Strategy fundamentals

Strategy Fundamentals

The diagnostic layer that sits above every marketing channel. It answers what to work on next, which fit is actually the bottleneck, which channel to concentrate on, and whether your growth compounds or just spikes. Get this right and execution has leverage; get it wrong and no amount of ad spend or content fixes it.

When to use this

Reach for this skill when the question is about direction, not execution:

• Growth has stalled or plateaued; CAC is climbing, retention is slipping, and no one agrees why.
• Someone wants to scale paid or content but the funnel underneath may be leaking.
• "Which channel should we use?" / "We're on ten channels and none is working."
• "Do we have product-market fit?" / "How do we find PMF?"
• Positioning, differentiation, or value-prop work: "why do we win," "how are we different."
• "What's our unfair advantage / moat?" / fundraise or recruiting readiness checks.
• Finding a new idea, wedge, segment, or the next growth curve as the current one saturates.
• "Should I start this / quit my job for this?" and how to validate an idea before going all-in.
• Any plan that leaps to tactics with no written diagnosis behind it.

If the request is "run the ads," "write the article," "fix the checkout," that is execution, route below.

When NOT to use this (reach for instead)

• To execute the channel you picked here, use the execution disciplines: paid-acquisition (run and scale ads), seo-content / organic content (build the content engine), organic-social, email-marketing, product-led-growth, launch-gtm.
• To measure whether a change worked, use analytics-data (funnel metrics, experiment readout, North Star instrumentation). Strategy points at the leak; analytics confirms it.
• To write the words inside a landing page, ad, or positioning statement once the strategy is set, use copywriting-messaging and brand.
• To design the price/tiers/offer, use pricing-monetization (this skill decides Market-Model fit; that skill builds the actual pricing page).
• To pick the buyer's psychological triggers, use marketing-psychology.

Strategy decides the bet. These decide how to place it and how to read the result.

How this works (decision path)

Start from the symptom and route to the right play. The recurring principle: fix the upstream blocker before pouring more into the top of the funnel.

1. Growth is stalling and you don't know why → run the Five Fits diagnosis first. PMF is only one of five fits (Market, Product, Model, Brand, Channel); growth is capped by the lowest-scoring one, and founders usually blame the channel when the real issue is model, brand, or market. This is the master diagnostic; almost every other play flows from what it surfaces.
2. Fit looks fine but the system leaks → walk the growth funnel / AARRR in order (competitive audit → landing → onboarding → paid → content). Do not scale acquisition until onboarding stops the bucket leaking. The leakiest AARRR stage is where experiments go.
3. The market or persona is fuzzy → run market & persona definition (hyper-specific JTBD persona, starving-crowd test) and opportunity finding (whitespace, pain-point journal, problem-solution mapping) before anything downstream.
4. The channel question → run channel selection (Bullseye + the specificity×intent 2×2), cheap-test finalists, then concentrate on 1-2. Under-resourcing a channel produces false negatives.
5. You win but can't say why → run differentiation & positioning (Opposites Game, subtraction) and the unfair-advantages audit to find what to over-invest in.
6. Growth spikes then dies → design loops & flywheels so output feeds input, instead of one-off hacks.
7. The core curve is flattening → plan expansion / the next S-curve before saturation, not after.
8. You need to align the team or decide to start → write the strategy doc (Rumelt's kernel); if pre-launch, run distribution-first validation (validate with cash, not opinions).
9. Deciding who executes / how to apply AI → growth execution paths and the AI-native operating model.

Cross-cutting tool: the competitive-analysis framework feeds steps 4-6 (teardown competitors' ads, funnels, and A/B tests for tested ideas).

The plays

Diagnose the whole system

• Score the Five Fits and find the real bottleneck (Market/Product/Model/Brand/Channel + 10 pairings; run the CAC-vs-ACV math)
• Work the funnel in order, fix the leak before scaling (competitive → landing → onboarding → paid → content)
• Run a scientific AARRR growth program (PMF check → North Star → experiment loop across the five stages)

Market, persona & opportunity

• Define a hyper-specific JTBD persona and test for a starving crowd (5 JTBD questions, 4 starving-crowd factors, non-user TAM reframe)
• Scan systematically for opportunities and whitespace (pain-point journal, orphaned markets, problem-solution mapping)

Channel choice

• Pick channels with Bullseye + the specificity×intent 2×2, then concentrate on 1-2 (cheap-test finalists, avoid the false-negative trap)

