3 typically affect multiple areas and require management decision.

The escalation contract

Each tier has an implicit contract with the tier below: if an issue is escalated to me, I will either resolve it or explain why I cannot. The failure mode is silence — issues escalated upward that disappear into a management vacuum. If Tier 1 escalates a maintenance issue to Tier 2 and hears nothing for three days, the MDI system is broken regardless of how nice the boards look.

Visual management means the state of the process is visible to anyone who looks. Not visible in a software dashboard that requires login and navigation. Visible on the floor, at the point of work, to the person doing the work. A visual management system that only managers look at is a reporting system, not a management system.

The SQDC board

Most Tier 1 boards organize around Safety, Quality, Delivery, Cost. Each category has a daily indicator: green (on target), red (off target). The power of SQDC is not the categories — it is the forcing function. Every day, the team reviews all four dimensions and addresses red conditions immediately.

What makes visual management work

• Real-time or near-real-time data. Yesterday’s data is acceptable for daily review. Last week’s data is stale. The board should reflect what happened in the last 24 hours.
• Owned by the team, not by management. The team lead updates the board. The operators read it. If management maintains the board and operators ignore it, the visual management is theater.
• Action-oriented, not decorative. A red indicator must trigger an action. If Quality is red three days in a row and nobody has assigned a countermeasure, the board is decoration. The board should have a space for actions: what are we doing about the red condition, who owns it, when will it be resolved.
• Standard format across areas. Every Tier 1 board in the plant should follow the same layout. When a manager walks from Line 1 to Line 4, they should not have to relearn how to read the board. Standardization enables pattern recognition across areas.

Hour-by-hour boards

For high-volume production, SQDC is supplemented by hour-by-hour (or pitch-by-pitch) tracking. Each hour: planned output, actual output, delta, reason for gap. This is the most granular layer of visual management. It catches problems in the hour they occur, not at the end of the shift.

Hour-by-hour boards require discipline. Someone must update them every hour. If they go blank for three hours, the system is failing. The consistency of the updates tells you more about the culture than any survey.

In MDI, “newspapers” are action registers. They are called newspapers because in the original TPS implementations, they were printed on large-format paper and posted visually. But the format is not the point. The mechanism is.

A newspaper item is not a task assignment. It is a bilateral contract. There are two parties and both have obligations:

The bilateral contract

Read the third row. Williams agreed to cross-train Chen by March 20. But Williams needs a backfill operator for three days, and Chen is on vacation. If the area manager does not assign a backfill, Williams cannot deliver. When March 20 arrives and the cross-training is not complete, the question is not “why didn’t Williams do it?” The question is “did the area manager provide the backfill?”

This is the fundamental shift in accountability. The newspaper makes both sides of the contract visible. The person doing the work commits to a deliverable. The leader commits to providing resources. If either side fails, it is visible on the board.

What makes a good newspaper item

• Specific deliverable, not vague activity. “Improve changeover on Press 4” is not an action item. “Fabricate SMED shadow board for Press 4 tools” is.
• Named owner, not a department. “Maintenance” does not own action items. Tony Martinez in maintenance does.
• Date, not “ASAP.” Every item has a date. “When I get to it” is not a commitment. March 15 is.
• Resource requirement stated explicitly. What does the owner need that they cannot provide themselves? If nothing, fine. If something, it goes on the newspaper so it is visible and trackable.
• Barrier field used actively. When an item is blocked, the barrier field says why. This is not an excuse — it is information that the next tier needs to resolve the block.

Newspaper items that exceed local authority

Some newspaper items cannot be resolved at the tier where they originate. The SMED shadow board costs $2,000 in fabrication time. The team lead does not have that budget authority. This item escalates to Tier 2 or Tier 3. If it aligns to a hoshin objective, it becomes a hoshin project. The newspaper item does not disappear — it promotes. And the hoshin project tracks it to completion with monthly savings verification.

The Tier 1 meeting is the foundation of MDI. Get this right and the rest works. Get this wrong and every tier above it operates on bad information.

Structure

Rules

• Stand up. Sitting invites long conversations. Standing keeps it to 10-15 minutes.
• At the board. Not in an office. Not in a conference room. At the visual management board where the data is visible.
• Same time, every day. No exceptions. Consistency is the signal that management takes this seriously. Skip one day and you are telling the team it is optional.
• No problem-solving. If a topic requires more than two minutes of discussion, take it offline. Write the action item. Assign an owner. Move on. The Tier 1 meeting surfaces issues — it does not solve them.
• Leader facilitates, does not lecture. The team lead asks questions and documents answers. The operators share information. If the leader talks for 10 of the 12 minutes, the meeting is a report-out, not a review.

