ROI Case File No.521: The Ten-Year-Old Purchasing Ledger Was a Sanctuary No One Dared Touch
The Ten-Year-Old Purchasing Ledger Was a Sanctuary No One Dared Touch
Chapter 1: No One Knows Where to Begin
"We have four systems running, all disconnected. No one can decide which one to tackle first."
In the business planning office at GlobalTech, division head Makoto Ujihara said this while spreading out a list of the company's internal systems. Sales management, inventory management, accounting, and a purchasing management system built from scratch more than ten years ago. "Each one runs independently. Reflecting sales results in inventory, handing the numbers over to accounting—it's all manual re-entry."
"What concrete harm has the lack of unified management caused?" Claude asked.
"Inventory consistency breaks down," Ujihara answered. "The sales-side numbers and the inventory-side numbers don't match, and reconciling them at month-end takes a full two days. What's more troublesome is the ten-year-old purchasing management system. Only a specific department uses it, and the person who built it has already left the company. No one will touch the inside of it. It's become a sanctuary."
"What is the structure of the business planning office?" I confirmed.
"It's just me," Ujihara answered. "And my specialized knowledge is limited. There was a period when we brought in consultants, but business declined and we cut them as a cost-reduction measure. Now the need has come up again, and we're at the stage of hearing proposals from several firms. Honestly, though, I can't establish a priority order for what to address first."
"You're in a situation where there's too much to do, so the order won't settle," I responded. "Let's quantify the priorities with ICE."
Chapter 2: ICE Asks About Impact, Confidence, and Ease
"This case needs ICE."
Claude wrote three letters on the whiteboard: "I, C, E."
"ICE is a framework that scores measures along three axes—Impact, Confidence, and Ease—to decide priorities," I explained. "It's a technique used in the growth-hacking world, but its essence is judging, by numbers, a situation where there's too much to do and the order won't settle. In what order do we integrate the four systems? We decide by score, not by gut."
"First, let's measure the current cost," Gemini said, opening ROI Polygraph. He entered the data Ujihara had provided.
"The monthly cost of fragmentation came out," Gemini read aloud. "Manual re-entry labor between systems averages 280 hours a month; at an hourly rate of 3,800 yen, that's 1,064,000 yen a month. The month-end inventory consistency reconciliation work averages 48 hours a month, or 182,000 yen. Losses from mis-ordering and stockouts caused by data inconsistency average 900,000 yen a month. The expected value of the personnel-dependency risk from the purchasing system's black-box state averages 700,000 yen a month. The opportunity loss from delayed decision-making, because the numbers management needs don't come out instantly, averages 600,000 yen a month. The duplicated cost of maintaining four separate systems averages 550,000 yen a month. The total is 3,896,000 yen a month. Annualized, that's about 46.75 million yen."
Ujihara stared at the figures. "I thought it was only the re-entry labor. When you put a price on the black box's personnel-dependency risk and the delayed decisions, the scale is completely different."
"Then let's design with ICE," I continued.
[Impact—The Size of the Effect Integration Brings]
"First, we score each measure's Impact," Claude said. "Integrating sales management and inventory management has the highest impact. It directly improves inventory turnover and directly reduces losses from mis-ordering and stockouts. Nine out of ten. Accounting linkage scores eight; replacing the purchasing management system scores seven for its mid-to-long-term effect. Impact is the measure of 'how much will fixing this help.'"
[Confidence—The Likelihood of Success]
"Next is Confidence," Gemini continued. "We score the likelihood of success from past integration track records and the quality of the existing data. The sales-inventory integration has the data in place, so confidence is nine. Accounting linkage has clear specifications, so eight. The purchasing management system, because its contents are a black box, scores a low four. Confidence is the measure of 'can we truly say it will succeed.'"
[Ease—How Easy It Is to Execute]
"Last is Ease," I continued. "Using a cloud-based unified management system, the linkage of sales, inventory, and accounting scores an ease of eight. It can be achieved with standard API integration. The purchasing management system, on the other hand, has no designer present and no documentation, so its ease is two. If we tackle it, it's the heaviest, and the numbers indicate we should leave it for last."
[Multiplying the Three Axes to Decide the Order of Attack]
"When you multiply Impact, Confidence, and Ease, the order of attack comes out automatically," Claude continued. "First priority: sales-inventory integration (9 × 9 × 8). Second: accounting linkage (8 × 8 × 8). The purchasing management system has effect, but low confidence and low ease, so it's last. Not cutting into the sanctuary first becomes the numerically correct judgment."
[Estimating the Investment Recovery]
"Let's estimate with ROI Proposal Generator," Gemini proposed.
