🔧 Techniques and Tools

The d.school's 5-phase model: Empathize, Define, Ideate, Prototype, and Test. It's not linear — you revisit phases as many times as needed.

Having a shared framework reduces language conflicts between teams and speeds up alignment on the problem before discussing a solution.

How Might We, empathy map, point of view (POV), low-fidelity prototype, test success criteria.

Observation, in-depth interview, and shadowing techniques to capture what users CANNOT articulate in questionnaires.

Most products fail because they solve the wrong problem. Real empathy drastically reduces wasted development effort.

“5 whys” interview, shadowing, diary study, user journey map, insight vs. observation.

Phase for generating the maximum number of ideas without judgment, using techniques like Crazy 8s, brainwriting, and forced analogies.

Teams that jump straight to solutions come up with average ideas. Divergent-convergent thinking produces solutions no one would have come up with alone.

Divergent vs. convergent thinking, Crazy 8s, dot voting, impact x effort matrix.

Build quick, inexpensive representations (paper, cardboard, wireframe) to test specific hypotheses before investing in actual development.

A R$20 prototype that reveals a critical problem saves months of misguided development. Too much fidelity creates attachment that prevents honest feedback.

Fidelity vs. focus, paper prototype, storyboard, wizard of oz, paper vs. digital prototype.

Build and facilitate teams with T-shaped profiles (specialists with a broad perspective) from different areas to tackle the problem from multiple angles.

Cognitive diversity improves the quality of solutions. The facilitator creates a safe environment where conflict over ideas is productive, not personal.

T-shaped profile, facilitation vs. management, IDEO brainstorming rules, "one conversation at a time," defer judgment.

How to adapt the DT process to contexts with constraints on hierarchy, budget, and timelines without turning it into "Post-it theater."

Corporate design thinking fails when it becomes a ritual without a commitment to iteration. Knowing how to navigate internal politics is what separates real practitioners from workshop consultants.

IBM Enterprise Design Thinking, Hills statement, sponsor user, post-workshop feedback cycle, Go/No-Go criteria.

The central loop of Lean Startup: build a minimum experiment (MVP), measure real data, learn, and decide whether to pivot or persevere.

Reduces the waste of building a complete product before validating the most critical hypothesis. Each loop iteration accelerates validated learning.

Validated learning, value hypothesis, growth hypothesis, cycle time, minimum batch size.

A version of a product that lets you gather the most validated learning about customers with the least effort.

An MVP isn't a bad product: it's an experiment. Understanding this shifts the mindset from "launch" to "learning," reducing risks.

Concierge MVP, Wizard of Oz MVP, landing page test, smoke test, activation vs. vanity metrics.

AARRR Framework: Acquisition, Activation, Retention, Revenue, and Referral — the 5 funnel stages that reveal where the product is bleeding.

Vanity metrics (pageviews, downloads) hide real problems. AARRR forces you to measure actual user behavior at each stage.

North Star Metric, cohort analysis, churn rate, retention curve, NPS as a referral proxy.

The moment of making a structured, data-driven decision to change a fundamental assumption (pivot) or continue on the current path (persevere).

Without clear criteria, the team perseveres out of inertia or pivots out of panic. Validated learning data provide a basis for rational decisions.

Zoom-in pivot, zoom-out pivot, customer segment pivot, platform pivot, predefined pivot criteria.

An adaptation of Lean Startup for the corporate context, where compliance, brand, and process constraints require adjustments to the pace of experimentation.

GE, Intuit, and Toyota apply variations of Lean Startup internally. Knowing how to adapt it lets you use the method without breaking governance.

FastWorks (GE), Design for Delight (Intuit), innovation accounting, stage-gate vs. continuous iteration.

The most common mistakes: an MVP that doesn’t validate a real hypothesis, pivoting in a panic without enough data, and vanity metrics disguised as KPIs.

Poorly applied Lean leads to fast cycles of learning the wrong things. Recognizing the anti-patterns saves months of misdirected effort.

