ROI Case File No.543: They Were About to Decide Where to Start by Gut Feeling
They Were About to Decide Where to Start by Gut Feeling
Chapter 1: We Want to Unify, but Can't Decide Where to Start
"We'd like you to handle unified management of our operations and accompany us through system adoption. But we can't set priorities on where to start."
Kazuhiko Hasumi, head of the business planning office at Redicul, said this as he described the situation. "Our multiple systems can't link up, and nearly all our operations are manual. Order-form issuance, accounting software—each is installed individually, yet there's zero linkage between systems. We rely on manual work using Excel."
"Which department carries the heaviest load?" Claude asked.
"Accounting," Hasumi answered. "They process about three hundred order-form issuances and billing tasks a month by hand. The burden on the clerical staff is large, and oversights and mistakes happen constantly. Data management is scattered department by department, and information is siloed. Personnel dependency is serious too."
"How are you trying to decide the starting order?" I confirmed.
"Honestly, by gut feeling," Hasumi answered. "We have a policy to unify by 2029. But where to start is about to be decided by the say of the loudest department. We can't measure in numbers what works best."
"If you're deciding the starting order by gut feeling, then we need to quantify the dissatisfaction and judge from it," I responded. "Let's measure with NPS."
Chapter 2: What NPS Asks—The Bottleneck Surfaced by Recommendation Score
"This case needs NPS."
Claude wrote "NPS" on the whiteboard.
"NPS—Net Promoter Score, the recommendation score—is originally an indicator that measures, on a scale, whether a customer would recommend something to others," I explained. "Turned inward at the company, it's powerful. Ask each department's employees, 'Could you confidently recommend the current workflow to someone?' and measure it on a scale. Not by loudness but by the score of dissatisfaction, you surface the bottleneck. It's a tool that judges, by recommendation score, the starting order you'd been deciding by gut feeling."
"First, let's measure the current cost," Gemini said, opening the ROI Polygraph. He entered the data Hasumi had provided.
"The monthly cost is in," Gemini read out. "Manual processing labor for order-form issuance and billing: 200 hours per month on average, at ¥3,500 per hour, ¥700,000 per month. Double-entry labor from non-linked systems: ¥480,000 per month on average. Rework and losses from oversights and mistakes: ¥450,000 per month on average. Expected value of handover risk from departmental silos and personnel dependency: ¥380,000 per month on average. Management labor from start stalling due to absent priorities: ¥300,000 per month on average. A total of ¥2,310,000 per month. Annualized, about ¥27,720,000."
Hasumi stared at the figures. "I thought it was just the accounting handwork. Add the double entry, and the cost of priorities not getting decided in the first place, and it's this much?"
"Now, let's design with NPS," I continued.
[Measuring Internal NPS—Turning Dissatisfaction Into a Score]
"First, we ask each department's employees for a recommendation score," Claude said. "We gather 'could you confidently recommend the current work?' as a score. Accounting will come out conspicuously low in numbers. We pinpoint the bottleneck not by gut feeling but by score."
[Identifying the Bottleneck—Choose the Single Lowest Point]
"Next, we choose the work with the lowest score," Gemini continued. "The lowest recommendation score is accounting, running three hundred items a month by hand. Not loudness but the score points to where we start first."
[Setting the Starting Order—Line Them Up in the Order That Works]
"From the identified bottleneck, we build the starting order," I continued. "Accounting automation first, then inter-system linkage, and finally cross-department data sharing. Line them up from the lowest score, and priorities shift from gut feeling to numbers."
[Re-measuring—Track the Effect of the Moves by Score]
"Finally, after adoption, we measure NPS once more," Claude continued. "We confirm the effect by whether the score rose, and decide the next target. Measure recommendation score continuously and improvement becomes a structure that keeps going."
[Calculating the Investment Recovery]
"Let's run the estimate with the ROI Proposal Generator," Gemini proposed.
