10 Think Toolkits to Test New Ideas Fast and Cheap

The ability to validate ideas quickly and inexpensively before committing significant resources is crucial for innovation success. These ten toolkits will help you design rapid, low-cost experiments that reveal whether ideas have merit without risking substantial time, money, or reputation.

1. The Assumption Extraction Framework

Identify and test critical assumptions underlying your idea before building anything.

How to apply it:

• List all assumptions: What must be true for this idea to work?
• Categorize assumptions: Desirability, feasibility, viability
• Rank by risk: Which assumptions, if wrong, kill the idea?
• Identify testable elements: What can be validated quickly?
• Design minimal tests: Smallest experiment that tests assumption
• Sequence strategically: Test highest-risk assumptions first
• Set decision criteria: What results would validate or invalidate?
• Think: "Ideas are bundles of assumptions—test assumptions, not complete ideas"

Assumption categorization:

Desirability (Do people want this?):

• "Customers have problem X"
• "Current solutions are inadequate"
• "People would pay for this solution"
• "The value proposition resonates"
• "Target audience is accessible"

Feasibility (Can we build this?):

• "We have/can acquire necessary skills"
• "Technology is available/achievable"
• "We can deliver at required quality"
• "Timeline is realistic"
• "Resources are obtainable"

Viability (Should we build this?):

• "Unit economics work"
• "Market is large enough"
• "We can compete effectively"
• "Business model is sustainable"
• "This fits our strategy/capabilities"

Assumption testing example:

Idea: Online fitness coaching for busy professionals

Critical assumptions:

1. Busy professionals struggle with fitness consistency
2. They prefer online to in-person coaching
3. They'll pay $200/month for service
4. We can deliver effective coaching remotely
5. We can acquire customers profitably

Risk ranking:

• Highest risk: #3 (pricing assumption)
• High risk: #2 (delivery preference)
• Medium risk: #1 (problem exists)
• Lower risk: #4, #5 (can test after validating market)

Test sequence: 3 → 2 → 1 → 4 → 5

2. The Smoke Test Constructor

Create the appearance of a solution to test demand before building anything real.

How to apply it:

• Create landing page: Describe offering as if it exists
• Include call-to-action: "Buy now," "Join waitlist," "Get early access"
• Drive targeted traffic: Small ad spend to ideal customers
• Measure conversion: How many people take action?
• Collect feedback: Talk to those who expressed interest
• Set success threshold: What % conversion validates demand?
• Decide based on data: Build, pivot, or abandon based on results
• Think: "Market demand is proven by actions, not opinions—make it testable"

Smoke test components:

Landing page elements:

• Compelling headline addressing problem
• Clear value proposition
• Features/benefits overview
• Pricing (even if approximate)
• Strong call-to-action
• Email capture for waitlist
• Optional: Video explanation

Traffic sources:

• Facebook/Instagram ads ($50-200 budget)
• Google Ads (targeted keywords)
• Reddit/forum posts (free if allowed)
• Direct outreach to potential customers
• Your existing audience/network

Metrics to track:

• Click-through rate (CTR) on ads
• Landing page conversion rate
• Email signups or purchase attempts
• Time on page, scroll depth
• Qualitative: Email responses, questions asked

Success criteria examples:

• 5%+ conversion to email signup = strong interest
• 1%+ conversion to purchase attempt = very strong
• 10+ quality conversations with interested prospects
• Compared to industry benchmarks for similar offers

Smoke test variations:

Fake door test:

• Add "coming soon" feature to existing product
• Track clicks to see interest level
• Simple implementation in existing system

Concierge MVP:

• Deliver service manually as if automated
• Test if people value outcome before building automation
• Example: Manual curation before building algorithm

Wizard of Oz test:

• Appears automated but is actually manual behind scenes
• Tests if solution works before building technology
• Example: "AI" that's actually human operators

3. The Minimum Viable Test Designer

Create the smallest possible experiment that yields meaningful learning.