MVP that's too minimal (doesn't validate), pivot mania, vague hypothesis, cohort vs. total aggregate, analysis paralysis.

Monday: map and choose a focus. Tuesday: sketch solutions. Wednesday: decide and storyboard. Thursday: prototype. Friday: test with 5 real users.

5 days may seem like little time, but the forced pace eliminates decisions by endless consensus. The deadline creates a creative constraint that accelerates quality.

Long-term goal, sprint question, Lightning Demos, crazy 8s in the Sprint, 8-frame storyboard, tester recruitment.

Build a map of the problem (stakeholders, flow, pain points) and formulate the “Sprint Question”—the most critical hypothesis the sprint will test.

A Sprint without a clear focus produces a prototype that answers nothing. The Sprint Question is the compass that ensures Friday’s test validates what really matters.

Sprint map, How Might We in the context of the Sprint, target selection on the map, decider vs. facilitator.

Build a prototype that looks real enough for users to suspend disbelief and react as if it were the real product—in just 1 day.

A Sprint prototype doesn't need to work — it needs to provoke authentic reactions. Knowing the threshold between “realistic enough” and “too finished” is a critical skill.

Goldilocks quality, Keynote/Figma as a prototyping tool, division of work among Maker/Writer/Asset, prototype rehearsal.

Structured 60-minute interviews with 5 representative users while the team watches from a separate room and records patterns in real time.

Nielsen Norman Group shows that 5 users reveal 85% of usability problems. More interviews have rapidly diminishing returns.

Interview guide, think-aloud protocol, one sticky note per room, recruitment sniff test, pattern analysis in 1h.

Comparison of the 3 frameworks: when to use each, how they work together, and how to avoid confusion between each methodology’s vocabulary and pace.

Teams that mix all 3 without a clear rationale create dysfunctional hybrids. Knowing the purpose of each tool is a prerequisite for using them well together.

Discovery sprint vs. delivery sprint, DT for problem framing, Sprint for validation, Agile for development, double diamond.

Sprint adaptations for distributed teams: digital tools, pace adjustments, remote facilitation techniques, and how to maintain the energy of 5 days in an asynchronous-synchronous format.

Post-pandemic, most product teams are hybrid or remote. Adapting the Sprint to this context without losing effectiveness is a direct competitive advantage.

BOARD/Miro for voting, Loom for async Lightning Demos, Lookback.io for remote testing, video fatigue, 4+1 rhythm (4 intense days + a light Friday).

IDEO’s 7 brainstorming rules: defer judgment, encourage wild ideas, build on others’ ideas, stay focused on the topic, one conversation at a time, be visual, and aim for volume.

Brainstorming without rules produces average ideas dominated by the most extroverted people. The rules create a safe space for unusual ideas to emerge.

Production blocking, social loafing, brainwriting as an alternative, quantity over quality, HMW as a warm-up.

SCAMPER: Substitute, Combine, Adapt, Modify/Magnify, Put to other uses, Eliminate, Rearrange/Reverse. Each verb is a forced angle on the existing product/service.

When the team is stuck on incremental improvements, SCAMPER forces it out of its mental comfort zone with structured prompts that break the pattern.

Osborn checklist, functional fixedness, forced analogy, reversing assumptions, examples for each SCAMPER letter using real products.

5–10-minute exercises to get the team into creative mode: 30 Circles, Worst Possible Idea, Yes And!, and other improv games adapted for business.

Jumping straight into a product brainstorm in a corporate meeting brings a judgment mindset. A warm-up calibrates the mental state for divergence in minutes.

30 circles (IDEO), Worst Possible Idea (reversal), Yes-And (improv), flow state, social inhibition vs. psychological safety.

Convergence techniques: dot voting, impact x feasibility matrix, effort x novelty 2x2, and how to present selection criteria WITHOUT killing the creative energy.

Poorly facilitated selection brings back HiPPO (Highest Paid Person's Opinion) and kills ideas that have no “sponsor” in the room.

The HiPPO effect, dot voting, a 2x2 matrix, “how weird is it” as a novelty criterion, a portfolio of ideas vs. a single bet.

