Skip to main content
These foundational thinking models sharpen your strategy and help you solve the right problems. Use them to make better decisions, avoid common pitfalls, and focus on what matters most.
Thinking frameworks are meta-tools that help you approach problems more effectively. Master these before diving into specific tactics.

Strategic Thinking

Definition

Break problems down to their most basic truths and build solutions from there. Instead of copying what others do, ask “why” repeatedly until you reach fundamental realities.

How It Works

Use the 5 Whys technique to tunnel down to root causes:
  • Why do we need this feature?
  • Why do customers want that?
  • Why does that problem exist?
  • Why hasn’t it been solved?
  • Why would our solution work?

Marketing Application

Don’t assume you need content marketing because competitors do. Ask:
  • Why do we need marketing?
  • Why would customers pay attention?
  • Why is content the right format?
  • Why would this content be different?
You might discover you need a completely different approach.

Examples

Wrong Approach

“Our competitor has a blog, so we need one too.”

First Principles

“We need to reach decision-makers. They consume podcasts during commutes. Let’s start a podcast.”
When stuck, ask: “If I had to solve this problem from scratch with no existing solutions, what would I do?”

Definition

People don’t buy products—they “hire” them to get a specific job done. Focus on the outcome customers want, not product features.

How It Works

Customers have jobs they need to accomplish (functional, emotional, or social). Your product is hired when it’s the best available solution for that job.

Marketing Application

Frame your marketing around the job, not the tool:
  • A drill buyer doesn’t want a drill—they want a hole in the wall
  • A CRM buyer doesn’t want software—they want to never lose a customer
  • A course buyer doesn’t want videos—they want a career transformation

Examples

Slack’s positioning:
  • Feature-focused: “Cloud-based team chat with channels and integrations”
  • JTBD-focused: “Where work happens” (the job: coordinate team communication)
Intercom’s evolution:
  • Started: “Customer messaging platform”
  • Evolved: “Make internet business personal” (the job: personalize at scale)
Interview customers about the circumstances that led them to seek a solution, not just what features they like.

Definition

Know what you’re genuinely good at and stay within that circle. Venture outside only with proper learning or by hiring experts.

How It Works

Your circle of competence is where you:
  • Have deep knowledge from experience
  • Understand nuances others miss
  • Can predict outcomes better than average
  • Have a competitive advantage

Marketing Application

Don’t chase every channel because it’s trendy. Double down where you have genuine expertise:
  • Great writer? Focus on SEO content
  • Video natural? Prioritize YouTube
  • Technical product? Developer communities
  • Visual brand? Instagram/Pinterest

Examples

Inside your circle: “We’ve run Google Ads for 5 years and know our CPA. Let’s scale this.”Outside your circle: “TikTok is hot. We should try it!” (but no video experience, young audience you don’t understand)
Expand your circle deliberately through learning, not randomly through experimentation.

Definition

Instead of asking “How do I succeed?”, ask “What would guarantee failure?” Then systematically avoid those things.

How It Works

It’s often easier to identify what doesn’t work than what does. By avoiding catastrophic errors, success becomes more likely.

Marketing Application

Before launching a campaign, list everything that would make it fail:
  • ❌ Confusing value proposition
  • ❌ Wrong target audience
  • ❌ Slow landing page (3+ seconds)
  • ❌ Broken forms or checkout
  • ❌ No mobile optimization
  • ❌ Weak call-to-action
Then prevent each systematically.

Examples

Planning a webinar:Don’t ask: “How do I get 1,000 attendees?”Instead ask: “What would guarantee no one shows up?”
  • Bad timing (Friday 5pm)
  • Generic topic everyone covers
  • No promotion
  • Complicated registration
  • No reminder emails
Avoid these, and attendance improves.
Inversion is especially powerful for risk management and preventing disasters.

Decision-Making Tools

Definition

The simplest explanation is usually correct. Avoid overcomplicating strategies or attributing results to complex causes when simple ones suffice.

How It Works

When faced with competing explanations, favor the one requiring the fewest assumptions. Complexity is often a sign you’re overthinking.

Marketing Application

If conversions dropped:
  • ✅ Check the obvious first: broken form, page speed, mobile display
  • ❌ Don’t immediately assume: complex attribution issues, algorithm changes, market shifts
Most problems have simple causes. Start there.