Positioning & advantage

• Prove differentiation is real (Opposites Game, PMF by subtraction, demonstrate don't state)
• Audit personal + company unfair advantages and over-invest (4 company classes, Ladder of Proof rungs)

Compounding & durability

• Build growth loops and classify your flywheel (referral/paid-reinvest/free-wedge; don't fake network effects)
• Layer the next S-curve before the first saturates (dual innovation, adjacent niches, attribute-flip)

Direction & alignment

• Write a real strategy doc from Rumelt's kernel (diagnosis → guiding policy → coherent actions → one sentence)
• Sequence and validate a new business with cash, not opinions (audience-first vs service-first, exit metric, LOIs)
• Reverse-engineer competitors for tested tactics (ad libraries, hidden LPs, A/B variants, funnel teardown, tech stack)

Operating model

• Choose an execution path: DIY, hire, or outsource (resource-escalation ladder)
• Run an AI-native / solo operating model and gate AI to defensible levers (avoid busywork parallelization)

Key numbers & benchmarks

• PMF is one of five fits, not the whole game: only ~34% of startup failures are genuinely product-market fit; the other ~66% are Market-Model, Market-Brand, Market-Channel, or Model-Channel misfits.
• CAC reality math: enterprise CAC runs 15-30% of ACV. A $100K friends-and-family round at $15-30K CAC only buys 3-7 customers. 1,000 customers x $100K ACV = $100M (venture scale).
• Channel focus: 20% effort on a channel is not 20% of results, it is often 10-20x worse than 100% effort because you never climb the learning curve. Concentrate on 1-2.
• Aha-moment benchmarks: Facebook 7 friends in 10 days; Twitter follow 5+ accounts; Mailchimp first campaign sent; Grubhub first order; Candy Crush first game finished.
• Viral coefficient > 2 signals runaway-growth potential; niche/expensive products rarely go viral, don't force it. Dropbox hit 39x growth in 15 months via an embedded viral loop.
• Durability: 52% of the 2000 Fortune 500 were gone by 2020; the survivors ran dual (sustaining + disruptive) innovation and layered new S-curves during hypergrowth.
• Validation bar for a new venture: cash transactions, signed LOIs, or design partnerships, never interview enthusiasm. Favor fast time-to-value, high-margin, short sales cycles; SMB/mid-market (not 12-18mo enterprise cycles) first.
• AI-native shift: 36.3% of new startups are solo-founded (up from 23.7% in 2019); 52.3% of successful exits are solo-founded; AI-native startups run ~34% leaner. Cost savings everyone gets are not a moat.
• This sequencing has produced $1B+ in cumulative ad-client revenue: the evidence that paid scales only once the funnel underneath is fixed.

---

Reference library

Every play above, in full.

Choose channels with the Bullseye method and a 2×2 of audience specificity × search intent, then concentrate on 1-2 to escape the false-negative trap

The strategy

Don't spread thin across every channel or copy whatever worked for someone else. Brainstorm all channels, cheap-test the plausible ones, then pour effort into the one or two that win, because under-resourced channels look like failures even when they'd work.

When to use it

Choosing acquisition channels for a new product, or when current channels feel like they're plateauing and you're tempted to add more.

How to execute (steps)

1. Brainstorm across all 19 traction channels (from Traction, Weinberg/Mares): score each on audience fit, cost, and feasibility.
2. Narrow with a 2×2 (issue #188): audience specificity × active search intent. Broad + searching → Google/Amazon. Specific + searching → trade shows, niche review-site listings. Broad + not-searching → mass CPG-style awareness. Specific + not-searching → cold email, LinkedIn targeting, custom audiences, communities (Circle/Skool/Slack/Meetup).
3. Cheap-test the finalists: small ad buys measuring CPA, cold PR outreach, two months of LinkedIn, a basic affiliate program.
4. Concentrate on 1-2 winners (issue #257): 20% effort isn't 20% of results, it's often 10-20x worse than 100% effort because you never climb the channel's learning curve. Under-resourcing produces false negatives.
5. Match the channel to your growth lever (issue #180): each quarter go hard on one of, increase revenue per customer, optimize acquisition efficiency (conversion + CAC), scale existing channels while building a secondary engine before saturation, or enter new markets via small RICE-scored experiments.

Notes / caveats / examples

• The core discipline: brainstorm → rank → cheap-test → focus. The failure mode is testing everything at 20% and killing channels that would have worked at 100%.
• Build a secondary channel engine before your primary one saturates so you're not caught flat when CAC rises.

→ Skill conversion note

Strong skill candidate: a "channel picker" that scores candidate channels on the 2×2 (audience specificity × intent), suggests a cheap validation test per finalist, and enforces focus on 1-2 winners.

