ROI Case File No.301 | 'Carelink's Invisible Users'

Chapter 1: The Trap of Grand Visions—A Dream to Save 2,300 Facilities

The week after resolving the EcoWaste OODA incident, a consultation arrived from Kanagawa regarding a care service company's new system development. Episode 301, opening Volume 25 "The Pursuit of Certainty," tells the story of transforming a massive plan into a chain of small improvements.

"Detective, we provide services to 2,300 care facilities nationwide. We're now developing a new web application. Care records, vital sign management, family information sharing... consolidating everything into one app. However, we've been struggling with selecting a development partner for six months."

Misaki Kawashima, system planning manager at Carelink Inc., originally from Yokohama, visited 221B Baker Street unable to hide her confusion. In her hands were a requirements document exceeding 300 pages and, in stark contrast, meeting minutes stamped "Selection Postponed."

"We're based in Kanagawa, providing IT services to care facilities nationwide. Digitization of care records, staff shift management, billing support... We've offered various tools. However, they're all independent systems, not integrated."

Carelink's New App Vision: - Establishment: 2012 (IT services for care facilities) - Annual Revenue: ¥4.2 billion - Customer Facilities: 2,300 (nationwide) - Existing Services: 5 independent systems - New App Vision: Integrated web app for all services - Development Scale: Large-scale project from requirements to maintenance - Budget: ¥280 million (5 years) - Status: Stalled for 6 months on development partner selection

Deep anxiety filled Kawashima's voice.

"The problem is that the requirements are overwhelming. Care records, vital sign management, family contact, staff shifts, billing... We've incorporated 38 features. We requested proposals from three development partner candidates, but they all say 'the risk is too high.'"

Reactions from 3 Development Partner Candidates:

Company A (Major SI firm): "Developing 38 features at once is reckless. Minimum 3 years development period, budget needs to exceed ¥500 million."

Company B (Mid-sized development firm): "We cannot handle this. We lack knowledge of the care industry. The risk is too high to accept."

Company C (Startup): "It might be possible with agile. However, the requirements are too fixed, lacking flexibility."

Result: Six months passed, zero contracts

"Management says 'build it quickly.' But from facilities in the field, we hear voices of concern: 'Will it really be usable?' We created a 300-page requirements document, but I no longer know if it's truly correct."

Chapter 2: The Illusion of Planning—Those Who Seek Perfection Gain Nothing

"Ms. Kawashima, how was this requirements document created?"

To my question, Kawashima answered.

"We spent a year creating it. We interviewed 100 existing customers and collected voices from the field. Then we incorporated all their requests. To create the 'perfect system.'"

Current Approach (Plan-Heavy Type): - Requirements Period: 1 year - Interview Target: 100 facilities - Requirements Count: 38 features - Premise: "Perfect planning leads to success" - Problem: Plan too large to execute

I explained the importance of starting small.

"Perfect planning is the enemy of execution. PDCA—Plan, Do, Check, Act. Turning this cycle small and fast is the path to success."

⬜️ ChatGPT | Catalyst of Vision

"Abandon the massive plan. With PDCA, accumulate small successes"

🟧 Claude | Alchemist of Stories

"The dream to save 2,300 facilities is beautiful. But that dream begins with the reality of saving one user"

🟦 Gemini | Compass of Reason

"PDCA is a spiral of improvement. Plans minimal, execution immediate, evaluation honest, improvement certain"

The three members began analysis. Gemini developed the "PDCA Cycle Framework" on the whiteboard.

PDCA's 4 Steps: 1. Plan: Form hypotheses with minimum features 2. Do: Pilot introduction on small scale 3. Check: Collect user feedback 4. Act: Reflect in next cycle

"Ms. Kawashima, let's start small with the most important features, not all 38 at once."

Chapter 3: The First Step—Courage to Narrow 38 Features to 3

Phase 1: Plan - 1st Cycle (2 weeks)

First, we classified the 38 features by priority.

Re-interviewing 100 Existing Customers: "Of the 38 features, which troubles you most?"

Responses (multiple choice): 1. Care record digitization: 82 facilities (82%) 2. Vital sign management: 68 facilities (68%) 3. Family information sharing: 58 facilities (58%) 4. Staff shift management: 45 facilities (45%) 5. Billing: 38 facilities (38%) ... (remaining 33 features under 15%)

Discovery: Just the top 3 features cover the main issues of 82% of facilities.

Kawashima was astonished.

"So of 38 features, only 3 are truly necessary...?"