Examples

Conversion rate dropped 30%:Complex theory: “Algorithm changes affected our quality score, reducing impression share in our target demos.”Simple reality: Last week’s deploy broke the checkout form on mobile. (Check this first!)
When data is weird, check for tracking issues before building elaborate theories.

Definition

Roughly 80% of results come from 20% of efforts. Identify and focus on the vital few; minimize or eliminate the trivial many.

How It Works

In most systems, inputs and outputs are not evenly distributed. A small portion of causes creates most effects.

Marketing Application

Find your vital 20%:
  • 20% of channels drive 80% of conversions → focus there
  • 20% of content gets 80% of traffic → create more like it
  • 20% of customers generate 80% of revenue → market to lookalikes
  • 20% of features drive 80% of value → emphasize these in messaging

Examples

Channel analysis:
  • Google Ads: 60% of budget, 15% of conversions
  • Content/SEO: 10% of budget, 40% of conversions
  • Referrals: 5% of budget, 30% of conversions
  • Social: 25% of budget, 15% of conversions
Action: Reallocate budget to SEO and referrals.
Use 80/20 analysis quarterly to identify what’s working and ruthlessly cut what isn’t.

Definition

A local optimum is the best solution in your immediate area. A global optimum is the best solution overall. Don’t get stuck optimizing the wrong thing.

How It Works

Imagine hiking in fog. You climb to the nearest peak (local optimum) but can’t see there’s a much taller mountain nearby (global optimum).

Marketing Application

Local optimization:
  • A/B testing email subject lines (+2% open rate)
  • Tweaking ad copy (+5% CTR)
  • Optimizing landing page buttons (+3% conversion)
Global optimization:
  • Is email the right channel?
  • Are we targeting the right audience?
  • Are we solving the right problem?
Zoom out before zooming in.

Examples

Stuck at local optima: Spending months optimizing Google Ads when:
  • Your ideal customers don’t click ads
  • They find solutions through communities
  • You’d get better ROI from content marketing
Finding global optima: Step back and ask: “If I could only do one thing, what would it be?”
Periodically audit whether you’re optimizing the right thing, not just optimizing well.

Definition

Every system has one bottleneck limiting overall throughput. Find and fix that constraint before optimizing anything else.

How It Works

Your marketing funnel is only as strong as its weakest link. Improving non-constraints doesn’t improve results.

Marketing Application

Identify your constraint:If traffic is low but conversion is high:
  • Constraint: Top of funnel
  • Fix: Invest in traffic generation
  • Don’t waste time: Optimizing conversion (already good)
If traffic is high but conversion is low:
  • Constraint: Conversion rate
  • Fix: Improve messaging, offer, page experience
  • Don’t waste time: Buying more traffic

Examples

SaaS funnel:
  • 10,000 visitors/month
  • 100 trial signups (1% conversion)
  • 50 paying customers (50% trial→paid)
Where’s the constraint? Trial signup rate (1% is low)Fix this before improving trial→paid conversion (already strong at 50%)
Measure your funnel to identify constraints. Don’t guess—calculate which step has the weakest conversion.

Definition

Every choice has a cost—what you give up by not choosing alternatives. The real cost isn’t just money spent, but value you could have created elsewhere.

How It Works

When evaluating options, don’t just ask “Is this worth it?” Ask “Is this the best use of resources compared to all alternatives?”

Marketing Application

Time spent on a low-ROI channel is time not spent on high-ROI activities:
  • Building a TikTok presence (10 hours/week, unclear ROI)
  • Opportunity cost: Could have written 2 SEO articles (proven ROI)
Always compare against your next best alternative.

Examples

Decision: Sponsor a conference ($10,000)Don’t just ask: “Will we get $10,000 in value?”Ask: “What else could we do with $10,000?”
  • Run Google Ads (known CPA: $200, 50 customers)
  • Create content (10 articles, 2 years of traffic)
  • Hire a contractor (40 hours of specialized work)
Which creates the most value?
Before saying yes to anything, explicitly list what you’re saying no to.

Definition

After a certain point, each additional unit of investment yields progressively smaller gains. The first effort matters most.

How It Works

Early optimizations produce big gains. Later ones produce smaller improvements at higher cost.

Marketing Application

Content marketing example:
  • 1st blog post: Establishes presence, ranks for brand terms
  • 10th blog post: Filling out topic coverage, good ROI
  • 100th blog post: Incremental traffic, topic overlap
  • 1000th blog post: Minimal unique value
Know when to diversify rather than double down.