Reverse-engineer competitors' growth tactics, ads, content, and funnels

The strategy

Mine high-growth competitors for proven, testable growth ideas: source ideas → prioritize → A/B test → measure vs baseline (and implementation cost) → implement only winners. Study companies with similar audience and business model; avoid early-stage ones whose success may be non-replicable.

When to use it

Planning a growth roadmap, entering a market, or hunting for tested tactics to A/B test.

How to execute (steps)

• Reverse-engineer A/B tests: use incognito to reset variants, reload pages until visuals change, watch for URL params like "?abc=123", screenshot all variants with the Full Page Screen Capture Chrome extension. Find hidden landing pages by appending "/sitemap.xml" (e.g. forhims.com/sitemap.xml), or use Ahrefs/SEMrush to see pages receiving Google ad traffic.
• Audit ads: Facebook Ads Library (facebook.com/ads/library/), search the company's Page name; find high-frequency ads by browsing their site then watching retargeting; study copy, hooks, and comment sections. For Google Ads, use Ahrefs for paid keywords, click costs, volume, and the ads themselves (long-term keyword spend = effective).
• Read traffic signals: Google Trends (search volume for competitor name as a growth proxy), SimilarWeb (traffic), Sensor Tower (app rankings, keywords, install estimates).
• Audit content marketing: Ahrefs for highest-volume organic keywords + ranking posts, cross-referenced with BuzzSumo social shares; target the intersection of high search volume + high social engagement.
• Reverse-engineer the funnel: sign up for the newsletter, start the free trial, buy, click every drip-email tutorial link, and document every touchpoint; use Built With (builtwith.com) to reveal their tech/onboarding stack.

Notes / caveats / examples

• Purpose: establish baseline growth patterns, uncover overlooked tactics, and learn industry norms so you can deliberately break them to stand out.

→ Skill conversion note

A "competitor teardown" skill: given a competitor URL, produce a structured teardown (A/B variants, ad library findings, content gaps, funnel touchpoints, tech stack) with prioritized test ideas.

Prove your differentiation is real, pass the Opposites Game, build PMF by subtraction, and demonstrate value instead of stating it

The strategy

"Best-in-class" positioning is usually fake differentiation. Test whether your strategy is a real choice, differentiate by removing what customers don't want, and for undifferentiated products, demonstrate the value physically rather than claiming it.

When to use it

Positioning a product, writing a value prop, or deciding what to build/cut to stand apart.

How to execute (steps)

1. Run the Opposites Game (issue #150): strip subjective language ("best-tasting yogurt" → "ice-cream flavors in yogurt"). If the opposite strategy could also make sense (low-cost vs premium), it's real strategy. If the opposite is absurd (win vs lose), it isn't a choice, it's a platitude.
2. Build PMF by subtraction (issue #273): map competitor features and find what customers actively DON'T want, then remove it. Slate removed speakers, power windows, and paint to hit affordability.
3. Demonstrate, don't state (issue #214): show value physically, Tyrolit knife cutting real materials; use environment (Economist lightbulb installation); dramatize the objection (Surreal cardboard-taste ad); transform the mundane (LEGO bus-stop); create cognitive friction (Specsavers blurred ad). Ask: how can I demonstrate not state, where can I add delight, what won't competitors try?
4. Go deep on the problem (issue #313): when you understand the problem deeply, small solutions produce outsized results, Grammarly's two-word CTA lifted signups 8x; Duolingo modeled retention as 5x more impactful than other levers, then added Zynga-style leaderboards.

Notes / caveats / examples

• The Opposites Game is the fastest filter for whether a positioning statement is a genuine strategic choice.
• Subtraction is a differentiation lever most teams ignore because instinct is to add features.

→ Skill conversion note

Strong skill candidate: a "differentiation tester" that runs a value prop through the Opposites Game, suggests subtraction candidates from a competitor feature map, and brainstorms physical demonstrations of the core benefit.

Sequence a business the low-risk way, build the audience or a productized service first, and validate with cash (not opinions) before quitting your job

The strategy

Distribution is the scarce asset, and interview enthusiasm is not validation. De-risk a startup by building demand before the product (audience-first) or funding it with a capped service (service-first), and prove demand with real money before you go all-in.

When to use it

Deciding whether/when to start or quit for a startup; sequencing product vs distribution; validating an idea.