Common Tier 1 failures

• Too long. If Tier 1 regularly exceeds 15 minutes, you are solving problems instead of surfacing them. Stop.
• Data not current. If the board shows yesterday’s data on Thursday morning but the last update was Tuesday, nobody trusts it. Data freshness is non-negotiable.
• No actions assigned. A red condition discussed but not assigned to anyone is a complaint, not management. Every red condition needs: what, who, when.
• Escalation not followed up. Tier 1 escalates an issue to Tier 2. Two days later, no response. The team stops escalating because “nothing happens anyway.” This kills MDI.

Tier 2 is the coordination layer. Area managers meet daily to resolve cross-functional issues, allocate resources, and ensure escalations from Tier 1 are being addressed. This meeting is harder than Tier 1 because it requires managers to negotiate resources across competing priorities.

Structure

The resource conversation

Tier 2 is where the resource conversation happens. Maintenance has a backlog of 40 requests. Production has three lines running. Quality has two investigations open. The area manager must decide: which maintenance request takes priority? Does Line 2’s PM get done today or does Line 4’s breakdown take precedence?

These decisions are explicit at Tier 2. In a weak MDI system, they are implicit — whoever shouts loudest gets maintenance first. In a strong MDI system, the decisions are visible on the board, documented, and aligned to the priorities set by hoshin.

Tier 2’s obligation to Tier 1

When Tier 1 escalates an issue, Tier 2 owes a response. Not necessarily a resolution — sometimes the answer is “we cannot address this until next week because of X.” But the response must come. Silence kills escalation. If Tier 1 escalates three issues and hears nothing on two of them, Tier 1 learns that escalation is pointless. Once that lesson takes hold, MDI collapses. Issues get buried at the frontline because “management does not care anyway.”

MDI is a system. But how you employ the system determines whether it builds trust or destroys it. The same tier meeting structure can produce accountability and engagement or surveillance and resentment. The difference is not the structure — it is the intent behind it.

The trust progression

Building trust through the newspaper

Trust is not built by telling people to trust you. It is built by honoring commitments. The newspaper makes commitments visible. When a manager commits to providing a backfill operator by Wednesday and the backfill arrives on Wednesday, trust increments. When the backfill does not arrive and the manager says nothing, trust decrements.

This is why the “resource needed” column matters. It is not bureaucracy — it is the leader’s commitment made visible alongside the operator’s commitment. When both sides deliver, the system builds trust. When one side consistently fails, the newspaper shows it objectively.

How you employ it matters more than how it looks

A beautifully designed SQDC board with color-coded magnets and laminated headers is worthless if the team lead uses the meeting to lecture operators about performance. A hand-drawn board on butcher paper that operators actually own and update themselves is infinitely more valuable.

The test is not “does the board look professional?” The test is: “If I stand here for 30 minutes, will three different people walk up and look at it without being asked?” If yes, your visual management is working. If no, it is decoration.

Preventing culmination

This is the Clausewitzian concept that Toyota calls “respect for people.” Culmination is when an attacking force exhausts itself beyond recovery. In operations: if you push MDI so hard that people feel surveilled rather than supported, the system produces compliance but destroys engagement. If you use the newspaper to blame people instead of resolving barriers, you teach people never to put items on the newspaper.

Respect for people in MDI means: when someone raises an issue, you resolve it. When someone misses a due date, you ask about barriers before assigning blame. When the system produces a red indicator, the first question is “what do you need?” not “why did you fail?”

MDI without hoshin is activity without direction. Hoshin without MDI is strategy without execution. They must be connected, and the connection has specific mechanism:

Four connection points

1. Newspaper items that exceed local authority escalate to hoshin. A Tier 1 item that requires $15K for tooling is too big for the area manager. It becomes a hoshin project with a kaizen charter, savings calculation, and monthly tracking. The newspaper item does not vanish — it promotes.
2. KPIs are shared. The OEE number on the Tier 1 SQDC board is the same number on the hoshin X-matrix east quadrant. One data source, two views. If hoshin shows OEE at 85% and the Tier 1 board shows 72%, someone is lying. Shared data prevents this.
3. Hoshin action items appear in MDI newspapers. When a hoshin project generates an action item (“install shadow board on Press 4”), that item appears in the relevant team’s newspaper. The team lead tracks it alongside other daily items. It is visible in both the hoshin project view and the team’s daily newspaper.
4. Leadership standard work bridges the gap. The plant manager’s weekly LSW includes: “Review hoshin project status. Walk to Tier 1 boards for active hoshin areas. Verify KPIs match.” Without LSW, the connection between hoshin and MDI depends on individual discipline. With LSW, it is a structured verification built into the management routine.

When to promote from MDI to hoshin

• The improvement requires cross-functional resources that the area manager cannot authorize
• The improvement aligns directly to a strategic or annual objective on the X-matrix
• The expected savings exceed a threshold (typically $25K+, varies by organization)
• The improvement requires a kaizen event with dedicated team time and formal charter

Not everything belongs on the X-matrix. Most MDI newspaper items are local improvements handled within the team or area. Only those that require strategic-level resources or align to breakthrough objectives should promote. The X-matrix should have three to five projects per site, not thirty.