ROI Case File No.336 | 'The Trapped Drawings of CAD Solutions'

Chapter 1: The Futility of Searching — Can't Know Contents Without Opening Files

The day after the Globex Corp. core system replacement case was solved, a consultation about design operations efficiency arrived. Volume 27, "The Pursuit of Reproducibility," Episode 336 tells the story of transforming internal indifference into interest.

"Detective, in our design department, finding similar past drawings takes an average of 40 minutes per case. And new staff can't find them at all. Because you can't know the contents without opening CAD files."

Kenta Suzuki, Design Director of CAD Solutions, originally from Nagoya, visited 221B Baker Street with a confused expression. In his hands, he held CAD file lists alongside contrasting operation manuals marked "Search Method: Filename Guessing."

"We specialize in mold design. We receive requests from customers and create CAD drawings for molds. Approximately 350 projects annually. Twelve designers. However, there's a problem. Finding similar past drawings takes enormous time."

CAD Solutions' Design Structure: - Established: 2005 (mold design specialty company) - Annual projects: Approximately 350 - Designers: 12 (5 veterans, 4 mid-level, 3 newcomers) - CAD files: Approximately 8,500 files (past 18 years) - Problem: 20-30 drawings in one CAD file, can't verify without opening contents

Suzuki's voice carried deep frustration.

"In mold design, we often reference similar past drawings. 'We've made this shape before,' 'If we modify that drawing, we can finish quickly.' However, finding that drawing takes 40 minutes. As for newcomers, they have no idea where it is."

Typical Problem Cases:

Case 1: Veteran Designer A (15 years experience): - New project: Automotive parts mold design - Task: Find similar past drawings - Method: 1. Guess from 8,500 file server filenames 2. Filter with keywords like "automotive," "parts," "around 2020" 3. Narrow to about 50 candidate files 4. Open CAD files one by one to check contents 5. Discover applicable drawing - Time required: 35 minutes

Case 2: New Designer C (1 year experience): - New project: Home appliance parts mold design - Task: Find similar past drawings - Method: 1. Look at file server but can't tell which is applicable 2. Ask senior designer "Where are home appliance parts drawings?" 3. Senior searches for them (30 minutes) 4. Or gives up thinking "faster to design from scratch" - Time required: 30 minutes (stealing senior's time), or give up and create new (3 hours)

Case 3: Quotation Staff D (non-designer): - Task: Reference similar past drawings for customer quotation requests - Method: 1. Ask designer "Have we done this project before?" 2. Designer searches relying on memory (20 minutes) 3. Or gives up saying "don't know" - Problem: Quotation accuracy decreases, loss of order opportunities

Suzuki sighed deeply.

"There's more. We're considering implementing a 'CAD Drawing Auto-Search System.' A system that reads drawings with AI-OCR and can search by shapes and dimensions. However, internal reactions are poor. 'Current way is sufficient,' 'Learning new systems is troublesome,' 'Doubtful if it's truly useful.' They're reluctant to implement."

Chapter 2: The Wall of Indifference — The Organization That Resists Change

"Suzuki-san, when implementing new systems internally, what do you think is the biggest barrier?"

To my question, Suzuki answered immediately.

"Budget. System implementation costs 5 million yen. We don't know if management will approve."

Current Understanding (Budget Problem Model): - Recognition: Can't implement because can't secure budget - Problem: Can't see the true barrier of internal indifference

I explained the importance of gradually raising internal interest.

"The problem isn't budget. It's internal 'indifference.' If designers strongly feel 'we want this system,' management will approve the budget. However, if they think 'current way is sufficient,' no matter how excellent the system, it won't be implemented. AIDMA — Attention, Interest, Desire, Memory, Action. Draw attention, create interest, arouse desire, remain in memory, prompt action. This process moves organizations."

⬜️ ChatGPT | Catalyst of Concepts

"Transform indifference into interest. Move organizational hearts gradually with AIDMA."

🟧 Claude | Story Alchemist

"Trapped drawings always await 'interest.' What opens that door is not demos but experience."

🟦 Gemini | Compass of Reason

"AIDMA is persuasion technology. Move people through five stages: attention, interest, desire, memory, action."