How to execute (steps)

1. Audience-first (distribution before product, issue #248): build content to 1K, 10K engaged followers, then launch a micro-SaaS with better UX. Linus Tech Tips did $5M in screwdriver pre-sales off 16M subs; Kane Kallaway → sandcastles.ai.
2. Service-first bridge ("manual before mechanical", issue #248): fund the SaaS with a capped productized service and define an exit metric (e.g. $300K ARR triggers the product transition) to avoid "service purgatory." DesignJoy ~$1.2M/yr.
3. Validate with cash before quitting (issue #284): run a wizard-of-oz / analog MVP while still employed; require cash transactions, signed LOIs, or design partnerships, not interview sentiment. The founder's opportunity cost was $1.65M, $2.4M over 7.5 years.
4. Pick a favorable business model (issue #284): fast time-to-value, high-margin, short sales cycle; SMB/mid-market before enterprise (12-18mo cycles). Confirm 2+ scalable channels to your ICP exist before building, and set an explicit 12-18-month clock with investors.

Notes / caveats / examples

• The exit metric on a service-first bridge is what prevents you from getting stuck running an agency forever.
• "Cash, LOIs, or design partnerships" is the validation bar, enthusiasm in interviews is the thing that fools founders.

→ Skill conversion note

Strong skill candidate: a "startup sequencing advisor" that recommends audience-first vs service-first based on the founder's existing distribution, sets an exit metric, and defines cash-based validation milestones before a quit decision.

Layer a second S-curve before the first saturates, run dual (sustaining + disruptive) innovation, pivot into adjacent niches, and flip your core-user attributes to find new segments

The strategy

Every growth curve flattens. The companies that endure launch the next curve, a new product, market, or channel, while the current one is still climbing, and they invest in disruptive bets that could cannibalize themselves before a competitor does.

When to use it

Planning the next phase of growth, or when your core product/market shows early signs of saturation (rising CAC, falling retention).

How to execute (steps)

1. Layer a new S-curve before saturation (issue #242): across product (Uber X→Pool→Eats→Freight), market (Slack devs→startups→enterprise), or channel (Dropbox ads→referral→B2B). Ask "what's the biggest pain our current customers still have?", leverage existing distribution, MVP-test without disrupting the core, and time the launch during hypergrowth (before CAC rises / retention drops).
2. Run dual innovation (issue #166): sustaining (incrementally improve for current customers) AND disruptive (invest in cheaper/radical alternatives that could cannibalize you). Kodak, Google, and Boeing failed by only optimizing existing products; 52% of the 2000 Fortune 500 were gone by 2020.
3. Decouple exec comp from short-term metrics (issue #166): tie incentives to 3-5 year innovation goals, not annual bonuses, so disruptive bets survive.
4. Pivot into adjacent niches using core strengths (issue #276): Garmin reused GPS/mapping/battery tech for Forerunner (running), Approach S1 (golf, 14K+ courses, 2011), and fēnix (mountaineering) as smartphones killed car GPS.
5. Flip your core-user attributes (issue #224): document your highest-engagement user's traits (Instacart: "women, urban, affluent, willing to spend an hour ordering"), invert each (men, suburban, price-sensitive) to surface adjacent segments, then remove the small friction blocking them.

Notes / caveats / examples

• Timing is the hard part: launch the second curve during hypergrowth, not after the first curve has already stalled.
• Attribute-flipping is a cheap, concrete way to generate adjacent-segment hypotheses.

→ Skill conversion note

Strong skill candidate: an "S-curve planner" that identifies expansion vectors (product/market/channel), runs the attribute-flip exercise for adjacent segments, and checks whether a disruptive bet is being funded alongside sustaining work.

Diagnose growth with the Five Fits (Foundational Five), PMF is just one of ten pairings, and the lowest-scoring fit is your real bottleneck

The strategy

Founders obsess over product-market fit and blame channels when growth stalls. But PMF is only one of the alignments that matter. Audit five elements, Market, Product, Model (pricing), Brand, Channel, and their pairings; growth is bottlenecked by the lowest-scoring fit, not by whatever showed up last in the metrics.

When to use it

When growth stalls, before killing a channel, before scaling with paid or AI, or as a periodic strategic health check.