Skills for facilitating ideation sessions: space design, time management, handling blocks, energizing tired groups, and neutralizing domineering participants.

The same exercise with different facilitators produces drastically different results. Facilitation is a learnable skill that multiplies the value of any technique.

Facilitator vs. participant, physical/digital environment design, timeboxing, "parking lot" for tangents, productive silence technique.

Use analogies from distant fields (nature, sports, the military, the arts) to transfer solutions to the problem at hand—the Synectics method by Gordon.

Analogies force the brain to step outside its usual frame for a problem. Biomimicry, for example, has generated engineering solutions by copying billions of years of optimization in nature.

Synectics, direct vs. personal analogy, biomimicry, De Bono’s lateral thinking, and the 6 Thinking Hats applied to ideation.

Jobs are functional (getting something done), social (how others perceive us), and emotional (how we feel). All 3 dimensions are present in every purchase decision.

Focusing only on the functional job ignores 2/3 of the real motivation. Christensen's milkshake study revealed that the morning job was "keeping a hand busy and avoiding hunger until lunch."

Canonical job statement (when [situation], I want [motivation], so I can [outcome]), functional/social/emotional dimensions, hiring context.

Interview customers about the specific moment they decided to buy/hire to reveal the real job, not what they consciously say.

Traditional research asks, "What do you want?" JTBD asks, "Tell me about when you decided to hire/fire product X." The story reveals the real job.

Switch interview (Intercom), forces of progress, pushing/pulling/anxieties/habits, decision timeline.

Tony Ulwick’s framework that turns jobs into measurable “desired outcomes”—metrics people use to evaluate success when completing the job.

ODI connects JTBD with quantification: which outcomes are most underserved (high importance, low current satisfaction) indicate where innovation will yield the greatest returns.

Opportunity score, desired outcome statement, importance x satisfaction matrix, 8-step job map, overserved vs. underserved.

Group customers by similar jobs rather than by age or income bracket—people with opposite demographic profiles may have the same job and respond to the same product.

Traditional demographic segmentation is imprecise for predicting purchasing behavior. Job-based segmentation improves communication accuracy and feature prioritization.

Job-based segmentation, “same job, different persona,” context switching, non-obvious competitors (the milkshake vs. a banana vs. eating nothing).

When the focus is on the job, the competition expands: Netflix competes with sleep, Spotify with conversations among friends, Duolingo with Instagram. “Doing nothing” is always a competitor.

Companies that only look at direct competitors are disrupted by alternatives that weren’t on their radar. JTBD reveals the real competitive landscape.

Jobs-based competition map, nonconsumption as a market, workaround solution, “fire the product” (firing the product), Christensen’s theory of disruption.

How to use JTBD insights to prioritize features, define positioning, and build a job-oriented roadmap instead of one driven by requests from vocal users.

A feature roadmap without JTBD is a wish list. With JTBD, every feature has a clear rationale: "Which job does this feature help you do better?"

Jobs-based roadmap, opportunity solution tree (Teresa Torres), “outcome over output,” jobs as the unit of value measurement, retiring features because a job is obsolete.

Objective = an inspiring qualitative direction. Key Results = 3 to 5 quantitative metrics that prove the objective was achieved. Without metrics, it’s a mission; without direction, it’s a standalone KPI.

Most poorly written OKRs are disguised KPIs ("increase sales by 10%"). Innovation OKRs require aspirational goals that go beyond the status quo.

Moonshot OKRs (70% is success), rooftop OKRs (operational), leading vs. lagging indicators, input vs. output OKRs, the rule of "how will we know we got there".

Innovation (exploration) and operations (exploitation) OKRs have completely different cadences, failure tolerances, and success criteria—and they need to coexist.

Applying the same operating OKR to innovation teams is a recipe for failure: risk tolerance and progress metrics are fundamentally different.

Organizational ambidexterity through OKRs, McKinsey’s Horizon 1/2/3 as an OKR framework, quarterly vs. annual cadence, learning OKRs for innovation.