- Initial cost: Cloud-based unified management system deployment, sales-inventory integration, accounting linkage, data migration, phased-migration design, and internal staff training—9,800,000 yen total
- Monthly cost: System usage fee plus ongoing operating cost—320,000 yen a month combined
- Monthly reduction effect: Re-entry labor reduction = 850,000 yen a month (assuming 80% reduction), inventory consistency work reduction = 160,000 yen a month, mis-order/stockout loss reduction = 630,000 yen a month, elimination of decision-making delay = 420,000 yen a month, duplicate maintenance reduction = 400,000 yen a month—2,460,000 yen a month total
- Monthly net reduction: 2,460,000 yen − 320,000 yen = 2,140,000 yen a month
- Payback period: 9,800,000 yen ÷ 2,140,000 yen = about 4.6 months
"Recovery within five months," Gemini summarized. "What matters is not touching the sanctuary—the purchasing management system—first. The ICE scores show the order in which to start from the integration that is most effective and easiest to execute. By leaving the low-confidence target for later, the early success experience comes out quickly."
Ujihara confirmed the figures and said, "Whenever we tried to integrate everything at once, the discussion always stopped at the wall of the purchasing management system. When you score on three axes, the order of what to touch becomes visible."
"ICE is a tool for assigning numbers to the queue of things you should do," I responded.
Chapter 3: A Deployment Plan That Proceeds by Phased Integration
"Let me organize how we'll proceed," I said, standing before the whiteboard.
"Months one and two—selecting the cloud-based unified management system and designing the data structures for sales and inventory management. Months three and four—building and trial-running the sales-inventory integration and implementing automatic inventory-consistency reconciliation. Month five—building the linkage with the accounting system and automating monthly aggregation. Month six—putting the integrated system into production and transferring operations to internal staff. Month seven—analyzing the current state of the purchasing management system and beginning to document the black box's specifications. Month eight onward—phased migration of purchasing data to the integration platform and establishment of an in-house operating structure."
"So in the end, what happens to the purchasing management system?" Ujihara confirmed.
"We don't replace it abruptly," Claude responded. "First we analyze its contents and document the specifications. Then we migrate the data to the integration platform in phases. A target ICE flagged as low-confidence shouldn't be rushed for effect; we lower the risk before tackling it. If you rush to destroy a sanctuary, there's a danger of bringing operations to a halt."
Taking notes, Ujihara said, "I'd never had the idea of deciding priorities with numbers. Even with a one-person structure, once the order is set, I can move."
Chapter 4: A Queue of Improvements, Now Numbered
Nine months later, a report arrived from Ujihara.
The sales-inventory integration cut 80% of manual re-entry three months after going live. The month-end inventory reconciliation shrank from a full two days to half a day. "Sales results are reflected in inventory in real time. The work of matching up numbers at month-end has nearly vanished," Ujihara wrote.
Mis-ordering and stockouts also fell sharply. With inventory quantities unified in the system, ordering errors stemming from data inconsistency were nearly eliminated. "The habit of over-ordering on the assumption that inventory wouldn't match has disappeared. We now run on appropriate stock levels," the report said.
The most unexpected change appeared in the speed of decision-making. The numbers management needed now came out instantly, and the quality of meetings changed. "Before, it was 'gather the numbers, then decide.' Now it's 'look at the numbers and decide on the spot,'" Ujihara wrote.
The sanctuary—the purchasing management system—also started to move. Current-state analysis began in month seven, and the logic that had been a black box for ten years was documented. "The contents of a system no one had touched since its designer left were put into words for the first time. That may have been the biggest gain," the report said.
As a secondary effect, the business planning office's one-person structure changed. With the integrated system's operation standardized, members on loan from other departments could now join in running it. "The knowledge that had been concentrated in me alone became something multiple people could share. The personnel dependency was resolved," Ujihara wrote.
His view of the past consulting investment that had been cut also changed. "Before, I saw consultants as 'an expense that gets cut when business is bad.' This time, because we locked down the investment recovery with numbers before starting, it's being treated internally as an 'investment,' not an 'expense,'" the report said.
At the end of Ujihara's report, he had written: "When there's too much to do, people can't decide the order, and in the end they do nothing. The moment we scored the three axes with ICE, the first move was decided. Priorities can be decided by calculation, not by debate."
It was recorded that a division head who had frozen before four systems had turned to days spent clearing a numbered queue one item at a time.
"The reason integration projects stall is usually 'trying to do everything at once.' You cut into the hardest sanctuary first, and the discussion stops there. What ICE asks is the three axes—impact, confidence, ease. Start from what has large effect, where success can be assured, and which is easy to execute. The low-confidence sanctuary you touch last, after lowering the risk. The day a division head who had been standing paralyzed before a purchasing ledger no one had touched in ten years could finally assign numbers to his queue of improvements, what changed was not the system configuration but the very order of decision-making."
Related Files
Tools Used
- ROI Polygraph — Visualizing re-entry labor, data-inconsistency losses, and black-box personnel-dependency risk
- ROI Proposal Generator — Investment-recovery simulation for unified system management through phased integration