How to execute (steps)

1. Score the five core fits 1-10: Market-Product, Market-Model, Market-Brand, Market-Channel, Model-Channel. Growth is capped by the lowest.
2. Check all ten pairings before blaming a channel: product-channel (does the product suit how people arrive there?), model-channel (can unit economics survive that channel's CAC?), brand-channel (is there enough narrative space?), market-channel (is your market reachable there?).
3. Run the CAC reality math (Market-Model): enterprise CAC runs 15-30% of ACV; a $100K F&F round at $15-30K CAC only buys 3-7 customers. 1,000 customers × $100K ACV = $100M (venture scale), check the math fits the model.
4. Watch for false negatives: early traction hides weak foundations because early users self-select; growth mode exposes all five at once, so founders wrongly blame channels.
5. Audit structure before adding hires/tools/experiments (issue #304): "by the time a problem shows up in metrics, the real issue was baked in earlier." Review each fit with a co-founder or advisor.

Notes / caveats / examples

• Only ~34% of startup failures are actually product-market fit; the other ~66% are the other four fits. Foundational Five reframes PMF as one pairing (Market↔Product) among ten.
• Channel-model mismatch examples: senior healthcare on Snapchat, luxury on Groupon. Failure examples: a premium watch under an untrusted brand, a $10k/mo toy subscription, LinkedIn ads for fidget spinners.

→ Skill conversion note

Strong skill candidate: a "Five Fits diagnostic" that walks a founder through scoring all five fits and ten pairings, runs the CAC-vs-ACV math, flags the bottleneck fit, and warns about false negatives from early self-selected users.

Turn growth knowledge into action by choosing an execution path, DIY, hire full-time, or outsource, and climbing the resource-escalation ladder

The strategy

Knowledge is useless unless applied. Once you understand the growth playbook, the next decision is how you'll execute, one of three paths, chosen by stage and budget, and how you'll keep leveling up via an escalation ladder of resources.

When to use it

When you've learned the fundamentals and need to decide who does the growth work and how to keep improving.

How to execute (steps)

1. Pick an execution path (decision matrix):
   • DIY growth, best for early-stage startups; lowest cost; requires continuous self-education to progress.
   • Hire full-time, a growth/marketing co-founder (high risk, relationship-dependent), a single junior or senior hire (expensive, knowledge gaps likely), or multiple specialized roles (resource-intensive).
   • Outsource, generalist agencies (full scope), specialist agencies (focused expertise), or a mix-and-match of freelancers/agencies.
2. Climb the resource-escalation ladder: self-education via guides/playbooks → growth newsletter (~100k+ founders) → structured growth program (deep training, cited for 1,000+ startups) → paid services / professional growth teams.
3. Keep learning from the specific playbook library: TikTok Ads, Content-Led SEO, Marketing Psychology, Influencer Marketing, Landing Page Teardowns, LinkedIn Organic.

Notes / caveats / examples

• Primary directive verbatim in spirit: take action on what you've learned first, application beats more reading.
• Subscriber benchmark cited: newsletter targets "110,000+ operators."

→ Skill conversion note

Lower skill-value (mostly a decision aid); could become a short "growth execution path chooser" that recommends DIY vs hire vs outsource based on stage and budget inputs.

Treat growth as the intersection of marketing, sales, and product, and fix the funnel in order: don't scale paid until onboarding stops the bucket leaking

The strategy

Growth is the intersection of marketing (attract the right audience), sales (convert), and product (retain), cross-disciplinary and systematic, not siloed. The operating principle: remove the blocker to the next stage before pouring more into the top, and sequence the work so you don't scale a leaky funnel.

When to use it

As the orienting model when building or auditing a company's whole growth system, and to decide what to work on next.

How to execute (steps)

Work the funnel in this order:

1. Competitive analysis, de-risk strategy by auditing competitors' growth assets first.
2. Landing pages, convert visitors; identify your unique selling features and value props.
3. Ad creation, write/design social ads (Facebook, Instagram, TikTok).
4. Onboarding, the retention system. "Marketing fills the bucket, onboarding prevents leaking." Fix this before scaling spend.
5. Paid advertising, the scalable channel, but only "once the bucket isn't leaking, fill it."
6. Content marketing, organic lead generation and authority building.
7. A/B testing, systematic conversion-rate improvement on any of the above.

Notes / caveats / examples

• Organizational requirement: growth must be a "core focus and principle" led by founders, it can't succeed in silos.
• Credibility marker: $1B+ in cumulative revenue has been generated across dozens of ad clients using this sequencing (evidence paid scales when the funnel underneath is fixed).

→ Skill conversion note

Becomes a "growth funnel diagnostic" SKILL.md that asks about each stage (landing page, onboarding, paid, content) and tells the user which blocker to fix next before scaling acquisition.