OKRs should not be “cascaded” top-down like traditional goals—the ideal is for 50-60% to be defined bottom-up and aligned with the company’s objectives.

Cascaded OKRs become bureaucracy in disguise. The system’s power lies in autonomous alignment: each team contributes to the larger objective with the freedom to define the “how.”

Bidirectional OKRs (John Doerr), alignment vs. cascading, CFR (Conversations, Feedback, Recognition), OKRs as a conversation, not a contract.

The rituals that make the system work: quarterly planning (write), weekly check-in (update confidence), retrospective (learn), final grading (score without punishing).

OKRs without check-in rituals become forgotten documents. The rituals are what turn the framework into a habit of execution and continuous learning.

Confidence rating (0.0-1.0), 0.7 grading as success, “set and forget” as an anti-pattern, OKR scoring without salary consequences, celebration of learning.

Why tying OKRs to bonuses destroys the system: teams set easy goals (sandbagging), lose ambition, and OKRs become a political game to protect pay.

Google, Intel, and most success stories explicitly separate OKRs from performance reviews. Understanding why is essential to defending the system at your company.

Sandbagging, grade inflation, OKR vs. MBO (Management by Objectives), separating OKRs from performance, “John Doerr’s superpowers”: focus, alignment, tracking, stretch.

How to structure an OKR pilot in 1 quarter: choose 1–2 teams, facilitate the writing process, define minimum rituals, and gather learnings before scaling.

Implementing OKRs company-wide all at once is the fastest way to get rejected. A controlled pilot builds organizational confidence and allows adjustments without risk.

OKR champion, tracking software (Lattice, Perdoo, Google Sheets), common resistance (“another methodology”), a quick win to generate buy-in, rollout in the second quarter.

Genrich Altshuller analyzed 400,000 Soviet patents and discovered that 95% of invention problems had already been solved in another field — cataloging 40 recurring inventive principles.

TRIZ shows that creativity has structure. Instead of waiting for spontaneous insight, you apply cataloged principles to solve technical and physical contradictions.

40 inventive principles, contradiction matrix, Ideality (ideal solution), Contradiction Matrix, levels of innovation (1 to 5).

TRIZ identifies 2 types: technical contradiction (improving A makes B worse) and physical contradiction (the same element needs to have property X and not-X at the same time).

Most technical problems are disguised trade-offs. Precisely naming the contradiction points to the right inventive principles and avoids mediocre compromise solutions.

Technical contradiction (39x39 matrix), physical contradiction (separation in space/time/condition), IFR (Ideal Final Result), segmentation principle.

Jacob Goldenberg and Roni Horowitz's SIT (Systematic Inventive Thinking) distills TRIZ into 5 templates: Subtraction, Multiplication, Division, Task Unification, and Attribute Dependency.

SIT can be applied to product and service innovation without an engineering background. Harvard Business Review documented that SIT produces more actionable ideas than free-form brainstorming.

5 SIT templates, "closed world" assumption, function follows form (the inverse of DT), product examples (iPod, Bic pen, Swiffer).

Platforms like InnoCentive, Kaggle, and Topcoder show that external communities solve complex technical problems faster and more cheaply than internal R&D.

Eli Lilly, NASA, and P&G demonstrated massive ROI from crowdsourcing innovation. The challenge isn’t the platform — it’s framing the problem so an outsider can solve it.

InnoCentive case studies, prize vs. contract as motivation, problem decomposition for crowds, intellectual property in crowdsourcing, quality filters.

Research by Adam Grant and Charlan Nemeth shows that individual creativity produces more novelty, while collective creativity produces more refinement. Session design should reflect the goal.

Groupthink is real. Teams that always brainstorm together converge on the group's average. Alternating individual and group cycles maximizes the quality and quantity of ideas.

Brainwriting 6-3-5, solo → group → solo cycle, devil’s advocate as a structured role, productive dissent, Charlan Nemeth’s research on creative minorities.

Csikszentmihalyi, Amabile, and neuroscience research show that physical and digital space affects creative output: color, noise, light, ceiling height, and “accidental distraction” influence results.

