METHODOLOGY · CUSTOMER M EVIDENCE

Digitalization Success
Methodology

7 iron rules + 5 red lines distilled from 8 years of digitalization practice at a high-end manufacturing enterprise
Not theoretical deduction, but crystallized real-world experience

Customer M (a global leading FPC flexible printed circuit manufacturer) with revenue of 40B+, serving the Apple supply chain, underwent 8 years of digitalization, iterating from 50 dashboards to 2000+, moving from the big data era into the Agent era. The following methodology is derived from its real-world practice and validated by the Leansight platform.
ORIGIN STORY

A Severe Criticism That Launched an 8-Year Digitalization Journey

From the IT manager being severely criticized, to CIO, to factory GM—behind this career path is the complete story of digital transformation going from "taking the blame" to "carrying the flag"

2017
2017
IT Manager · Darkest Hour
The GM called the IT manager into his office and severely criticized him.
The digitalization reports and dashboards (Halo BI) were too ineffective, too slow to respond, had too many IT staff, delivered no value.
This moment became the starting point of Customer M's comprehensive digital transformation—not "should we do it," but "we can't not do it anymore."
2022
2022
IT Manager → CIO · Vindication Moment
5 years of deep work, the IT manager was promoted to CIO.
From the severely criticized "scapegoat" to the CIO in charge of factory-wide data strategy—what happened in those 5 years?
IT-OT Fusion Center established Company-wide low-code training 50 dashboards as starting point Apple supply chain compliance driving force—IT was no longer a cost center, but a value engine.
2024
2024
CIO → Factory GM · Summit Moment
At year-end, the CIO was promoted to factory General Manager.
From IT manager → CIO → GM, a three-level jump in 8 years. A severely criticized IT manager ultimately took charge of the entire factory—digital transformation is not a destination, but a path to the core of business operations.
2000+ dashboards Big Data → Low-Code → Agent Smart Factory certification successful Data-driven business decisions—what was once "no value" has become today's core competitiveness.

The severe criticism in 2017 hurt not the person, but the four words "no value".
8 years later, the IT manager who was severely criticized became the factory GM—the best response is not to defend, but to spend 8 years making data truly deliver value.
The following 7 principles + 5 red lines are the crystallization of this 8-year journey.

METHODOLOGY OVERVIEW

Methodology Spiral

7 principles like a DNA double helix, intertwined and mutually driving

1 Demand-Driven Top-Down Pull 2 Org Enablement Rotation · Fusion Center 3 Vendor Selection Service + Productization 4 Architecture Stable Core, Agile Edge 5 Talent Evolution Company-Wide Development 6 Implementation Small-Step Iteration 7 Strategic Alignment Smart Factory · Annual Theme Spiral upward · each loop is deeper and broader than the last
SEVEN PRINCIPLES

Seven Principles and Cases

Each comes from Customer M's real practice, each with Leansight fit explanation

1
DEMAND-DRIVEN

Top-Down Demand Pull

Digitalization is not an optional exercise for the IT department, but is pulled top-down by the dual forces of external customer compliance requirements and internal management's data-driven decision-making. When the top decision-maker truly uses data as the management language, the entire organization will treat data as a core asset rather than an auxiliary tool.
Customer M Practice
External pull: Apple's supply chain requires process stability and controllability, demanding investment in IT systems for end-to-end traceability—this is a rigid demand brought by external compliance.
Internal pull: The GM's own management style is data-driven; decisions in business meetings are based not on presentation materials but on real-time data dashboards. This top-down data culture ensured sustained strategic-level investment in IT construction.
Leansight Fit
CEO Headline is purpose-built for "top-down demand pull"—productizing the GM's data-driven management philosophy. AI autonomously discovers business blind spots, pushing high-confidence insights daily, so decision-makers don't need to proactively look at reports—data comes to them. The paradigm shift from "people looking for data" to "data finding people" is the technical implementation of top-down pulling power.
2
ORGANIZATIONAL ENABLEMENT