- Initial cost: Internal NPS measurement design, expense-settlement/accounting automation system adoption, core-system linkage, cross-department data-sharing infrastructure, and training—¥5,400,000 total
- Monthly cost: System operation and update ongoing fees combined, ¥220,000 per month
- Monthly reduction effect: Automation of order-form and billing work = ¥560,000 per month (assuming 80% reduction), resolution of double entry = ¥400,000 per month, reduction of oversight and mistake rework = ¥360,000 per month, linkage effect from silo resolution = ¥280,000 per month, totaling ¥1,600,000 per month
- Net monthly reduction: ¥1,600,000 − ¥220,000 = ¥1,380,000 per month
- Payback period: ¥5,400,000 ÷ ¥1,380,000 = approximately 3.9 months
"Recovery in just under four months," Gemini summarized. "What works is starting from accounting, where the recommendation score is lowest. Rather than linking all systems at once, we crush from the single lowest-scoring point. Because we re-measure the effect with NPS as we expand, we avoid the failure of starting by gut feeling and missing."
Hasumi confirmed the figures. "We were about to decide 'from the loudest department.' Measure by score, and the truly heavy place becomes visible. We can decide from there."
"NPS is a tool that turns dissatisfaction into a score and judges the starting order," I responded.
Chapter 3: A Deployment Plan That Crushes From the Lowest-Scoring Point
"Let me organize the approach," I said, standing at the whiteboard.
"Month one—internal NPS measurement, gathering each department's recommendation score. Month two—identifying the bottleneck, fixing accounting as the first target. Months three and four—adopting the accounting-automation system. Month five—strengthening linkage with the core system. Month six—re-measuring post-adoption NPS, effect verification. Month seven onward—rolling out cross-department data sharing, phased expansion toward the 2029 unification."
"Will the 2029 unification make it in time?" Hasumi confirmed.
"It will," Claude responded. "Aim for unification all at once and everything ends up half-done. Crush from low-scoring accounting, confirm the effect with NPS, and move to the next. Crushing steadily from the heavy single point gets you to unification far faster than reaching for everything by gut feeling."
Hasumi said, taking notes, "Turn dissatisfaction into a score and crush from the lowest. I can see how to decide the starting order now."
Chapter 4: The Day the Starting Order Was Decided by Numbers
Nine months later, a report arrived from Hasumi.
Accounting's manual processing was reduced 80% from before after the automation system was introduced. "Three hundred order-form issuances and billings a month now run almost automatically. The state where clerical staff process while dreading oversights has vanished," Hasumi wrote.
Double entry also dropped sharply. With inter-system linkage advanced, the work of re-keying the same data over and over disappeared. "Enter on one side and it reflects on the other. The waste of typing twice vanished," the report read.
The biggest change appeared in how the starting order was decided. Priorities, nearly set by gut feeling, were fixed by numbers. "What had stalled on 'where to start' got decided by the NPS score. We came to move not by loudness but by score," Hasumi wrote.
Personnel dependency also headed toward resolution. With siloed data now shared, the state of depending on a specific person thinned. "'Only that person knows' decreased. Handovers stopped being frightening," the report read.
As a secondary effect, it became a structure where improvement continues. By measuring NPS periodically, the next place to crush kept becoming visible. "It doesn't end with a one-time improvement. Measure the score, and where to fix next stays visible," Hasumi wrote.
At the end of Hasumi's report it said: "Unification was stalled on where to start. The moment we turned dissatisfaction into a score with NPS, the heaviest single point got decided. Starting order can be judged by score, not gut feeling."
The day a company about to decide the starting order by gut feeling became a company that could judge by score, system adoption had changed from a matter decided by loudness into a design that judges the starting order by recommendation score, the report noted.
"Consultations on unification and system adoption usually stall on 'where to start.' There's a policy, but priorities can't be set, and it's about to be decided by the loudest department's say. What NPS asks is a score called recommendation. Ask each department 'could you recommend the current work to someone?' and dissatisfaction surfaces as a score. The day a company about to decide the starting order by gut feeling could judge by score, what changed was not the system but the very perspective that turns dissatisfaction into a score and decides priorities."
Related Files
Tools Used
- ROI Polygraph — Visualizing accounting handwork labor, double entry, and start-stalling cost
- ROI Proposal Generator — Investment-recovery simulation for recommendation-score-rooted operations unification