Creativity doesn’t emerge from the individual alone—it emerges from the interaction between the individual and the context. Designing the innovation environment is an underused lever with measurable ROI.

Innovation labs, "messy desk" effect, ambient noise at 70dB, natural light, Pixar campus design, "happy accidents" as intentional design, transition space.

The prototyping continuum: paper (low-cost, fast), wireframe (structure), mockup (visual), interactive prototype, proof of concept (technical), MVP (minimal production version).

The right level of fidelity depends on the question. High fidelity too early kills feedback; low fidelity too late doesn’t validate what needs to be validated. The right mapping saves weeks.

Paper prototype, lo-fi/hi-fi wireframe, Figma prototype, Storybook, technical spike, production MVP, "prototype as a question".

Techniques specific to hardware (cardboard, foam, 3D printing, Arduino), software (Figma, Framer, no-code), and services (role-play, service blueprint with actors, Wizard of Oz).

Each domain has its own tools. Innovators who move between hardware, software, and services multiply their prototyping options and speed up validation.

Cardboard prototype, Arduino prototype, no-code (Bubble, Webflow), rough service blueprint, role-play for services, prototype "smell test."

Before prototyping, formulate the hypothesis the prototype will test. “Users will pay X for Y” is the hypothesis; the prototype is the experiment that confirms or refutes it.

Without a clear hypothesis, the test produces friendly data. The question determines the required level of fidelity: testing visual appeal requires visuals; testing a workflow requires interactivity.

Risky assumption first, hypothesis, test card (Strategyzer), learning card, predefined pass/fail criteria, invalidation vs. partial validation.

Document just enough to preserve what you learned without creating bureaucracy that slows iteration: a photo of the prototype + 3 learning bullets + a decision = an adequate record.

Teams that document too much iterate slowly. Teams that document too little repeat mistakes. The minimum viable documentation is the balance point you need to find.

Learning card, "just enough documentation," version in the file name vs. Git, assumption graveyard, one-paragraph decision log.

The organizational ability to discard a prototype that didn’t validate without drama—treating the discard as valuable information, not failure.

Attachment to invested effort (the sunk cost fallacy) is the biggest enemy of rapid iteration. Teams with a culture of letting go reach the right product much faster.

Sunk cost fallacy, “kill your darlings,” prototype as hypothesis (not product), celebrating invalidation, “what did we learn” as a closing ritual.

Corporate constraints (compliance, legacy system integration, IT approval) change what can be prototyped and how quickly—and how to navigate this without getting stuck.

Ignoring corporate constraints leads to prototypes that never make it to production. Understanding the "safe sandbox" for prototyping within large companies is the difference between theory and practice.

Innovation sandbox, shadow IT vs. approved IT, compliance-first prototype, "ask forgiveness vs. ask permission," internal user proxy as the first customer.

Reinforcing loops (R) amplify changes—virality, exponential growth. Balancing loops (B) stabilize systems—controlled churn, budget constraints.

Understanding the type of loop that governs your product or process is essential for knowing where to intervene and predicting unintended effects of changes.

Reinforcing loop (R), balancing loop (B), System Dynamics (Forrester), loop delay, Donella Meadows’ leverage point.

Leading indicators predict the future (onboarding speed, depth of engagement); lagging indicators confirm the past (revenue, quarterly NPS). Innovative organizations measure both.

Making decisions based only on lagging indicators is like driving by looking in the rearview mirror. Leading indicators let you correct course before the problem shows up in the results.

Leading vs. lagging, predictive vs. retrospective metrics, "health metrics" vs. business KPIs, time-to-value as a leading indicator of retention.

Systems that continuously collect user feedback (in-app surveys, intercept interviews, session replay, support ticket mining) instead of quarterly research rounds.

Quarterly satisfaction surveys are too slow for digital products. Continuous feedback lets you detect problems in hours, not weeks, reducing churn before it turns into a crisis.

Micro-surveys (Hotjar, Sprig), session replay (FullStory), ticket mining, continuous discovery (Teresa Torres), and weekly user interviews as a ritual.

