ROI Case File No.540: Only the DX Rallying Cry Echoed Down the Hall
Only the DX Rallying Cry Echoed Down the Hall
Chapter 1: There's a Company-Wide DX Cry, but No Starting Point
"Company-wide DX is our top priority this term. But where to start—no concrete plan comes from any department."
Kenichi Mikuriya, CEO of PharmaTech, spoke about the current state. A major pharmaceutical company. "I've issued the 'do DX' rallying cry. But all that comes back from the floor is the question 'where' and 'what works most,' and nothing concrete surfaces."
"What's the state of the current systems?" Claude asked.
"They only multiply," Mikuriya answered. "Each department brings in its own system, and they're not unified. There's a lot of individual-dependent work, and paperless is behind. The issues are piled high, yet no one can judge which to tackle first."
"Have you ever set DX priorities?" I confirmed.
"Never," Mikuriya answered. "The pharma industry has few success examples, so it's hard to picture concretely. That's why we're looking for a partner with AI-consulting expertise. Honestly, only the rallying cry echoes down the hall, and the first move never gets decided."
"If the issues are piled high and the starting point won't settle, you need to judge by impact, confidence, and ease," I responded. "Let's quantify the priority with ICE."
Chapter 2: ICE Asks Impact, Confidence, and Ease
"This case needs ICE."
Claude wrote "I・C・E" on the whiteboard.
"ICE is a framework that scores measures on three axes—Impact, Confidence, Ease—and decides priority," I explained. "It's a method used in growth hacking, but its essence is judging by numbers the situation where 'there's too much to do and the order won't settle.' A DX rallying cry alone won't move the floor. We decide by score which issue, tackled first, works, can be confident, and is easy to execute."
"Let's measure the current cost first," Gemini said, opening ROI Polygraph and entering the data Mikuriya had provided.
"The monthly cost is in," Gemini read out. "Inefficiency and handover effort from individual-dependent work averages 280 hours a month; at ¥4,600 an hour, that's ¥1,288,000 a month. Processing cost from paper operations and non-paperless work averages ¥900,000 a month. The cost of system proliferation and duplicated maintenance averages ¥750,000 a month. The management effort spinning its wheels for lack of DX priorities averages ¥650,000 a month. The expected value of technical-succession risk from individual dependence averages ¥550,000 a month. The total is ¥4,138,000 a month—roughly ¥49.66 million a year."
Mikuriya stared at the figures. "Just issuing the rallying cry, even the management effort spinning its wheels turns into money. The very absence of priorities was cutting this much every month."
"Then let's design it with ICE," I continued.
[Impact—Which Issue Works Most]
"First, we score each issue's Impact," Claude said. "Resolving individual-dependent work and going paperless directly affect company-wide productivity—high impact. Nine out of ten. System unification works over the medium-to-long term but its immediacy is mid—seven. Automation via AI and RPA, eight. Impact is the measure of 'how much it works if you fix this.'"
[Confidence—How Sure of Success]
"Next, Confidence," Gemini continued. "With few success examples in pharma, large-scale system unification scores lower in confidence—five. On the other hand, surfacing individual-dependent work and going paperless have a track record regardless of industry—confidence eight. AI/RPA deployment has a clear scope of application—seven. Confidence is the measure of 'can you assert it will truly go well.'"
[Ease—How Easy to Execute]
"Finally, Ease," I continued. "Surfacing individual-dependent work and going paperless are easy to start—eight. AI/RPA automation, if you choose the targets, ease seven. Unifying the proliferated systems is heavy to coordinate with many existing systems—ease four. If you tackle it, it's the heaviest, and the numbers show it should be put off."
[Multiply the Three Axes and Decide the Order of Attack]
"Multiply the three axes and the order of attack comes out," Claude continued. "First priority is surfacing individual-dependent work and going paperless (9×8×8). Effect can be felt in a relatively short period. Second is automation via AI and RPA (8×7×7), with system unification advanced in parallel alongside it. Not cutting into the low-confidence, low-ease system unification from the start is the numerically correct judgment."
[Estimating the Payback]
"Let's run the numbers with ROI Proposal Generator," Gemini proposed.