How to apply it:

• Define learning goal: What specific question are you answering?
• Identify minimum test: Absolute smallest thing that answers question
• Remove everything else: Strip away non-essential elements
• Set clear metrics: How will you know if it worked?
• Define timeline: How long is "fast"? (Usually days/weeks, not months)
• Set budget limit: Maximum spend (typically $100-$1000)
• Execute quickly: Speed matters more than polish
• Think: "Don't build to test—test to decide whether to build"

MVT (Minimum Viable Test) design:

Question: Will people pay for this?

• Not MVT: Build full product, launch, see if sales happen
• MVT: Pre-sell before building, see if people actually pay

Question: Do people have this problem?

• Not MVT: Survey asking "Do you have problem X?"
• MVT: Observe behavior, track existing workarounds they use

Question: Which version works better?

• Not MVT: Build both fully, compare over months
• MVT: Prototype both, show to 20 people each, gather feedback

MVT principles:

Speed over perfection:

• Rough prototype > polished wrong thing
• Quick learnings > slow comprehensiveness
• Iteration beats initial accuracy

Focus over breadth:

• Test one thing at a time
• Clear variable isolation
• Specific learning objective

Action over opinion:

• What people do > what they say
• Real behavior > stated preferences
• Actual transactions > interest expressions

Examples by test type:

Demand test:

• MVT: Pre-sale offer, track purchases
• Time: 1-2 weeks
• Cost: Landing page (~$50) + ads ($100-500)

Solution test:

• MVT: Manual delivery to 5-10 customers
• Time: 2-4 weeks
• Cost: Your time + minimal tools

Pricing test:

• MVT: Show different prices to different segments
• Time: 1 week
• Cost: A/B testing tool (often free tier)

Channel test:

• MVT: Small spend across 3 channels
• Time: 1 week per channel
• Cost: $100-300 per channel

4. The Prototype Spectrum Selector

Choose the right fidelity level for your testing needs—from sketches to working prototypes.

How to apply it:

• Match fidelity to question: Higher fidelity only when needed
• Start lowest possible: Sketches before wireframes before prototypes
• Increase fidelity progressively: Only when lower fidelity can't answer question
• Consider audience: Who needs to understand this? What fidelity do they need?
• Balance speed vs. feedback quality: Lower fidelity = faster, but less realistic feedback
• Know your options: Range of prototyping approaches available
• Choose appropriate tools: Right tool for fidelity level needed
• Think: "Use the lowest fidelity that yields valid learning"

Fidelity spectrum:

1. Verbal description (Lowest fidelity)

• What it is: Explain idea in conversation
• When to use: Very early validation, concept testing
• Time: Minutes
• Cost: Free
• Good for: Initial reactions, obvious flaws

2. Sketch/wireframe

• What it is: Hand-drawn or basic digital sketch
• When to use: Testing concepts, flows, layouts
• Time: Hours
• Cost: Free to $20 (paper/pen or basic tool)
• Good for: Structure, layout, basic functionality

3. Clickable mockup

• What it is: Designed screens with simulated interaction
• When to use: Testing user experience, workflows
• Time: Days
• Cost: Free to $50 (Figma, InVision, Balsamiq)
• Good for: User journey, interaction patterns, design feedback

4. Facade/fake backend

• What it is: Real-looking interface, simulated data
• When to use: Testing UI with realistic experience
• Time: 1-2 weeks
• Cost: $100-500 (development time or no-code tools)
• Good for: Usability testing, realistic interactions

5. Functional prototype

• What it is: Working code, limited features
• When to use: Testing if solution actually works
• Time: 2-4 weeks
• Cost: $500-5000 (depends on complexity)
• Good for: Technical validation, real user testing

6. Minimum viable product (MVP)

• What it is: Basic working version with core features
• When to use: After validation, testing at small scale
• Time: 1-3 months
• Cost: $5,000-50,000+ (highly variable)
• Good for: Market validation, initial customers

Fidelity selection questions:

• "What do I need to learn?"
• "Who needs to interact with this?"
• "What's the simplest way to test this specific aspect?"
• "What's the risk of low-fidelity misrepresenting the idea?"