Rotation-Driven IT-OT Fusion

The biggest obstacle to digitalization is not technology, but the cognitive gap between IT and OT. Establishing an "IT-OT Fusion Center" as a cross-boundary hub, paired with job rotation and performance schemes, lets those who understand process learn IT, and those who understand IT go to the workshop floor. When an IT director can be transferred to GM, it means the organization has connected the meridians of technology and business.
Customer M Practice
IT-OT Fusion Center: Established a dedicated organization, mixing IT and production technical personnel in teams, breaking down departmental walls.
Rotation mechanism: Encouraged cross-departmental transfers, with adjusted performance evaluation schemes to eliminate the fear that "rotation = demotion." The most typical case: IT director transferred to GM, showing that digital talent had gained holistic business vision.
Performance alignment: Rotation is not "exile" but "cultivation"—performance schemes aligned with rotation goals, ensuring transferred personnel have both motivation and evaluation criteria in their new roles.
Leansight Fit
LeanFusion Data Capability Platform is itself the technical carrier of "IT-OT fusion"—IT personnel develop applications using the low-code engine, OT personnel drive business with BI dashboards, both collaborating on the same platform rather than being siloed. LeanCodee Low-Code Engine lowers the participation threshold for OT personnel, enabling production line engineers to become application developers—this is the technical foundation of "company-wide development."
3
VENDOR SELECTION

Service Support + Productization Capability

Selection is not based on brand size or how beautiful the PPT is, but on two things: service support capability (can they respond quickly when problems arise, deep co-running) and productization capability (can they continuously upgrade without being dragged down by customization). Outcome-driven, with deliverables as the measurement standard.
Customer M Practice
Service first: During selection, prioritize evaluating the vendor's service support team size, response speed, and industry understanding depth. Vendors who can station teams, deeply co-run, and iterate together with the internal team are qualified partners.
Productization standard: Reject the "one project, one codebase" customization trap. Require vendors to have productization capability—the same platform, adaptable to different scenarios, meeting needs through configuration rather than coding.
Outcome-driven: Don't accept by "system go-live," but by "business outcomes"—are dashboards being used? Are processes running? Has efficiency improved?
Leansight Fit
Leansight's platform product strategy naturally aligns with this selection philosophy: LeanFusion / LeanCodee / LeanBI are all standardized platforms, meeting scenario needs through configuration rather than customization. The Leansight team insists on "tower-building" implementation—not doing a project and leaving, but building a tower that can grow. Service support is not after-sales, it's continuous co-building. This is the core reason Customer M chose Leansight and has continued the partnership for 8 years.
4
ARCHITECTURE STRATEGY

Big Systems Stable, Small Needs Agile

Core big systems like MES and ERP remain stable—they carry the production lifeline and cannot be frequently modified. But the report needs, process optimization needs, and ad-hoc analysis needs that arise daily cannot wait for IT scheduling. Through a low-code platform for rapid development, "big systems steady as a rock, small apps fast as lightning."
Customer M Practice
Big system lock-down: Core systems like MES and ERP are kept stable after selection, with changes going through strict change management processes to ensure production continuity is not affected.
Derivative demand diversion: The large volume of "small needs" generated daily—temporary dashboards for a specific process, analysis reports for quality issues, drill-through queries for specific KPIs—are rapidly built through low-code platforms, going from requirement to launch in "days."
Data operations system: Starting from 50 dashboards, through continuous low-code iteration, accumulated 2000+ dashboards over 8 years, covering the full-chain data visibility from production lines to business meetings.
Leansight Fit
The combination of LeanCodee Low-Code Engine + LeanBI Analytics Components is the best implementation of the "stable core, agile edge" strategy. LeanCodee lets business personnel quickly build applications through drag-and-drop configuration without touching MES/ERP core code; LeanBI turns dashboard building from "project" to "configuration." The Leansight platform connects with MES/ERP through standard interfaces—big systems provide the data foundation, the low-code platform provides the agile upper layer, with clear architecture and distinct boundaries.
5
TALENT EVOLUTION

Company-Wide Development · Continuous Training

Digitalization is not the work of a few IT engineers, but a company-wide capability upgrade. From the big data era to the low-code era to the Agent era, continuously training the team, making everyone a participant and contributor in digitalization. Not "IT builds systems for business to use," but "business builds their own systems using tools."
Customer M Practice
Big data era: Cultivated the team's data analysis capabilities, learning SQL, ETL, data modeling, building data thinking.
Low-code era: Comprehensively shifted to low-code development, lowering the threshold for more business personnel to participate, from "people who submit requirements" to "people who build applications."
Agent era: The current phase being entered—training the team to understand AI Agents, prompt engineering, multi-agent collaboration. Not passively waiting for AI to replace, but proactively making the team AI drivers.
Continuous training mechanism: Not one-off training, but annual plans—setting capability upgrade themes each year, with training courses and hands-on projects.
Leansight Fit
Leansight's platform evolution path is completely synchronized with Customer M's talent evolution path: LeanFusion (data fusion → big data era) → LeanCodee (low-code → company-wide development era) → CEO Headline + Agent Factory (AI Agent era). Leansight is not just a tool vendor, but a capability upgrade co-runner—providing training, certification, and best practices to help customer teams complete leaps in each technology era.
6
IMPLEMENTATION RHYTHM

