Three-Lens Evaluation

Purpose

Evaluate each prototype through three complementary lenses—desirability (do users want it?), viability (can we build a business?), and feasibility (can we build and operate it?)—requiring all three to pass before recommending go. The three-lens framework prevents the most common evaluation failure: strong performance on one dimension masking fatal weakness on another.

Single-lens evaluation produces predictable failures:

• Desirable but not viable: Users love it; no business model (Google Wave, many social apps)
• Viable but not desirable: Business model works on paper; users reject it (subscription fatigue, dark patterns)
• Desirable and viable but not feasible: Users want it, economics work, but technical complexity or operational cost makes delivery impossible at scale

The intersection of all three is innovation that succeeds.

When to Use

Use three-lens evaluation when:

• Conducting A3.4 evaluation after pilot completion (primary use case)
• Making any investment decision requiring multi-dimensional assessment
• Screening concepts at earlier stages (A2.4) with lighter-weight evidence
• Reviewing portfolio health at I1 level (are existing products still desirable, viable, feasible?)

Do NOT use when:

• Only one dimension matters (rare—but possible for internal tools where viability = cost savings only)
• Evidence is insufficient for even one lens (evaluation becomes speculation—extend A3.3 or declare “inconclusive”)

Sample Size and Duration

Duration: 3–5 days for analysis (within A3.4's 1–2 week timebox)

Team effort:

• Product Owner: 20–30 hours (synthesis, writing)
• UX Researcher: 8–12 hours (desirability evidence)
• Financial Analyst: 8–12 hours (viability modelling)
• Solution Architect: 8–12 hours (feasibility assessment)
• Total: 45–65 person-hours per prototype

Prerequisites

• Complete A3 data: A3.1 build learnings, A3.2 validation results, A3.3 pilot data all available
• Success criteria: Per-lens pass/fail thresholds defined (ideally from A2.4, confirmed before A3.4 starts)
• Cross-functional input: UX Researcher (desirability), Financial Analyst (viability), Solution Architect (feasibility)—Product Owner synthesises
• Benchmark data: Industry comparisons for key metrics (retention, NPS, unit economics)

Complete Procedure

Step~1: Desirability Assessment (4–8 hours)

Synthesise all user-facing evidence:

Beyond metrics, include qualitative evidence:

• User quotes demonstrating value perception (from think-aloud, interviews, NPS follow-up)
• Behavioural patterns from cohort analysis (retention curve shape, feature adoption)
• Unmet needs identified (user interview themes)

Desirability verdict: Pass / Marginal / Fail, with narrative explaining the verdict.

Step~2: Viability Assessment (4–8 hours)

Synthesise business and economic evidence:

• Unit economics: Customer Acquisition Cost (CAC) vs. Lifetime Value (LTV). Target: LTV/CAC ≥ 3:1
• Revenue potential: If pilot monetised, actual revenue and conversion rates. If not, willingness-to-pay signals from interviews
• Operational costs: Infrastructure, support, maintenance—actual from A3.3, projected for production
• Market size validation: Does pilot evidence confirm A1.4 market assumptions?
• Competitive position: Has the competitive landscape changed during A3?

Feed into ROI modelling (the referenced method) and sensitivity analysis (the referenced method) for financial projections.

Viability verdict: Pass / Marginal / Fail. Marginal is acceptable if a clear path to positive economics exists within 12~months.

Step~3: Feasibility Assessment (4–8 hours)

Synthesise technical and operational evidence:

• Build complexity: A3.1 actual effort vs. A2.4 estimates (was it harder than expected?)
• Technical performance: A3.3 uptime, response times, error rates vs. targets
• Scalability: Can the architecture handle 10×–100× pilot load? What changes are needed?
• Technical debt: What shortcuts were taken in A3.1 that must be addressed in A4?
• Support burden: A3.3 support ratio—is it sustainable at production scale? (from customer success management analysis, the referenced method)
• Regulatory/compliance: Any blockers identified during pilot?

Feasibility verdict: Pass / Marginal / Fail. Marginal means “feasible with known investment” (quantify the investment).

Step~4: Synthesise Across Lenses (2–4 hours)

Combine the three verdicts into an overall recommendation:

Key principle: A single “Fail” on desirability is almost always a kill—business models can be adjusted, technology can be rebuilt, but forcing users to want something they don't is the hardest problem in innovation.

Step~5: Confidence and Caveats (1–2 hours)

Rate overall confidence (1–10) in the recommendation. Document:

• Evidence gaps (what data is missing or weak?)
• Assumptions made (what did the analysis assume that could be wrong?)
• Risks (what could change the verdict?)
• Sensitivity (how much would key metrics need to change to flip the recommendation?—see the referenced method)

Quality Criteria

1. All three lenses assessed: No lens skipped or superficially treated
2. Evidence-based verdicts: Each lens verdict supported by specific metrics and qualitative data
3. Pre-defined criteria: Pass/fail thresholds established before evaluation, not adjusted to fit desired outcome
4. Cross-functional input: UX, finance, and engineering perspectives represented
5. Synthesis explicit: Overall recommendation follows logically from lens verdicts (synthesis matrix applied)
6. Confidence rated: Recommendation includes confidence level and evidence gaps
7. Caveats documented: Assumptions, risks, and sensitivity noted for A3.5 decision-makers

Theoretical Foundation

Seminal references:

• : Tim Brown (IDEO CEO) established the three-lens framework as the foundation of design thinking: innovation lives at the intersection of desirability, viability, and feasibility. This Venn diagram has become the most widely used innovation evaluation model globally.
• : Operationalised the three lenses into a practical evaluation methodology with specific questions, evidence types, and pass/fail criteria per lens. Positioned the framework as applicable from early concept screening (A2) through post-pilot evaluation (A3.4).

Contemporary references:

• : Applied the three lenses to product management: product managers must ensure desirability (PM + design), viability (PM + business), and feasibility (PM + engineering) —no single function owns all three.
• : Extended the framework with specific hypothesis types and experiments per lens, enabling teams to test each lens independently before A3.4 synthesis.

The Three Lenses

Challenges and Solutions

Challenge 1: Lens Weighting Disagreement

Symptoms: Engineering says “feasibility is marginal—too risky.” Product says “desirability is so strong it justifies the risk.” No framework for resolving the tension.

Solutions: Pre-commit to the synthesis matrix (Step~4) before A3.4 begins. The matrix makes the logic explicit: desirability pass + feasibility marginal = conditional go with tech investment. The framework decides, not the loudest voice.

Challenge 2: Cherry-Picking Within Lenses

Symptoms: Desirability “pass” based on NPS alone, ignoring 45% Week~8 retention (below 60% target).

Solutions: Each lens requires all primary metrics to pass, not just one. If retention fails, desirability fails regardless of NPS. No metric trading within a lens.

Challenge 3: Marginal Everywhere

Symptoms: All three lenses are “marginal”—not clearly pass, not clearly fail. Team is stuck.

Solutions: Triple-marginal is a pause, not a go. The concept has potential but needs more evidence or iteration. Define specific conditions that would convert each marginal to pass (e.g. “If retention reaches 55% with improved onboarding, run 4-week extension pilot”). If conditions are unrealistic, the honest assessment is kill.

Relationship to Other Methods

Three-Lens Evaluation receives input from:

• Cohort Analysis (the referenced method): Retention data for desirability lens
• NPS (the referenced method Loyalty metric for desirability lens
• SUS (the referenced method Usability perception for desirability lens
• Customer Success Management (the referenced method): Support burden for feasibility lens
• User Interviews (the referenced method): Qualitative evidence across all three lenses