Example progression:

Product idea: Task management app for creative teams

Stage 1: Verbal description to 10 creatives → Mixed reactions Stage 2: Sketch of interface → Positive, requested specific features Stage 3: Clickable mockup → Tested workflow, identified friction Stage 4: Facade prototype → Users loved it, simulated using it Stage 5: Functional MVP → Built core features only Stage 6: Launch → Validated market with real users

Each stage costs 10x previous but provides 2x validation confidence.

5. The Pre-Sale Validation Method

Prove demand by selling before building through strategic pre-commitments.

How to apply it:

• Create compelling offer: Describe future solution as if available
• Set pre-order/waitlist: Collect commitments (ideally monetary)
• Offer early-bird incentive: Discount or special access for early commitment
• Drive targeted traffic: Reach ideal customers efficiently
• Track conversion rates: How many commit vs. browse?
• Engage with buyers: Talk to everyone who committed
• Set go/no-go threshold: "If X people pre-order, we build it"
• Think: "Money talks—pre-sales are the ultimate validation"

Pre-sale approaches:

Crowdfunding campaign:

• Platform: Kickstarter, Indiegogo
• Benefit: Built-in audience, credibility, all-or-nothing option
• Timeline: 30-60 day campaign
• Best for: Physical products, creative projects, clear concept

Direct pre-sale:

• Platform: Your own landing page, Gumroad, simple payment
• Benefit: Keep all proceeds, control timeline
• Timeline: Ongoing or time-limited
• Best for: Digital products, services, existing audience

Waitlist with deposit:

• Platform: Landing page with Stripe deposit collection
• Benefit: Serious commitment signal, refundable reduces risk
• Timeline: Flexible
• Best for: High-value items, B2B services

Letter of intent (B2B):

• Platform: Direct sales conversations
• Benefit: Enterprise validation, relationship building
• Timeline: Sales cycle dependent
• Best for: Business services, enterprise software

Pre-sale elements:

Offer structure:

• Clear description of what's being sold
• Delivery timeline (be realistic)
• Pricing (typically 20-50% discount for early adopters)
• What happens if project doesn't launch
• Limited quantity or time (creates urgency)

Trust building:

• Your background/credibility
• Why you're building this
• Progress updates commitment
• Refund policy if project fails
• Social proof (press, testimonials, early supporters)

Communication:

• Regular updates to pre-buyers
• Transparent about progress and challenges
• Engage for feedback
• Build community around project

Success metrics:

• Conversion rate: 2-5%+ is strong
• Total committed: Enough to justify building?
• Quality of buyers: Right target audience?
• Unsolicited interest: Organic shares, press inquiries

6. The Concierge MVP Strategy

Deliver service manually to test value before automating or scaling.

How to apply it:

• Define core value: What outcome are you delivering?
• Deliver manually: Do by hand what will eventually be automated
• Limit initial customers: 5-20 people to learn from
• Charge real money: Even small amounts validate value
• Document everything: What's hard, what works, what customers love
• Gather deep feedback: Close relationship with early users
• Identify automation priorities: What's most worth building?
• Think: "Prove people want the outcome before building the machine that delivers it"

Concierge MVP process:

1. Manual service delivery: Instead of building software/system, deliver outcome manually

• Curation: You select instead of algorithm
• Recommendations: You analyze instead of AI
• Matching: You connect people instead of automated system
• Content: You create custom instead of templated

2. Learn through doing:

• What takes time vs. what's quick?
• What do customers actually value vs. what you assumed?
• What patterns emerge from manual work?
• What's enjoyable vs. tedious?
• Where's the real value creation happening?