Small-Step Iteration · Continuous Evolution

Reduce upfront investment, validate quickly, then scale. Don't do a "comprehensive" top-level design and invest all at once, but validate with "small and beautiful" pilots then replicate quickly. Each iteration has clear goals, measurable outcomes, and replicable experience. Over 8 years, growing 50 dashboards into a forest of 2000+.
Customer M Practice
Data operations system: Started with 50 dashboards, not planning 2000 from the start. First got dashboards working for one workshop, one process, then replicated to other workshops after validating usefulness. Each iteration accumulated templates and experience, reducing the next round's development cost.
Reduce upfront investment: Don't heavily invest development resources when requirements are unclear. First build an MVP (Minimum Viable Product), let users see something tangible, then iterate requirements.
Validate then scale: Once validated as effective, immediately scale in a standardized manner. MVP phase allows trial and error; scaling phase requires standardization.
8-year iteration: 50→200→500→1000→2000+, each doubling based on the previous round's accumulated experience, forming a compound interest effect.
Leansight Fit
Leansight's "tower building" philosophy is the methodization of this rhythm—not pouring all at once, but growing layer by layer. LeanCodee's "Lego-style" components enable each application to be quickly built, validated, and iterated. LeanBI's dashboard templates make the "from 1 to N" replication cost approach zero. Leansight's implementation methodology clearly distinguishes the MVP phase (rapid validation, roughness allowed) from the standardization phase (rapid replication, standardization required), fully aligned with Customer M's iteration rhythm.
7
STRATEGIC BENCHMARKING

Certification Benchmarking · Annual Theme

Planning must be practical and deeply pathed, with a theme set each year. Through applying for Smart Factory and other government certifications to gain outsized returns—not for a plaque, but benchmarking against national standards to force capability building, while gaining policy dividends and brand endorsement. Annual themes ensure focus, breakthrough, and no diffusion each year.
Customer M Practice
Smart Factory application: Proactively benchmarked against national/provincial Smart Factory evaluation standards, using the application process as a capability health check—each standard requirement is a compass for construction. After successful application, gained government subsidies, tax incentives, and brand endorsement, achieving outsized returns.
Annual theme system: Setting one digitalization theme each year (e.g., "Data Governance Year," "Equipment Interconnection Year," "AI Application Year"), with all resources focused on the theme for the year. Avoiding doing a bit of everything and nothing deeply.
Practical planning: Not pursuing a grand "all at once" blueprint, but formulating a deep path suited to current capability level—this year's theme is next year's foundation, advancing one step each year.
Leansight Fit
Leansight has rich co-running experience in Smart Factory application—helping customers benchmark against evaluation standards, identify capability gaps, and formulate gap-closing paths. Leansight's Four-Layer Architecture (Autonomous Discovery Layer → Smart Control Layer → Lean Collaboration Layer → Agile Foundation Layer) naturally maps to the Smart Factory evaluation standard's capability dimensions, allowing customers to benchmark and close gaps layer by layer. Meanwhile, Leansight's annual planning service helps customers formulate yearly themes, aligning platform capability building with strategic goals.
WHAT NOT TO DO

What Customer M Didn't Do

Success is not just about what you did, but also about what you firmly refused to do—5 red lines marking the forbidden zone

✕
ANTI-01

Trusting PPT
Focus on actual results

Don't make decisions because a presentation is beautiful. Look at deliverables, not presentation slides—is the system actually running? Are users actually using it? Has efficiency actually improved?

During vendor selection, don't look at PPT cases; require live demos of real systems connected to real data environments
✕
ANTI-02

Only submitting requirements
Bring in service provider early to co-build

For innovative projects, don't first write requirement docs then tender, but bring in service providers early to jointly formulate the plan—letting implementers participate when requirements aren't yet clear, avoiding rework from "requirement misunderstanding."

When launching the data operations system, Leansight team was on-site from the requirements research phase, co-designing with Customer M rather than "taking orders to develop"
✕
ANTI-03

Detached from business
All proposals business-driven

Don't accept projects where "IT thinks it's useful" but business can't articulate the value. Every proposal must answer: which business problem does it solve? Who will use it? How much efficiency improvement after use? IT investment without business scenarios is rejected outright.

