ROI Case File No.595: They Were About to Hand Out Tablets, but Had Never Opened Up Why Paper Was Being Chosen
They Were About to Hand Out Tablets, but Had Never Opened Up Why Paper Was Being Chosen
Chapter 1: They Want to Drop the Paper, but the Floor Won't Move
"We're looking for a way to visualize and manage our production performance data."
Shibori Hirose, IT department manager at Globex Corporation, said this as he laid out the situation. "We're replacing our core system in 2028. But the new core is specified to hold only financial information. Production data will have to be held separately by the division. We want to see KPIs too, so we need an environment that can capture it."
"How is it recorded today?" Claude asked.
"Handwritten on paper forms, then keyed into the system afterward," Hirose answered. "About 120 sheets a day across four plants. It's double work, and the way people write varies. I want to hand out tablets and have them enter directly, but resistance on the floor is strong. We tried once before and it came back unused."
"Have you opened up why that resistance occurs?" I confirmed.
"...We haven't," Hirose answered. "All I've thought about is switching to tablets. Why paper keeps getting chosen. Whether it's gloves, or lighting, or habit. I've never laid the reasons out."
"Start from the solution and you trip in the same place twice," I replied. "Let's break this down with DOUBLE DIAMOND."
Chapter 2: DOUBLE DIAMOND Asks—Widen Twice, Narrow Twice
"This case calls for DOUBLE DIAMOND."
Claude wrote "Discover / Define / Develop / Deliver" on the whiteboard.
"DOUBLE DIAMOND—the twin diamonds—is a four-stage framework that first widens the facts and narrows to a problem, then widens the solution ideas and narrows to a form," I explained. "The crux is not starting from the second diamond. A tablet is only one of the solutions. Hand them out without widening why paper is being chosen, and you get the same ending as last time."
"First, let's measure the current cost," Gemini said, opening ROI Polygraph. The data Hirose had provided went in.
"The monthly cost is out," Gemini read off. "Labor for re-keying handwritten forms into the system averages 195 hours a month across four plants; at ¥3,600 an hour, that's ¥702,000 a month. Illegible entries and the resulting queries, caused by inconsistent writing, average ¥400,000 a month. Delayed detection of abnormalities, because the data isn't visible in real time, averages ¥360,000 a month. Manual aggregation for KPI reporting averages ¥320,000 a month. Hesitation to reinvest and stalled deliberation, following the earlier failed deployment, averages ¥280,000 a month. Total: ¥2,062,000 a month. Annualized, roughly ¥24,740,000."
Hirose stared at the figures. "I was only looking at the re-keying burden. Once you add the queries about handwriting nobody can read, and the part that's been frozen since the failure, it comes to this much."
"Then let's design it with DOUBLE DIAMOND," I continued.
[Discover—widen out why it's paper]
"First, discover," Claude said. "Go into all four plants and watch the moment of writing. Are they writing with gloves on, is there dust, is one hand occupied, is there enough light? Ask the people themselves why it stopped being used last time. Here you don't narrow—you widen the facts only."
[Define—pull the problem away from the input device]
"Next, define," Gemini continued. "Restate the problem from the facts you gathered. In most cases the problem isn't the input device—it's that there are too many fields to fill in, or that the place you write is far from where you stand. Restate that, and you have a question in a solvable shape."
[Develop—widen the solutions beyond devices]
"After define comes develop," I continued. "Don't make tablets the only candidate. Fixed terminals, automatic capture from the machines themselves, entry narrowed to selections, voice. At the entrance to the second diamond, you put out ideas broadly. When there's only one idea, it means you haven't compared anything."
[Deliver—release small and let the floor correct it]
"Last, deliver," Claude continued. "Don't roll out to four plants at once. Release in one process at one plant and fix it together with the people there. Spread only the shape that got used. Last time's failure was handing out a finished product."
[Simulating the investment recovery]
"Let's run the numbers with ROI Proposal Generator," Gemini proposed.