3. Build based on learning:

• Automate tedious, repetitive parts first
• Keep human touch where it matters
• Build what you've already proven works
• Skip features users didn't value

Concierge MVP examples:

Food & You (Personalized meal planning):

• Full automation vision: AI analyzes preferences, generates meal plans, creates shopping lists
• Concierge MVP: Nutritionist manually creates personalized plans for 10 clients
• Learning: Discovered clients valued weekly check-ins more than automated plans
• Result: Built hybrid model with automation + human touch

Talent marketplace:

• Full vision: Automated matching of freelancers to projects
• Concierge MVP: Manually matched 20 companies to freelancers
• Learning: Quality of match matters far more than speed
• Result: Built curated marketplace, not fully automated platform

Investment newsletter:

• Full vision: Algorithmic stock picks
• Concierge MVP: Manual research and weekly email to 50 subscribers
• Learning: Subscribers valued reasoning more than picks
• Result: Educational content became primary offering

Concierge advantages:

• Start immediately (no build time)
• Learn what actually matters
• Build relationships with early customers
• Get paid while learning
• Pivot easily based on learnings
• Only build what's proven necessary

7. The Landing Page Validation Engine

Use landing pages as fast, cheap validation tools for multiple aspects of your idea.

How to apply it:

• Create multiple landing pages: Test different angles/offerings
• Vary key elements: Headlines, value props, pricing, positioning
• Drive small traffic: $50-200 per variation for statistical significance
• Measure micro-conversions: Email signups, clicks, time on page
• A/B test systematically: Change one variable at a time
• Collect qualitative data: Exit surveys, chat transcripts, email responses
• Iterate rapidly: Launch, measure, adjust, relaunch in days
• Think: "Landing pages are cheap, fast laboratories for testing market hypotheses"

What landing pages can test:

Value proposition:

• Different benefits emphasized
• Alternative problem framing
• Various use cases highlighted
• Emotional vs. rational appeals

Audience:

• Small business vs. enterprise
• Industry A vs. industry B
• Role A vs. role B
• Different pain points

Pricing:

• Different price points
• Pricing models (subscription vs. one-time)
• Free trial vs. paid immediately
• Various tier structures

Positioning:

• Premium vs. accessible
• Innovative vs. reliable
• Simple vs. comprehensive
• Category associations

Offer structure:

• Service vs. product
• Done-for-you vs. done-with-you vs. DIY
• Individual vs. group
• Delivery method variations

Landing page testing framework:

Week 1: Create variations

• Landing page builder (Unbounce, Carrd, Webflow)
• 3-5 different variations
• Each tests specific hypothesis

Week 2-3: Drive traffic

• Facebook/Google ads ($50-100 per variation)
• Targeted to specific audience
• Track conversions carefully

Week 4: Analyze and iterate

• Conversion rates
• Traffic quality
• Qualitative feedback
• Create new variations based on learnings

Tools needed:

• Landing page builder: Carrd ($19/year), Unbounce ($90/month)
• Analytics: Google Analytics (free), Hotjar (free tier)
• Ad platforms: Facebook Ads, Google Ads
• Total cost: $300-500 for comprehensive testing

8. The Wizard of Oz Prototype

Make it appear automated while manually operating behind the scenes to test viability.

How to apply it:

• Design the interface: What users see and interact with
• Manually fulfill backend: Humans doing what automation would do
• Hide the manual process: Users experience seamless automation
• Test at small scale: 10-50 users maximum
• Measure outcomes: Does the solution deliver value?
• Learn operational requirements: What does fulfillment actually require?
• Decide automation priorities: What's worth building vs. keeping manual?
• Think: "Fake it until you validate it—manual behind-the-scenes reveals if automation is worth building"

Wizard of Oz examples:

Zappos (early days):

• Appeared as: Online shoe store with inventory
• Actually was: Founder bought shoes from local stores when orders came in
• Tested: Would people buy shoes online without trying them?
• Result: Yes → built real inventory system