1st Cycle Plan: - Target Feature: Care record digitization (MVP: minimum viable version) - Test Facilities: 5 (small, medium, large scale) - Development Period: 2 months - Budget: ¥8 million (3% of total budget)

Phase 2: Do - 1st Cycle (2 months)

Developed a minimum care record app.

MVP (Minimum Viable Product) Specifications: - Care record input (text/voice) - Record search/viewing - Information sharing among staff - That's it. No other features.

Pilot Introduction at 5 Test Facilities: - Small facilities (20 beds): 2 facilities - Medium facilities (50 beds): 2 facilities - Large facility (100 beds): 1 facility

Phase 3: Check - 1st Cycle (2 weeks)

After 2 months, collected feedback from test facilities.

Voice from Small Facility A: "Simple and easy to use. However, we want to attach photos. Necessary for bedsore records"

Voice from Medium Facility B: "Voice input is convenient. But dialects aren't recognized. Only works with standard Japanese"

Voice from Large Facility C: "With 100 staff sharing, we need permission management. Anyone can see all records—that's a problem"

Quantitative Data: - Usage Rate: 68% (target 70%) - Record Input Time: Reduced to 65% of previous - Satisfaction: 3.8/5 ("Simple and good" majority)

Issues: 1. No photo attachment feature 2. Voice recognition doesn't support dialects 3. Insufficient permission management

Phase 4: Act - 1st Cycle (1 week)

Based on feedback, planned the next cycle.

2nd Cycle Improvements: 1. Add photo attachment feature 2. Improve voice recognition for dialects (support 5 dialects) 3. Add permission management (3 levels: staff, admin, family)

2nd Cycle Goals: - Usage Rate: 70% → 85% - Record Input Time: Additional 20% reduction - Satisfaction: 3.8 → 4.5

Chapter 4: Accelerating the Spiral—Small Cycles Generate Massive Results

2nd Cycle (3 months):

Plan: Added 3 improvement items to development plan

Do: Development, expanded test facilities to 15 (5 from 1st cycle + 10 new)

Check: - Usage Rate: 85% (goal achieved) - Photo Attachment: "Easier to track bedsore progression"—well received - Voice Recognition: Satisfaction improved with dialect support - Permission Management: "Can share safely with families"

Act: Add "vital sign management" feature in next cycle

3rd Cycle (3 months):

Added vital sign management feature (blood pressure, temperature, pulse recording).

Check: - Usage Rate: 92% - Highly rated: "Care records and vitals managed in one app"

4th Cycle (3 months):

Added family information sharing feature.

Check: - Usage Rate: 95% - "Inquiries from families decreased"—reduced field burden

Full Deployment After 12 Months:

Test Facility Results: - Test Facilities: Total 50 facilities - PDCA Cycles: 4 implemented - Implemented Features: 3 (care records, vital management, family sharing) - Usage Rate: 95% - Satisfaction: 4.7/5

National Deployment Decision: - Deployment Target: 2,300 facilities - Initial Features: 3 (tested) - Additional Features: Added sequentially via PDCA (4 cycles/year)

Results After 24 Months:

Deployment Record: - Deployed Facilities: 88% of 2,300 (2,024 facilities) - Usage Rate: 91% - Customer Satisfaction: 4.6/5

Business Results: - Development Investment: Total ¥120 million (43% of initial ¥280 million budget) - Annual Revenue: ¥4.2 billion → ¥5.6 billion (+33%) - Cancellation Rate: 8% → 2% (continuous improvement via PDCA)

Organizational Change: - Development System: "Planners" → "Companions" - Quarterly PDCA cycles became routine - Culture of immediately reflecting customer voices

Chapter 5: Detective's Diagnosis—Opening Volume 25 "The Pursuit of Certainty"

That night, I contemplated the essence of PDCA.

In the final episode of Volume 24, we learned about the OODA Loop. The power of high-speed cycles adapting to the field.

However, PDCA is different. It doesn't disregard planning but makes it "minimal." Then execute, evaluate, improve. By repeating this cycle, uncertain futures transform into "certain results."

Carelink abandoned the massive plan of 38 features. They started with 3. By spinning small PDCA cycles 4 times, they created an app accepted by 2,300 facilities.

"Perfect planning is an illusion. But the accumulation of small plans and certain execution changes reality."

Volume 25 "The Pursuit of Certainty" opens here.

The next case will also depict the moment PDCA transforms uncertainty into certainty.

"Abandon massive plans. Start small with PDCA, spin fast, accumulate certainly. Success lies beyond the spiral"—From the Detective's Notes

Related Files