Growth hacking as a process, PMF prerequisites, the AARRR funnel, and scientific experimentation

The strategy

Growth hacking isn't a bag of tricks, it's a disciplined process: confirm product-market fit, then run scientific experiments across the five AARRR funnel stages (Acquisition, Activation, Retention, Referral, Revenue), extracting learnings each cycle. Replace "tactic hunting" with hypotheses → tests → learnings → new tests.

When to use it

• Standing up a repeatable growth program instead of one-off campaigns.
• Diagnosing which funnel stage is leaking and where to focus experiments.

How to execute (steps)

1. Confirm PMF first, "are you building something people want?" Define your Core Product Value (Dropbox = easy centralized sharing; Duolingo = fun free language learning), set a North Star Metric (Patreon = creators earning above a threshold; Netflix = median monthly view hours; Miro = collaborative boards), and track counter metrics (e.g. downloads without usage).
2. Acquisition, test many channels (SEO, PPC, social ads, content, influencer, email, PR, referrals, sponsorship) judged on scale + targetability. Port learnings across channels (a high-converting Google headline → email subject line).
3. Activation, drive users to the "aha" moment fast. Cut signup friction: shorten forms ~50%, delay registration until value is shown, give topic-relevant recommendations immediately.
4. Retention, cheaper than acquisition. Build natural retention via product quality (not promos), personalize support, surprise customers with freebies, and fight involuntary churn (retry expired cards via mobile links + SMS).
5. Referral, turn engaged users into a sales force. Organic (Slack invites) vs. inorganic/incentivized (Outdoor Voices $20/friend; Dropbox extra storage per referral). Needs a product whose value grows with more users.
6. Revenue, optimize the three pricing levers: value metric (what you charge for), pricing model (tiered/flat/per-unit), and price point. Test monthly vs. annual, run willingness-to-pay research, use freemium/trials, upsell/cross-sell.
7. Run the experiment loop: hypothesis → prioritize → controlled test → collect data → validate/refute → report → next experiment. Not "A/B testing headers", systematic method applied to every funnel stage.

Notes / caveats / examples

• Aha-moment benchmarks: Facebook = 7 friends within 10 days; Mailchimp = first campaign sent; Grubhub = first order; Candy Crush = first game finished.
• Twitter onboarding: users who didn't follow accounts post-signup churned → forced following 5+ accounts (later reduced to 1) + topic recommendations.
• Dropbox: 39x user growth in 15 months via an embedded viral loop + incentivized referrals.
• High-velocity incrementalism: Booking.com, Spotify, Meta, Duolingo win via many small changes at scale, requires a large experimentation program.
• Viral coefficient > 2 signals runaway-growth potential. Niche/expensive products are unlikely to go viral, don't force it.

→ Skill conversion note

Becomes a "growth-audit" prompt: input product + funnel metrics → output PMF check (CPV + North Star + counter metric), the leakiest AARRR stage, and a prioritized experiment backlog with hypotheses.

Build compounding growth loops and flywheels instead of one-off hacks, and know which of the four flywheel types (if any) you actually have

The strategy

One-off tactics (Product Hunt launches, press, limited promos) give a spike and stop. Growth loops reinvest their output as input, so they compound. Design for loops, then classify the flywheel you have, because "data collection" and "multiple user types" are not network effects.

When to use it

Planning a durable growth engine, or auditing whether your growth compounds or needs constant fresh effort.

How to execute (steps)

1. Prefer compounding loops over hacks: referral programs and profitable ads reinvest outputs as inputs; PH launches, limited promos, and press don't compound.
2. Match loop type to stage, Racecar Growth (issue #162): non-scalable Kickstarts + Turbo boosts early, then self-perpetuating Growth engines. Lubricate the engine → add accelerants → launch parallel engines → expand products/segments.
3. Classify your flywheel (issue #307): network effects (each user raises value for others), scale (volume lowers unit cost), embedded (B2B switching costs, aids retention not acquisition), brand (taboo-breaking à la Hims/Airbnb, single-idea ownership à la Amazon, community-as-product).
4. If you lack a flywheel, decide deliberately: accept linear growth, use temporary boosters (pricing/channels/partnerships), or engineer a loop on purpose. Don't mistake data collection or multiple user types for network effects.
5. Use "give away free what rivals charge for" as a loop input (issue #182): the "Law of Conservation of Attractive Profits" / "your margin is my opportunity", build affinity with a free wedge, funnel into the paid product. Calm: $7.7M/month, 150M+ downloads.

Notes / caveats / examples

• The test for a real loop: does this month's output feed next month's input? If not, it's a booster, not an engine.
• Brand flywheels can be built deliberately (own one idea, break a taboo, make the community the product).