"AI" customer service:

• Appears as: Instant AI-powered responses
• Actually is: Human support team responding quickly
• Tests: Do users value instant responses? What questions do they ask?
• Result: Learn patterns before building actual AI

Personalization engine:

• Appears as: Algorithmic content recommendations
• Actually is: Human curator making selections
• Tests: Do users engage with personalized content? What patterns work?
• Result: Understand curation logic before automating

Matching platform:

• Appears as: Automated matching between parties
• Actually is: Team manually reviewing and matching
• Tests: What makes good matches? Do users value matches?
• Result: Learn matching criteria before building algorithm

Wizard of Oz implementation:

1. Design user-facing experience:

• Interface that suggests automation
• Appropriate response times (not instant if that reveals human)
• Professional, polished presentation

2. Create manual fulfillment process:

• Document exactly what needs to happen
• Train team on how to respond
• Create quality standards
• Set realistic turnaround times

3. Monitor and learn:

• Track what requests come in
• Document edge cases
• Note what's hard vs. easy to fulfill
• Identify patterns in user behavior

4. Scale considerations:

• How many users can you handle manually?
• What breaks first as volume increases?
• What absolutely requires automation?
• What's better kept human?

9. The Target Customer Interview Protocol

Systematically gather qualitative insights through strategic conversations before building.

How to apply it:

• Identify ideal interviewees: People who have the problem you're solving
• Reach out directly: Personal outreach, not broad surveys
• Ask about behavior, not opinions: "How do you currently..." not "Would you..."
• Listen for pain points: What frustrates them about current solutions?
• Understand current workflows: How do they solve this problem now?
• Avoid pitching your idea: Learn first, share later if at all
• Aim for 20-50 conversations: Patterns emerge around 15-20
• Think: "Interview to learn, not to validate—seek truth, not confirmation"

Interview structure:

Opening (5 min):

• Thank them for time
• Explain you're researching [problem space]
• Ask permission to record (for notes)
• Set expectations (30-45 min conversation)

Problem exploration (15 min):

• "Tell me about how you currently handle [problem area]"
• "What's frustrating about that approach?"
• "What have you tried to solve this?"
• "What prevented those solutions from working?"
• "How important is solving this to you?"

Current solution deep-dive (10 min):

• "Walk me through the last time you dealt with this"
• "What tools/services do you use?"
• "How much time/money do you spend on this?"
• "What do you love about current approach?"
• "What would make this significantly better?"

Context understanding (10 min):

• "Who else is involved in this process?"
• "How does this fit into your broader workflow?"
• "What constraints do you operate under?"
• "What outcomes matter most to you?"

Closing (5 min):

• "Is there anything I should have asked but didn't?"
• "Would you be open to a follow-up conversation?"
• "Do you know others I should talk to?"
• Thank them genuinely

Interview best practices:

Do:

• Ask open-ended questions
• Follow interesting threads
• Dig into specifics ("Tell me more about that")
• Take detailed notes
• Listen 80%, talk 20%

Don't:

• Pitch your solution
• Lead the witness ("Wouldn't it be great if...")
• Accept vague answers ("sometimes, usually")
• Argue with their experience
• Rush through to hit all questions

What to listen for:

• Frequency: "I deal with this daily/weekly"
• Intensity: Emotional reactions, strong language
• Workarounds: Elaborate solutions they've created
• Spending: Time/money currently invested
• Urgency: "I need this now" vs. "Would be nice"

Interview analysis:

• Pattern identification: What themes repeat across conversations?
• Segmentation: Do different types of users have different needs?
• Pain point ranking: Which problems are most acute?
• Solution validation: Do existing ideas resonate?
• Opportunity discovery: Problems you hadn't considered

10. The Rapid Iteration Cycle

Design systematic learning loops that compress validation timelines dramatically.