- Initial cost: On-site study across four plants, problem definition, comparative validation of input methods, building the production-data management system, provisioning terminals, and implementing the KPI aggregation base—¥5,100,000 in total
- Monthly cost: System usage, terminal maintenance, plus ongoing operation—¥230,000 a month
- Monthly savings: Elimination of form re-keying = ¥490,000 a month (assuming a 70% reduction), elimination of illegible entries and the resulting queries = ¥320,000 a month, loss suppression through earlier detection of abnormalities = ¥260,000 a month, automation of KPI aggregation = ¥260,000 a month—¥1,330,000 a month in total
- Net monthly savings: ¥1,330,000 − ¥230,000 = ¥1,100,000 a month
- Payback period: ¥5,100,000 ÷ ¥1,100,000 = approximately 4.6 months
"That's a payback of four and a half months," Gemini summarized. "What works is that you don't enter through tablets—you widen twice. An idea that only swaps the device never touches the floor's reasons, so it comes back again. Widen the reasons and then narrow, and you arrive at a shape that gets used. The investment doesn't swing at air."
Hirose looked over the numbers. "I thought choosing a good terminal would settle it. Without widening why it's paper, you can't even decide what to build."
"DOUBLE DIAMOND is a tool for widening twice and narrowing twice," I replied.
Chapter 3: A Deployment Plan That Widens Twice and Narrows Twice
"Let me lay out the approach," I said, standing at the whiteboard.
"Month one—observe the four plants and collect the reasons paper is being chosen. Month two—restate the problem and narrow down the fields to be entered. Month three—generate input-method candidates and compare them. Months four and five—pilot operation in one process at one plant, with revisions. Month six—roll out to the four plants and implement the KPI aggregation base. Month seven onward—design the connection toward the core replacement and expand to more processes."
"There's time before the 2028 replacement. Couldn't we wait until then?" Hirose confirmed.
"Wait, and you arrive with nothing to choose from," Claude replied. "To decide where production data lands at replacement time, you need to know which fields are genuinely required. Run it on the floor now and you come to the design with two years of real data. Making the material by running something small comes before waiting."
Hirose took notes as he spoke. "Widen out why it's paper before handing out terminals. I can see the sequence now."
Chapter 4: The Day They Widened the Reasons and the Floor Started Using It
Ten months later, a report arrived from Hirose.
Entry changed process by process, starting with the pilot. "Once we made it large selection buttons that can be pressed with gloves on, there was almost no resistance. We learned that last time's cause was trying to make them type small characters," Hirose wrote.
The re-keying vanished as well. Data existed the moment it was entered, and the re-entry step disappeared. "The work of re-keying 120 sheets a day is gone entirely. So are the phone calls to confirm illegible handwriting," the report said.
The largest change showed up in how they entered a deliberation. A state of starting from the solution became a state of widening the facts before narrowing. "We used to write approval requests under the heading 'tablet deployment.' Now we first collect why things are the way they are," Hirose wrote.
KPIs became visible too. Manual aggregation disappeared and the figures lined up daily. "What used to be compiled at month-end now appears the next morning. Responses to abnormalities got faster," the report said.
As a side effect, preparation for the replacement advanced. Real data became the material for requirements. "We can decide what to keep in the core and what to hold outside it based on the fields actually being used. We won't arrive at 2028 empty-handed," Hirose wrote.
The final line of Hirose's report read: "I thought the trouble with digitalization was resistance on the floor. But the real problem was that we were about to hand out tablets and had never opened up why paper was being chosen. The moment we widened twice with DOUBLE DIAMOND, the shape to build was decided. Before choosing terminals, widening the reasons came first."
The day a company about to hand out tablets became a company that could widen twice and narrow twice, the digitalization of production data had changed from replacing an input device into a design that widens the facts and restates the problem, the report noted.
"Consultations about digitalization usually arrive in the form of 'we want to bring in tablets.' But there is a question to ask before handing them out. Why is paper being chosen right now? What DOUBLE DIAMOND asks is discover and define, develop and deliver—widening twice and narrowing twice. Widen the reasons, and the answer turns out to lie somewhere other than the device. The day a company that started from the solution could widen the facts, what changed was not the performance of the terminal but the very perspective of widening before narrowing."
Related Files
Tools Used
- ROI Polygraph — Visualizing form re-keying labor, queries caused by inconsistent entries, and the manual cost of KPI aggregation
- ROI Proposal Generator — Investment-recovery simulation for production-data management built from widening the facts and restating the problem