Structured internal feedback rituals: agile retrospective (what went well / could improve / try), blameless post-mortem, pre-mortem (imagine the failure in advance).

Teams without internal feedback rituals repeat the same mistakes. Effective retrospectives increase the pace of continuous improvement in the innovation process itself.

Blameless post-mortem (Google SRE), pre-mortem (Gary Klein), 4Ls retrospective (liked/learned/lacked/longed for), action items with an owner and deadline, “team radar.”

The time between a hypothesis and the data that confirms or refutes it. Amazon reduces this with “two-pizza teams.” Spotify does it with squads. Netflix uses feature flags and continuous deployment.

A company that runs 1 experiment per month loses to one that runs 100. Learning speed is the only sustainable competitive advantage in the long run (Bezos, Hastings).

Cycle time, DORA metrics, deployment frequency, feature flags, A/B testing as learning infrastructure, "day-1 mentality" (Amazon).

How to turn a team’s learning into organizational knowledge: decision wikis, “lunch and learns,” experiment repositories, and internal knowledge-sharing systems.

Learning that doesn’t spread across the organization is wasted. A learning organization (Peter Senge) has a compounding advantage: every well-documented experiment speeds up the next ones.

Learning organization (Senge), after-action review (U.S. military), experiment repository, knowledge half-life, "build the knowledge, not the product."

Rituals are behaviors repeated with intention and meaning. In innovation, they institutionalize practices that would otherwise depend on conscious mental effort.

Neuroscience shows that habit (not motivation) sustains behavior over the long term. Without rituals, innovation techniques become one-off events with no cultural impact.

Habit loop (Charles Duhigg), cue-routine-reward, keystone habit, identity vs. outcome (James Clear), ritual as a signal of cultural belonging.

Examples of short-cadence rituals: innovation daily stand-up (what did I observe in the market today?), weekly trend news digest, a shared “experiment of the week.”

Daily/weekly rituals are the most powerful because frequent repetition builds habits faster. Five-minute daily micro-rituals build a mindset more than quarterly workshops.

Morning Pages (Julia Cameron), a daily "one thing," a weekly digest of weak signals, 15-minute peer learning, "what surprised you this week?" as an opening ritual.

Higher-frequency rituals: monthly internal shark tank (teams pitch experiments for funding), quarterly hackathon, "innovation showcase" to celebrate learnings (not just successes).

Monthly/quarterly rituals create accountability milestones and visibility for innovation initiatives. Without milestones, projects die quietly without anyone noticing.

Internal shark tank, corporate hackathon, innovation showcase, “failure resume” as a learning ritual, quarterly experiment portfolio review.

Rituals that institutionalize learning from failure: "Failure Resume" (résumé of mistakes), corporate "FuckUp Nights," a gallery of failed experiments with documented lessons.

A culture that punishes public mistakes creates risk aversion. Rituals that celebrate informed failure signal that failing fast and cheaply is a valued skill, not a weakness.

FuckUp Nights (global movement), failure resume (Tim Harford), "heroic failure" award (Tata Group), blameless culture, lesson learned as a deliverable from failure.

Rituals that bring customers into the innovation process: "customer for a day," resident customer panel, monthly customer advisory board, open usability lab.

Companies that interact with real customers more than once a month make product decisions with 2-3x more accuracy (Bain & Company, 2022). Contact is the antidote to the “ivory tower.”

Customer for a day, customer advisory board (CAB), continuous discovery (Teresa Torres), "follow me home" (IDEO), co-creation vs. traditional research.

How to keep innovation rituals from turning into bureaucracy: rotate facilitators, evolve the format every 6 months, track ritual vitality metrics, and set criteria for discontinuing them.

Rituals that don't evolve become pointless obligations. The difference between a living culture of innovation and a dead program is the ability to renew the ritual itself periodically.

Ritual lifecycle (launch, maturity, renewal, discontinuation), "ritual health check," rotation of stewards, energy as a vitality metric, sunset criteria.