Whether a dashboard goes live isn't about how good the charts look, but whether the production line supervisor opens it daily to make decisions
✕
ANTI-04

Blindly choosing big vendors
Service support capability is limited

Big vendor platforms are powerful but have limited service response radius—your needs can't get into their priority queue. Manufacturing digitalization requires deep co-running, not "can't reach them after signing the contract." Choose service depth, not brand size.

A big vendor platform had comprehensive features but rotated on-site teams every 3 months; Customer M switched to Leansight for continuous deep co-building
✕
ANTI-05

Copying methodology verbatim
Framework right, content iterated

Methodology is the skeleton, not the flesh. The house framework must be right, but the interior decoration is completed through practice iteration. Not copying a "best practice" and calling it done, but continuously filling, adjusting, and evolving under the right framework.

The QCDSM metric framework was set right in one pass, but each metric's definition, thresholds, and weights were iterated through dozens of versions over 8 years
The 7 "what we did" principles are the methodology skeleton of success, the 5 "what we didn't do" red lines are the guardrails preventing failure.
The skeleton ensures the right direction; the guardrails prevent fatal mistakes—correct framework + firm restraint = sustainable digitalization
🏗️
Framework Right
The house structure can't be wrong
🚫
Red Line Restraint
5 forbidden zones firmly avoided
🎨
Content Iteration
Interior decoration done in practice
FAILURE CASE STUDIES

Smart Operations Control Tower Failure Case Warnings

7 real failure modes—each is a cautionary tale, each is a reverse validation of Customer M's methodology

🎯
FAIL-01

Not Business-Led
IT-initiated, business doesn't cooperate

An automotive industry enterprise where the Smart Operations Control Tower project was initiated by the IT department rather than business. During rollout, business departments cooperated poorly—not providing data, not attending reviews, not using the system. IT fell into a "talking to itself" dilemma, and the project ultimately became IT's "one-man show."
📍 Automotive industry · IT-initiated, poor business cooperation, project stalled
✅ Correct approach: Customer M Principle 1 "Demand-Driven"—pulled top-down by the GM, business departments shift from passive to active
📢
FAIL-02

Lack of Strong Promotion
Middle management resists, transparency blocked

One core value of the Smart Operations Control Tower is data transparency—making metrics at every level visible and traceable. But this directly touches middle management's interests, triggering covert resistance or even open opposition. Without strong promotion from top leadership and mechanisms ensuring open transparency, transparency becomes "paper talk."
📍 Data transparency triggered middle management resistance, lack of strong top-down push, system became a mere formality
✅ Correct approach: Build open transparency mechanisms + strong top-down promotion; Customer M's GM personally used data for decisions in business meetings
🏭
FAIL-03

Cross-Site Metric Differences
Systems not unified, replication difficult

An automotive plant where Qingdao and Wuhu factories had ununified metric management systems, ununified business system interfaces, and inconsistent data definitions. After the control tower was validated in one site, replication to other sites encountered "acclimatization"—the same dashboard meant different things at different sites, the same metric had different calculation logic. Replication costs far exceeded expectations.
📍 Automotive plant · Qingdao and Wuhu factories' metrics/interfaces not unified, cross-site replication difficult
✅ Correct approach: Customer M Principle 4 "Architecture Strategy"—big systems with unified standards, small apps using low-code to adapt to each site's differences
⚙️
FAIL-04

Unstable Central Systems
EAP/MES missing, castle in the air

The Smart Operations Control Tower is a "superstructure" that needs EAP, MES, and other central systems to provide a stable data foundation. An enterprise imported the control tower before MES was stably running, resulting in unreliable data sources, guaranteed real-time performance, and frequent false alarms—the control tower was built on sand; no matter how beautiful, it would collapse.
📍 Importing control tower before MES/EAP was stable, data unreliable, system became decorative
✅ Correct approach: Customer M Principle 4 "Stable Core, Agile Edge"—first ensure MES/ERP and other big systems are stable, then build the agile upper layer
💾
FAIL-05

Data Middle Platform Driven
Heavy on means, light on purpose, over-investment

An approach driven purely by data middle platform focuses only on the technical means of "gathering data together" while neglecting the fundamental purpose of "what business problem the data should solve." Heavy upfront investment in building the middle platform, ETL, and data warehouse, with no one answering "who looks at this data, and what do they do after looking." Over-investment, extremely slow returns, ultimately becoming a "data swamp."
📍 Data middle platform driven, excessive upfront investment, no business scenarios, became a data swamp
✅ Correct approach: Customer M Principle 1 "Demand-Driven" + Principle 6 "Small-Step Iteration"—pulled by business scenarios, don't build useless middle platforms
🔄
FAIL-06