The three members began analysis. Gemini displayed the "AIDMA Framework" on the whiteboard.

AIDMA's 5 Stages: 1. Attention: First, make them know it exists 2. Interest: Have them interested in details 3. Desire: Make them think "I want it" 4. Memory: Remain in memory so they don't forget 5. Action: Have them actually act

"Suzuki-san, let's start by drawing internal 'attention.'"

Chapter 3: Persuasion Through Experience — Not Demos, Actual Use

Phase 1: Attention — 1 week

First, needed to make designers aware of "CAD Drawing Auto-Search System's existence."

Measures:

Internal Email Announcement: - Subject: "Isn't it time to end spending 40 minutes finding past drawings?" - Content: - "Conducting demo of AI-OCR drawing search system" - "See drawing contents without opening files" - "Date/time: XX/XX 15:00-16:00"

Poster Display: - Displayed A3-size poster in design department break room - "30 drawings in one CAD file. 40 minutes to search. Let's end this."

Results: - Internal email open rate: 58% (7 of 12) - Demo participation requests: 3 (1 veteran, 2 mid-level)

Problem: - Few participation requests (3 of 12, 25%) - No newcomers requested participation

Cause Analysis: - Word "demo" is passive and unattractive - Doubts exist about "is it truly useful"

Phase 2: Interest — 2 weeks

Instead of demos, decided to have them actually use it.

Measures:

Free Trial Provision: - Negotiated with system vendor, conducted 2-week free trial - Target: All 12 designers - Condition: Each tries in actual work

Usage Lecture (15 minutes): - Explained basic system operation - "Can search not by filename but by drawing shapes and dimensions" - "For example, search for 'circular' 'diameter 50mm' displays list of applicable drawings"

Actual Search Demo (5 minutes): - Search with veteran designer A's actual project - Before: 35 minutes with filename guessing - After: 2 minutes with AI search - Result: 5 applicable drawings hit, first one was sought drawing

Reactions: - Veteran designer A: "This might be convenient. I'll try it" - Mid-level designer B: "Doubtful if truly usable. But I'll try" - New designer C: "Can I use it too? I want to try"

Phase 3: Desire — 2 weeks

During 2-week free trial, had them use in actual work.

Usage Status:

Veteran Designer A (15 years experience): - Usage count: 8 times (2 weeks) - Average search time: 3 minutes (Before: 35 minutes) - Time reduced: 32 minutes/time × 8 times = 256 minutes (4.3 hours/2 weeks) - Feedback: "Search time became one-tenth. This is excellent"

Mid-level Designer B (7 years experience): - Usage count: 5 times (2 weeks) - Average search time: 5 minutes (Before: 30 minutes) - Time reduced: 25 minutes/time × 5 times = 125 minutes (2.1 hours/2 weeks) - Feedback: "Initially skeptical, but actually using it surprised me. Seeing drawings without opening files is revolutionary"

New Designer C (1 year experience): - Usage count: 12 times (2 weeks) - Before: Ask seniors (30 minutes) or give up and create new (3 hours) - After: Search and discover by self (5 minutes) - Time reduced: 25 minutes/time × 12 times = 300 minutes (5 hours/2 weeks) - Feedback: "Until now, felt bad asking seniors. But with this system, I can search myself. Very helpful"

Quotation Staff D: - Usage count: 6 times (2 weeks) - Before: Ask designers (20 minutes) or give up - After: Search and discover by self (3 minutes) - Feedback: "Quotation accuracy improved. Can reference similar past projects, so man-hour estimates are accurate"

Phase 4: Memory — 1 week

After 2-week trial ended, visualized results and implemented measures to remain in memory.

