Where Design Thinking Works — and Where Another Tool Fits Better

What Design Thinking assumes

Over the past fifteen years, Design Thinking—popularised by Tim Brown (Harvard Business Review, June 2008) and the Stanford d.school—has become the default innovation method taught in corporate academies around the world. Its core sequence is familiar to almost every manager: Empathize, Define, Ideate, Prototype, Test.

When applied to the problems it was built for, Design Thinking works remarkably well. In McKinsey's five-year study of 300 publicly listed companies (The Business Value of Design, October 2018), companies in the top quartile of the McKinsey Design Index grew revenues at roughly twice the rate of their industry peers.

Every problem-solving method, however, rests on a set of hidden starting assumptions. Design Thinking assumes that:

1. The primary unknown is an unmet or unarticulated user need that ethnographic observation and customer interviews will uncover.
2. The solution space is relatively open, allowing the team to diverge widely ("defer judgment") and invent new features, interfaces or service experiences.
3. Rapid, low-cost "fail fast" prototyping with end users is safe and feasible before committing capital.

Four operational conditions that break the fit

Walk out of a consumer digital studio and onto a petrochemical plant floor, a hospital ward, an automotive injection line or a bank's regulated back office, and those three starting assumptions frequently invert. Four conditions routinely break the fit:

1. You already know what the customer wants. In industrial turbine maintenance (Case 02) or convenience-store expansion (Case 01), interviewing the customer for three weeks will not reveal a secret psychological insight. The customer wants the turbine back online faster and the store opened on schedule at the agreed cost. The unknown is not what they value; it is how to rearrange your physical constraints to deliver it.
2. You cannot switch off or replace the existing asset. If a manufacturer has $40 million tied up in an injection-moulding line or cold-storage facility, an ideation workshop that proposes "redesigning the business model around a new platform" is commercially useless on Monday morning. The solution must work with the machines, permits and headcount already inside the building.
3. There is no budget or tolerance for "failing fast" in live operations. You cannot run a half-baked prototype on a high-voltage turbine, a regulated tax ledger or an Alzheimer outpatient ward where a single failure means a patient wanders into traffic.
4. The requirement arrives as a hard B2B specification or regulation. When an OEM customer or regulator sets a strict tolerance, the team does not need divergent brainstorming; it needs a systematic way to eliminate a physical contradiction inside a closed system.

Matching the method to the constraint: Outside-In vs. Inside-Out

Rather than treating Design Thinking as a universal hammer, effective organisations match the method to where the primary unknown sits:

What this looks like in practice

Three of our published cases illustrate what happens when a team switches from an outside-in "add a new feature or technology" reflex to an inside-out constraint method:

1. Public hospital Alzheimer ward (Designed)

In Patient Safety Through the Waiting Area, Not a Tracking Device, a public hospital needed to prevent outpatient Alzheimer patients from wandering out of the clinic while their family caregivers queued at the pharmacy or cashier. An outside-in brainstorming approach naturally points toward wearable GPS trackers, Bluetooth wristbands or smart-door sensors—solutions that require capital budget, battery charging, sanitising and staff monitoring that an overstretched public hospital cannot sustain.

Looking strictly inside the Closed World of the ward's existing rooms, nurses, caregivers and uniforms led to a zero-tech structural change: a supervised waiting area inside the ward with a formal nurse handover and pickup ID card. Under the new rule, anyone inside the zone not wearing a hospital uniform is immediately identifiable as a patient, reducing ongoing hardware and maintenance cost toward zero.

2. Automotive plastic injection quality (Client-measured)

In Defect Caught Before the Colour Goes On, the problem was not an unknown customer need; it was the high cost of coloured plastic scrap on an injection-moulding line. Applying the SIT Multiplication and Division tools to the existing inspection step—placing a quality check immediately before the irreversible colouring step—allowed uncoloured defective parts to be reground or bought back by the resin supplier, cutting the annual cost of waste by ~75% in the client's own accounting.

3. Finance travel-expense workflow (Designed)

In The Expense System That Did Not Need to Be Built, a user-journey workshop would have focused on making the expense-claim app friendlier for travellers to upload receipts and fix OCR errors. Starting from the existing components in the transaction environment showed that a company travel card statement already contained every verified foreign-exchange and merchant record the auditor needed—eliminating the custom software build altogether.

Before you conclude the tool is wrong: Fit vs. Execution

When a Design Thinking programme fails to produce commercial results, leadership should ask two questions in order:

1. Was it an execution failure? Did the team treat Design Thinking as a two-day sticky-note theatre without real customer contact, without engineering or finance in the room, and without decision rights to kill or fund the output (Field Note 02)? If so, no method would have survived that governance vacuum.
2. Was it a method-fit failure? Did we point an open-ended, user-empathy method at a closed-world operational, manufacturing or cost constraint where adding new features was never affordable? If so, keep Design Thinking for front-end customer discovery and equip your operational and technical teams with Systematic Inventive Thinking (Field Note 10).

Where to start on Monday

Before booking your next training cohort or project kickoff, run this three-step check:

1. Classify your top five active problems by where the unknown sits. If the unknown is "What will customers value and adopt?", assign Design Thinking. If the unknown is "How do we hit a known cost, speed, quality or safety target using the assets and headcount we already have?", assign an inside-out constraint method (SIT, Lean/TOC or TRIZ).
2. Draw the "Closed World" boundary before any ideation session. Tell the team upfront which budgets, machines, regulatory rules and timelines are fixed for the next six months. Prohibiting "wish-list" ideas that require new capex or new IT platforms forces the team to examine the levers they actually control.
3. Pair both methods in sequence when a project spans front-end and back-end. Use Design Thinking first to confirm what the customer genuinely values (and which features they do not care about), then hand the validated requirement to an SIT session to deliver it using existing operational resources.

What this looks like in Thailand

Across Thailand's banking, retail, healthcare and public sectors, Design Thinking has been the single most widely funded capability programme of the past decade. Even the Thai public sector adopted it at scale: in a 6-month programme led by the Office of the Public Sector Development Commission (OPDC), UNDP and The GovLab across 63 officials from 10 Thai government agencies, an academic evaluation by Promsiri and colleagues (Sustainability, 2022) found strong gains in participant mindsets and workshop satisfaction—yet noted that translating workshop prototypes into implemented institutional processes remained the primary hurdle.

In Thai manufacturing, agro-industry and infrastructure groups, we see the same mismatch every quarter: engineers, plant managers and supply-chain leads are sent to consumer-style Design Thinking bootcamps, return to a factory governed by strict OEM specifications and zero extra capex, and quietly abandon the sticky notes within two weeks. Giving those technical teams a structured, constraint-based method that speaks the language of physical steps, parameters and existing equipment unlocks immediate traction.