→ Skill conversion note

Strong skill candidate: a "growth loop designer" that classifies a business's current flywheel type, flags fake network effects, and proposes a compounding loop (referral, paid-reinvest, or free-wedge) matched to stage.

Define a hyper-specific core persona around jobs-to-be-done, test the market for "starving crowd" demand, and expand TAM by selling to non-users

The strategy

Vague personas cause shiny-object syndrome. Define one hyper-specific core persona built on jobs-to-be-done (not demographics), verify the market is a "starving crowd" with real desperation and money, and grow by reframing TAM toward non-users rather than fighting over existing users.

When to use it

Early positioning, GTM targeting, or when the team keeps chasing every possible customer segment.

How to execute (steps)

1. Name a hyper-specific core persona: Substack, "successful veteran online newsletter writers"; Cameo, "B-list athletes in Chicago." Specificity kills shiny-object drift.
2. Build the persona on jobs-to-be-done, not demographics (issue #112): five questions, what stresses them, where they search, what solutions they're testing, how they define success, what makes them nervous.
3. Test the market as a "starving crowd" (issue #145): four factors, desperate pain, purchasing power, easy targeting (findable congregations), growth trajectory. Find desperate demand first; optimize offer and persuasion second.
4. Reframe TAM toward non-users (issue #194): "sell oat milk to people who don't drink it" expands the market from milk-avoiders to all cow-milk drinkers.
5. Mine real stories via digital ethnography (issue #088): join customers' communities, watch post-intensity, and interview with "tell me about a time our product added value", not "what's your favorite feature."

Notes / caveats / examples

• Order matters: find the starving crowd before perfecting the offer. A great offer to an indifferent market fails.
• JTBD interviewing surfaces the emotional triggers demographics miss.

→ Skill conversion note

Strong skill candidate: a "core persona builder" that runs the 5 JTBD questions, scores the market on the 4 starving-crowd factors, and suggests a non-user TAM reframe.

Systematically find business and growth opportunities, whitespace research, dead-product gaps, a pain-point journal, and problem-solution mapping

The strategy

Opportunities aren't found by waiting for inspiration, they're surfaced by running repeatable scans of frustrations, abandoned markets, and unserved whitespace, then mapping each problem into a solution.

When to use it

Idea generation for a new product or feature line; finding a wedge or an underserved segment.

How to execute (steps)

1. Run opportunity-finding methods (issue #250): a pain-point journal (log daily frustrations, review monthly), a "painful payment" audit of expensive recurring charges ("could this be 50% cheaper?"), JTBD gaps, a business-model scan (subscription/rental/unbundle/DTC), a professional-to-consumer bridge, and a retail category scan for stale designs.
2. Find three kinds of whitespace (issue #136): audience whitespace (ignored segments), channel whitespace (presence + low competition + proven + brand fit), positioning whitespace (search queries + social listening). Tool: SimilarWeb.
3. Monitor abandoned markets (issue #105): watch killedbygoogle.com, discontinued products leave thousands/millions of orphaned users; move fast to capture them.
4. Turn problems into offers via Problem-Solution Mapping (Hormozi, issue #164): list every step to the customer's goal, name the problem at each step, reverse each into solution language, brainstorm solutions, then trim by value / likelihood / effort / time.

Notes / caveats / examples

• Auggie (issue #250): a silicone breast-implant sizer rental (7-14 days) replacing rice-in-pantyhose hit $20K/mo within 10 months, 1,000+ units, 100+ 5-star reviews, incumbents ignored a solution that cannibalized their model.
• Problem-Solution Mapping converts friction points directly into offer ideas.

→ Skill conversion note

Strong skill candidate: an "opportunity scanner" that runs the whitespace types, a killedbygoogle-style orphaned-market check, and a Problem-Solution Mapping pass over a stated customer goal.

Run an AI-native, often solo, operating model, position AI to defensible advantage, and gate AI use so you parallelize real work, not busywork

The strategy

AI is collapsing team sizes and shifting where advantage sits. Solo and lean AI-native startups now reach outcomes that used to require teams, but only if AI is pointed at defensible levers and real work, not content volume and busywork.

When to use it

Deciding team structure and where to apply AI as a founder or lean operator; setting an AI-era strategic position.