Examples

Landing page optimization:
  • Round 1: Fix broken mobile layout → +40% conversion
  • Round 2: Improve headline and CTA → +15% conversion
  • Round 3: Test button colors → +2% conversion
  • Round 4: Adjust spacing → +0.5% conversion
After round 3, time is better spent elsewhere.
Track cost-per-improvement. When gains get too expensive, move to a different optimization.

Definition

Consider not just immediate effects, but the effects of those effects. Think through consequences two or three steps ahead.

How It Works

First-order thinking: “This will increase revenue”Second-order thinking: “This will increase revenue, which will attract competitors, which will require us to…”

Marketing Application

Flash sale decision:
  • First-order: Boosts short-term revenue ✅
  • Second-order: Trains customers to wait for discounts ⚠️
  • Third-order: Degrades brand perception and margins 🚫
Aggressive affiliate program:
  • First-order: Drives lots of signups ✅
  • Second-order: Attracts affiliates who spam and use black-hat SEO ⚠️
  • Third-order: Google penalizes your domain, affiliates vanish 🚫

Examples

Freemium model:
  • 1st order: More users
  • 2nd order: Support costs increase
  • 3rd order: Free users attract more free users (referrals)
  • 4th order: Need to optimize free→paid conversion
Plan for each stage.
Before major decisions, map out 2-3 levels of consequences. What happens after the obvious result?

Definition

Think in probabilities and ranges, not certainties. Estimate likelihoods and plan for multiple outcomes.

How It Works

Instead of “This campaign will work” or “This won’t work,” think:
  • 60% chance of 3-5% conversion rate
  • 30% chance of 1-3% conversion rate
  • 10% chance of >5% conversion rate

Marketing Application

Don’t bet everything on one campaign:
  • Best case: Exceeds targets
  • Expected case: Meets 70% of targets
  • Worst case: Breaks even or slight loss
Spread risk across multiple initiatives with different probability profiles.

Examples

Budget allocation:Instead of: “Put $100K into this new channel”Think: “This channel has:
  • 20% chance of 10x ROI
  • 50% chance of 2x ROI
  • 30% chance of breaking even
Expected value: (0.2×10 + 0.5×2 + 0.3×1) = 3.3x”Bet size should match confidence level.
Keep a decision journal. Record your probability estimates and actual outcomes to calibrate your thinking.

Risk Management

Definition

Models and data represent reality but aren’t reality itself. Don’t confuse your analytics dashboard with the actual customer experience.

How It Works

Every model simplifies reality. The simplification is useful but incomplete and sometimes misleading.

Marketing Application

  • Your customer persona is a model—real customers are more complex
  • Analytics show correlation, not always causation
  • Quantitative data misses qualitative context
  • User behavior in tests may differ from real-world usage
Stay connected to reality:
  • Talk to actual customers regularly
  • Use your own product
  • Read support tickets
  • Watch session recordings

Examples

The model says: “Users spend 45 seconds on our pricing page”Reality: They’re confused and comparing to competitors in other tabs, not engaged and consideringThe model says: “Our ICP is ‘Marketing Director, 50-200 employees, B2B SaaS’”Reality: Your best customers include a freelancer, an agency owner, and a Fortune 500 team—none fit the persona perfectly
When data conflicts with customer feedback, trust the feedback. Data tells you “what,” customers tell you “why.”

Definition

Combine extreme safety with small high-risk/high-reward bets. Avoid the mediocre middle.

How It Works

Instead of moderate risk across everything:
  • 80-90%: Safe, proven, reliable approaches
  • 10-20%: High-risk experiments with asymmetric upside
  • 0%: Mediocre-risk, mediocre-reward options

Marketing Application

Budget allocation:
  • 80% in proven channels (known ROI, reliable)
  • 20% in experimental bets (new channels, wild ideas)
  • 0% in “maybe this works” middle ground
If experiments succeed, shift them to the proven bucket. If they fail, the loss is contained.

Examples

Content strategy:Safe side (80%):
  • SEO articles targeting proven keywords
  • Customer case studies
  • Product tutorials
Experimental side (20%):
  • Launch a podcast
  • Try interactive tools
  • Controversial thought leadership
Avoid the middle:
  • Generic blog posts (not risky, not safe, just mediocre)
The barbell protects downside while capturing upside. You can’t lose more than 20%, but experiments could 10x that bucket.

Behavioral Psychology

Understand how customers think and make decisions

Growth Models

Models for building and scaling growth systems

Build docs developers (and LLMs) love