- Initial cost: surfacing individual-dependent work, building a paperless base, deploying AI/RPA automation, system-unification design, and company-wide training—¥8.6 million total
- Monthly cost: base operation, licenses, and ongoing updates combined—¥360,000 a month
- Monthly savings: individual-dependent work standardized = ¥770,000 a month (assuming 60% reduction); paperless = ¥620,000 a month; duplicated maintenance cut = ¥500,000 a month; spinning-wheels resolved = ¥420,000 a month—¥2,310,000 a month total
- Net monthly savings: ¥2,310,000 − ¥360,000 = ¥1,950,000 a month
- Payback period: ¥8.6 million ÷ ¥1,950,000 = about 4.4 months
"Payback in just under four and a half months," Gemini summarized. "What matters is not tackling the low-confidence system unification first. The ICE scores show the order: start from individual-dependence resolution and paperless, which work most and are easiest to execute. Produce a success experience in the short term and head into the heavy issues on that momentum."
Mikuriya said as he checked the figures, "I'd been issuing the rallying cry as one big lump called company-wide DX. Score the three axes, and which issue to tackle first becomes visible. Not by rallying cry but by calculation."
"ICE is a tool for numbering the queue of things to do," I responded.
Chapter 3: A DX Plan That Tackles the Working Issues First
"Let me lay out the approach," I said, standing at the whiteboard.
"Months one and two—ICE-score all issues and surface individual-dependent work. Months three and four—advance paperless and tackle the short-term-effect domains. Month five—deploy AI/RPA business automation. Month six—verify the automation effect and start designing system unification. Month seven—advance unification in parallel while expanding the automation scope. Month eight onward—stage-wise execution of system unification and embedding a company-wide DX structure."
"What becomes of system unification in the end?" Mikuriya confirmed.
"We don't touch it first," Claude responded. "Rush an issue scored low in confidence and ease, and the discussion stalls. First produce short-term success with individual-dependence resolution and paperless, and advance automation with AI/RPA. Once the path to unification becomes visible in that process, we tackle it stage by stage. Issues ICE scored low we don't rush for effect; we head into them after building momentum. This is the order the numbers show."
Mikuriya said as he took notes, "I thought it would move if I issued the rallying cry. Without showing in numbers where to start, the floor couldn't move."
Chapter 4: The Day the First Move Was Decided
Ten months later, a report arrived from Mikuriya.
Surfacing individual-dependent work and going paperless were tackled as top priority, with effect felt within months of going live. Individual-dependent work was standardized, and paper processing decreased. "Because we started from the short-term-effect domains, the floor could feel 'DX works.' The air changed from when it was just a rallying cry," Mikuriya wrote.
Automation via AI and RPA advanced too. Automating routine work further backed up the individual-dependence resolution. "The routine work done by hand was automated. The individually held work started running by mechanism," the report said.
The biggest change showed up in the DX entrance becoming visible. From a state where only the rallying cry echoed, the order of attack was settled by numbers. "What had stalled at 'where to start' got numbered by ICE. Once the first move was decided, the floor started moving," Mikuriya wrote.
The spinning of wheels was resolved too. Priorities settled, and the management cry connected to concrete action. "The repetition of issuing the cry and spinning idle ended. Show the priorities in numbers, and the floor stopped hesitating," the report said.
As a side effect, the path to system unification became visible. The heavy issue not touched first revealed its outline in the process of automation. "Had we cut into unification straight away, the discussion would have stalled. Putting it later, the path actually became visible," Mikuriya wrote.
At the end of Mikuriya's report, he had written this: "The company-wide DX rallying cry couldn't move the floor as a lump. The moment we scored the three axes with ICE, the first move was decided. Even with few examples in pharma, score by impact, confidence, and ease and the order of attack is decided by calculation."
The day a company where only the DX rallying cry echoed down the hall became one that can judge the order of attack by numbers, DX had turned from a cry spinning its wheels into the work of clearing a numbered queue, the report read.
"Many companies see company-wide DX end at a rallying cry. The 'do DX' directive comes, but 'where' won't settle, and the issues spin idle, piled high. What ICE asks is the three axes of impact, confidence, and ease. Tackle the issues that work most, you can be most confident of, and are easiest to execute. The low-confidence big job you head into last, after building momentum. Even with few industry examples, score the three axes and the order of attack is decided by calculation. The day a company where only the DX rallying cry echoed down the hall decided its first move, what changed was not the DX tool but the very judgment that replaces a rallying cry with calculation."
Related Files
Tools Used
- ROI Polygraph — Visualizing individual-dependent work hours, paper-operation cost, and the management effort spinning its wheels
- ROI Proposal Generator — Payback simulation for DX that judges the order of attack by impact, confidence, and ease