Measures:

Results Briefing (30 minutes): - All 12 designers participated - Trial participants (all 12) presented results

Presentation Content: - Veteran designer A: "Reduced 4.3 hours in 2 weeks. Calculation shows about 100 hours annual reduction" - Mid-level designer B: "Truly convenient seeing drawings without opening files" - New designer C: "Could search myself without asking seniors. This is revolution for me" - Quotation staff D: "Quotation accuracy improved, seems order rate will also increase"

Overall Results: - Total reduced time of 12 participants: About 28 hours (2 weeks) - Annual conversion: 28 hours × 26 weeks = 728 hours - Monetary conversion: 728 hours × 5,000 yen (designer hourly rate) = 3.64 million yen/year

Poster Display: - "28 hours reduced in 2 weeks. Your 40 minutes to 2 minutes" - Displayed in break room, remain in memory

Phase 5: Action — 1 month

Finally, prompted organizational action to "implement."

Measures:

Request Letter from Designers: - All 12 trial participants submitted request letter to management - Content: "Request official implementation of CAD Drawing Auto-Search System" - Reasons: "Annual 3.64 million yen time reduction, eliminate design work personalization, efficient newcomer training"

Presentation to Management: - Director Suzuki presented to management (15 minutes) - Presented result data, designer voices, cost-effectiveness - Implementation cost: 5 million yen - Annual reduction effect: 3.64 million yen - Investment recovery period: 1.37 years

Management Reaction: - President: "If all designers request it, let's implement" - CFO: "Cost-effectiveness is clear. Approved"

Decision: - Official implementation of CAD Drawing Auto-Search System approved

Chapter 4: Results of Establishment — Changes After 12 Months

Results After 12 Months:

Search Time Reduction: - Before: Average 40 minutes/time - After: Average 3 minutes/time - Reduction: 37 minutes/time (93% reduction)

Annual Reduced Hours: - 12 designers × average 20 searches/month × 37 minutes = 8,880 minutes/month (148 hours/month) - Annual: 1,776 hours

Monetary Effect: - 1,776 hours × 5,000 yen (hourly rate) = 8.88 million yen/year

Investment Recovery: - Implementation cost: 5 million yen - Annual reduction effect: 8.88 million yen - ROI: 78% (first year) - Investment recovery period: 0.56 years (about 7 months)

Eliminate Design Work Personalization:

Before: - Newcomers can't find past drawings, depend on seniors - Veteran knowledge personalized

After: - Even newcomers can search past drawings themselves - Design know-how accumulated in organization

New Designer C's Growth: - Before trial: Average 5 days per project - 12 months later: Average 3.2 days per project (36% efficiency) - Reason: Can reference past drawings, no need to design from zero

Quotation Accuracy Improvement:

Before: - Quotation staff D's quotation accuracy: ±30% - Order rate: 42%

After: - Quotation accuracy: ±15% - Order rate: 58% (+16 points) - Reason: Can reference similar past projects, man-hour estimates accurate

Organizational Changes:

Designer Voices:

Veteran Designer A: "Initially thought 'current way is sufficient.' But actually using it changed my thinking. Search time became one-tenth, time to concentrate on design increased. 100 hours annual reduction is great value for me."

New Designer C: "Without this system, I think I'd still be completely dependent on seniors. But now I can search and reference past drawings myself. This accelerated my growth."

Suzuki's Feedback:

"Until using AIDMA, we thought 'if we implement system, everyone will use it.' But that was wrong. There was a big wall called 'indifference' internally.

Made them aware of existence with Attention, had them actually use it with Interest, had them experience results with Desire, visualized with Memory, moved organization with Action. This 5-stage process made implementation successful.

Now all designers master the system. Achieved 8.88 million yen annual reduction and eliminated design work personalization. Trapped drawings are no longer trapped."

Chapter 5: Detective's Diagnosis — Change Happens Gradually

That evening, I contemplated the technology of moving organizations.

CAD Solutions had the illusion that "if we implement good systems, internal operations will change." However, what moves organizations isn't system performance but people's interest.

By gradually raising interest with AIDMA, indifference transformed into desire, and the organization moved. And achieved 8.88 million yen annual reduction and eliminated design work personalization.

"Transform indifference into interest. Attention, interest, desire, memory, action. People move through these five stages. Excellent systems won't be implemented without people's interest."

The next case will also depict the moment of gradually raising organizational interest.

"Not demos, but experiences. Attention, Interest, Desire, Memory, Action. Transform indifference into action through these five stages. Trapped drawings are opened with the key called interest." — From the Detective's Notes

Related Files