Lack of Iterative Thinking
Siloed optimization, overall loss of control

Operations management is commander's holistic management, requiring starting from the whole and optimizing the whole, not siloing. An enterprise built production, quality, equipment, and warehouse separately—each module had its own dashboards and metrics, but modules couldn't link and data couldn't drill through. The control tower became "control tower cluster," and overall operations remained invisible.
📍 Modules built in silos, lack of overall iterative design, control tower became "control tower cluster"
✅ Correct approach: Customer M Principle 6 "Small-Step Iteration" + Principle 7 "Strategic Benchmarking"—overall framework + continuous iteration, no siloing
👤
FAIL-07

Users Not in Operations
Metrics change fast, system can't keep up

Lean and agile operations mean metrics change fast—new KPIs constantly emerge, old metric definitions constantly adjust, business processes continuously optimize. If only the IT team handles operations, every metric adjustment goes through IT scheduling, and the system always lags behind business. Business itself needs to participate in operations to keep the system evolving in sync with business.
📍 Business not participating in operations, metric adjustments all through IT scheduling, system always lagged behind business
✅ Correct approach: Customer M Principle 5 "Talent Evolution"—company-wide development, business maintains its own dashboards

7 failure cases, each corresponding to the reverse of one of Customer M's methodology principles.
Failure is not accidental; it's the inevitable result of methodology absence—success has methods, failure has patterns.
Customer M's 7 iron rules + 5 red lines are the "pitfall avoidance guide" distilled from these cautionary tales.

⚠️
7 Failure Modes
Not business-led · Lack of promotion · Metrics not unified
Unstable core · Middle platform driven · No iteration · Not in operations
✅
Customer M's 7 Iron Rules
Demand-driven · Org enablement · Vendor selection
Architecture strategy · Talent evolution · Implementation rhythm · Strategic alignment
WHY LEANSIGHT

Why Leansight Fits This Methodology

7 "what we did" + 5 "what we didn't do"—Leansight's product DNA is simultaneously written into these 12 principles

Methodology Principles × Leansight Product Fit Matrix

Each principle has corresponding product capability support—not "barely can do," but "born to do this"

Principle Customer M Experience Core Leansight Product Fit Logic
Demand-Driven Apple requirements + GM data-driven CEO Headline Make data proactively find people, achieving top-down pull
Org Enablement IT-OT fusion center + rotation + performance alignment LeanFusion LeanCodee IT/OT collaborate on same platform, low-code lowers cross-boundary threshold
Vendor Selection Service support + productization, outcome-driven Leansight Platform Three standardized platforms + tower-style continuous co-building
Architecture Strategy Big systems stable, small needs low-code fast LeanCodee LeanBI Standard interfaces connect MES/ERP, low-code builds agile upper layer
Talent Evolution Big data → low-code → Agent company-wide training LeanFusion LeanCodee CEO Headline Product evolution fully synchronized with talent upgrade, capability co-running
Implementation Rhythm Small-step iteration, 50→2000+ dashboards LeanCodee LeanBI Lego-style components + dashboard templates, MVP validation then rapid replication
Strategic Benchmarking Smart Factory application + annual theme Four-Layer Architecture Benchmark against Smart Factory standards layer by layer, annual planning service

5 Red Lines × Leansight's Natural Avoidance

Leansight's product model happens to avoid exactly what Customer M firmly refused to do

Red Line What Customer M Didn't Do Leansight Avoidance Why Naturally Avoided
✕ Trusting PPT Focus on actual results, look at deliverables MVP First Leansight insists on running MVP before signing contracts, speaks with deliverables
✕ Only Submitting Requirements Innovative projects bring in service provider early to co-build Co-Build Delivery Leansight is on-site from requirements research phase, not "order-taking development"
✕ Detached from Business All proposals business-driven LeanBI CEO Headline Dashboard usage = go-live criteria; Headline outputs business insights
✕ Blindly Choosing Big Vendors Big vendor service support capability limited Deep Co-Running Leansight team size controllable, short response radius, continuous co-building without rotating people
✕ Copying Methodology Verbatim Framework right, content iterated LeanCodee Platform framework standardized, scenario applications iterated through low-code

Customer M validated 7 "what we did" and 5 "what we didn't do" over 8 years; Leansight achieved all 12 simultaneously through product capability.
Methodology is the Way, products are the Art, restraint is the Boundary—Way, Art, and Boundary unified is the complete answer to digital transformation.