Where to Start Improvements After Completing a Value Stream Map? —— A Five-Step Method for Marking and Prioritizing Improvement Burst Points

1. After Hanging the Map on the Wall, Improvements Get Stuck

Many teams, after completing the current state value stream map, excitedly hang it in the meeting room, and then nothing happens. The map is filled with dozens of data frames, inventory triangles, waiting times, rework loops, and bottleneck processes, all pointing to waste. It seems like everything needs to be improved. However, when it comes to planning the improvements, it often turns into a situation where "whoever shouts the loudest gets improved first" or "whichever area the leader is currently focusing on gets improved." As the improvements proceed, they revert to departmental optimizations, rendering the entire map useless.

A common scenario is the map being covered with sticky notes, each marking a "problem," but no one can clearly explain which one to tackle first. After three rounds of improvement weeks, the results are contradictory. The reason is that the current state map only answers "what is the current situation," but not "which areas to improve first and why." To transform vague improvement intentions into an ordered action list, a specific symbol on the value stream map—the improvement burst point (Kaizen Burst)—and a straightforward prioritization method are essential.

2. Improvement Burst Points: Marking the "Areas to Improve" on the Map

Improvement burst points are typically marked with a starburst symbol on the current state map, indicating areas that require concentrated improvement. This forces the team to clarify each waste: what the issue is, how much waste there is, and how much smoother the flow will be after improvement. A qualified burst point must have three elements: a clear location, quantified waste data, and a corresponding improvement direction. For example, "12 days of inventory in the raw material warehouse, amounting to approximately 800,000 yuan, with the direction being to shorten changeover time and reduce batch size" is a proper burst point, while simply writing "high inventory" is not.

The locations for marking are generally chosen from three areas: between data frames (inventory and waiting between processes), within data frames (bottleneck processes, high rework processes), and on the information flow (slow order processing, chaotic scheduling). Each point should be accompanied by a note: the problem, the data, and the initial improvement direction. Marking is not the final conclusion but a target for team discussions—after marking, the team should review each point collectively, and any point with unclear data should be remeasured.

3. Five-Step Method for Marking and Prioritizing

Step 1: Walk through the material flow segment by segment, circling all points of flow interruption. Start from the customer demand end and check for four types of issues at each data frame: Is there inventory buildup? Is there waiting? Is there a rework loop? Is there a bottleneck process (cycle time close to or exceeding takt time)? Conduct the walk-through with a stopwatch and a process list on-site; don't just rely on old reports in the meeting room—actual conditions on-site often differ from reports. Use a pencil to circle candidate points; this round is about completeness, not filtering.

Step 2: Supplement data for each candidate point, marking the waste amount. For inventory points, mark the inventory days and amount; for waiting points, mark the waiting time; for rework points, mark the rework rate and rework hours; for bottleneck points, mark the difference between cycle time and takt time. Each point should be registered in a uniform format: location, type of waste, data, data source, and measurement date. Points with insufficient data should be marked as "pending measurement" and a special measurement should be arranged—never rely on feelings for scoring. The quality of the prioritization depends on the quality of the data, and a lack of uniformity makes it impossible to prioritize.

Step 3: Prioritize based on "Impact × Implementation Difficulty." The horizontal axis represents the impact on delivery time, quality, and cost after improvement, while the vertical axis represents implementation difficulty. Use a 1 to 5 scoring system: for impact, score each of the three dimensions (delivery, quality, cost) from 1 to 5 and take the average; for difficulty, score based on technical challenges, cross-departmental coordination, and investment. Prioritize points in the upper right quadrant (high impact, low difficulty); break down points in the upper left quadrant (high impact, high difficulty) into smaller tasks and proceed step by step; address points in the lower right quadrant (low impact, low difficulty) when there is spare capacity; and discard points in the lower left quadrant. The scorers must include on-site team leaders, who are most familiar with the implementation difficulty. This matrix effectively suppresses random prioritization based on impressions or personal relationships.

Step 4: Calibrate with the target state map. For each candidate burst point, ask: Is it a point that needs to be addressed in the future state map? If a point is not on the improvement path of the target state, even if it has a significant current impact, it should be approached with caution, as it might be a local optimization. If the target state map has not been drawn, complete it before prioritizing—prioritizing burst points without a target state is essentially guesswork. Ensure that each burst point on the list corresponds to the target state map to keep improvements on track.

Step 5: Control the number, form an improvement plan. Select 3 to 5 burst points for each improvement cycle, clearly defining quantified goals, responsible persons, completion deadlines, and required resources for each point. The remaining points should enter the "improvement pool" and be reviewed periodically: if the environment changes or data is updated, priorities can be adjusted, but points currently being addressed should remain unchanged to avoid frequent changes. Marking a dozen burst points and filling the plan with tasks means there is no clear priority—this is the most common reason for improvement plans to fail.

4. An Example: Start with Leverage Points, and Subsequent Tasks Become Easier

A machining workshop of an automotive parts company completed its current state map and identified 8 candidate burst points: 12 days of inventory in the raw material warehouse, 3 days of waiting between two processes, an 8% rework rate after heat treatment, a bottleneck process with a cycle time 15% over the takt time, etc. After prioritizing using the two-dimensional matrix, the top-ranked point was not the rework rate, which the leadership was most concerned about, but the "45-minute changeover time" point, which had high impact and low difficulty—because long changeover times led to large batch sizes, which in turn increased inventory and waiting times. The team first tackled quick changeover, reducing the changeover time to 12 minutes, halving the batch size, and significantly decreasing inventory and waiting times. The delivery cycle was shortened from 18 days to 11 days, and the rework task became easier due to the additional time buffer. Starting with leverage points makes subsequent tasks naturally simpler, demonstrating the value of prioritization.

5. Conclusion

The true output of a value stream map is not the map itself but the improvement burst points on the map and an ordered improvement plan. Being able to draw the map is just the entry point; being able to mark and prioritize is the key skill that makes value stream management truly effective.

Improvement burst points transform the value stream map from a "diagnosis report" into an "action map." Starting with high-impact, low-difficulty leverage points ensures that improvements are sustainable.

Knowledge code: 7.1.1

Version: v20260816

Author: Quality Think Tank Quality Think Tank is dedicated to providing systematic knowledge, methodologies, and practical tools for quality management professionals, helping enterprises continuously enhance their quality capabilities.