How to apply it:

• Set weekly iteration goals: What will you learn this week?
• Design quick experiments: 1-7 day cycles, not months
• Measure clearly: Specific metrics or learning objectives
• Analyze immediately: Review results within 24 hours
• Decide quickly: Continue, pivot, or stop based on data
• Document learnings: Build institutional knowledge
• Iterate relentlessly: Multiple cycles reveal truth faster than perfect single test
• Think: "Ten one-week experiments teach more than one ten-week experiment"

Rapid iteration framework:

Monday: Design

• What are we testing this week?
• What's the experiment?
• What metrics indicate success/failure?
• What resources needed?

Tuesday-Thursday: Execute

• Run the experiment
• Gather data
• Monitor in real-time
• Document observations

Friday: Analyze

• Review results against success criteria
• Identify learnings
• Document what worked/didn't
• Spot patterns with previous iterations

Weekend: Decide

• Continue this direction?
• Pivot to different approach?
• What do we test next week?
• Plan Monday's iteration

Iteration velocity tactics:

Ruthless scope limitation:

• Test one thing at a time
• Remove nice-to-haves
• Accept imperfection
• Focus on learning, not polish

Pre-commit to decisions:

• Before experiment: "If X happens, we'll do Y"
• Remove decision paralysis
• Act on data immediately

Parallel testing:

• Run multiple small experiments simultaneously
• Different team members on different tests
• Compress total time

Time-boxing:

• Hard deadlines for each phase
• Better to have partial data fast than perfect data slow
• Iteration speed compounds learning

Example iteration sequence:

Week 1: Landing page test

• Created 3 variations
• Ran $100 in ads
• Learning: Headline A converts 3x better

Week 2: Offer structure test

• One-time vs. subscription pricing
• 50 clicks each
• Learning: Subscription preferred 2:1

Week 3: Audience test

• Small business vs. enterprise targeting
• $50 each segment
• Learning: Small business responds better

Week 4: MVP test

• Manual delivery to 5 customers
• Charged $99 each
• Learning: Feature C unnecessary, Feature D critical

Week 5: Price test

• $99 vs. $149 offers
• Learning: $149 converts at 80% of $99 rate (better revenue)

Result after 5 weeks:

• Clear target: Small businesses
• Optimal offer: $149 subscription
• Key features identified
• Revenue validated
• Ready for small launch

Iteration acceleration tools:

• No-code platforms: Webflow, Carrd, Bubble
• Testing tools: Google Optimize, Optimizely
• Analytics: Amplitude, Mixpanel, Google Analytics
• Communication: Notion for documentation
• Payment: Stripe for quick checkout setup

Integration Strategy

To test ideas fast and cheap:

1. Start with Assumption Extraction to identify what needs testing
2. Use Smoke Tests for initial demand validation
3. Conduct Customer Interviews for qualitative depth
4. Build Concierge MVP to test value delivery
5. Run Rapid Iteration Cycles to compress learning timeline

Fast Testing Success Indicators

You're testing effectively when:

• Validating or invalidating assumptions in days/weeks, not months
• Spending hundreds, not thousands on initial tests
• Getting real behavioral data, not just opinions
• Making clear go/no-go decisions based on evidence
• Learning from failed tests without significant loss
• Multiple iterations completed while others build first version

The Testing Paradox

Investing more time in cheap, fast testing often leads to faster overall success by avoiding expensive wrong directions.

Common Testing Mistakes

• Testing too many variables at once
• Confusing opinions with behavior
• Building before validating core assumptions
• Insufficient sample size for statistical meaning
• Ignoring negative signals (confirmation bias)
• Testing without clear success criteria
• Perfecting tests instead of running them

When to Stop Testing and Build

Test until:

• Core assumptions validated with behavioral data
• Clear demand demonstrated (pre-sales, waitlist, engagement)
• Solution approach proven (concierge/manual delivery worked)
• You understand target customer deeply
• Risks are acceptable given validation

Then build minimum viable version and continue testing with real product.