How to execute (steps)

1. Consider the solo/AI-native model with the data (issue #328): 36.3% of new startups are solo-founded (up from 23.7% in 2019); 52.3% of successful exits are solo-founded; AI-native startups run 34% leaner.
2. Position AI by role (issue #315): marketers should build data systems tying AI to pipeline/revenue (not content volume); agencies should move upstream (strategy) or downstream (AI-scaled price play), not stay mid-market; founders should separate temporary cost savings from real improvement and protect defensible levers.
3. Gate every AI task (issue #293): ask "is it worth having AI on this, or am I just making busywork?" before parallelizing, this prevents low-value task proliferation.

Notes / caveats / examples

• Solo/AI-native case studies: Base44 ($1M ARR in 3 weeks, $80M Wix exit in <500 days), HeadshotPro ($3.6M/yr, 40K users, no employees), Marc Lou ($1M+ in 2025, TrustMRR built in 24h), Pieter Levels ($3M+/yr, 40+ products).
• The strategic warning: AI cost savings that everyone gets aren't a moat, protect the defensible lever underneath.

→ Skill conversion note

Lower skill value (a strategic-posture guide); could become an "AI leverage auditor" that classifies tasks as defensible-lever vs busywork and flags where AI is being used for volume instead of advantage.

Write a real strategy doc, Rumelt's kernel (diagnosis → guiding policy → coherent actions) distilled into a 4-section argument and one-sentence statement

The strategy

Most "strategy" is a list of goals. Real strategy is an argument: a clear diagnosis of the problem, a guiding policy, and coherent actions. Capture it in a tight four-section doc anchored by a one-sentence statement anyone on the team can repeat.

When to use it

Setting or communicating company/product/growth strategy; aligning a team; before a planning cycle or fundraise.

How to execute (steps)

1. Start from Rumelt's kernel (issue #139): diagnosis (what's really going on) → guiding policy (your overall approach) → coherent actions (concrete, mutually-reinforcing moves). Apple 1997 cut to 4 products before iPod/iPhone/iPad.
2. Write the 4-section doc: Argument (the narrative), Statement (one sentence), Implications (cross-functional changes), Execution flow (3-5 major steps).
3. Build the Argument as a chain: status quo → why it's broken → your insight → your fix.
4. Compress to one sentence: Duolingo, "gamify language learning so it feels like a game"; Tesla, "make electric cars cool and desirable, then accessible."
5. Pressure-test it with a challenge network (issue #096): 3-5 advisors who'll disagree, on a regular feedback cadence (Steve Jobs' team pushing back on iPhone skepticism). If no one can argue the opposite, it's not a strategy, it's a platitude.

Notes / caveats / examples

• The one-sentence statement is the test of whether the strategy is actually coherent, vague companies can't produce one.
• Pairs with the differentiation-and-positioning card (the Opposites Game is another test of whether a statement is real strategy).

→ Skill conversion note

Strong skill candidate: a "strategy doc builder" that prompts for diagnosis/guiding-policy/actions, drafts the 4 sections, forces a one-sentence statement, and generates the opposite-strategy pressure test.

Audit your unfair advantages, personal (assets, knowledge, timing, connections) and company (product, ecosystem, marketing fuel, marketing engine), then over-invest in the ones you hold

The strategy

Every founder and company holds a handful of advantages competitors can't easily copy. Name them explicitly, check which "rungs of proof" you've actually earned, and pour resources into your real advantages instead of playing everyone else's game.

When to use it

GTM planning, positioning, fundraising/recruiting readiness, or deciding where to concentrate resources.

How to execute (steps)

1. Audit your 5 personal unfair advantages (issue #141): assets, knowledge, location/timing, interpersonal skills, prestige/connections (Ali Abdaal's Cambridge, Andreessen's timing). Exploit them in your GTM.
2. Audit the 4 company advantage classes (issue #236) and over-invest:
   • Product, referrals (Slack/WhatsApp), freemium (Figma), obvious differentiator (Tesla/Superhuman).
   • Ecosystem, app store (Shopify), integrations (Zapier), category tailwind.
   • Marketing fuel, brand narrative (Patagonia/SpaceX), educational content (Ahrefs), loyalty/gamification (Duolingo), proprietary data (Spotify).
   • Marketing engine, GTM wedge (Stripe devs→enterprise), organic search (HubSpot), community (Webflow/Notion).
3. Check the Ladder of Proof (NfX, issue #150): audit which rungs, team, product, traction, you've earned as a fundraising/recruiting readiness check, and prioritize the critical "red rungs" you still owe.

Notes / caveats / examples

• The move is asymmetric: identify the advantage you have that rivals lack, then over-invest there rather than matching them feature-for-feature.
• The Ladder of Proof turns "are we ready to raise/recruit?" into a concrete rung-by-rung audit.

→ Skill conversion note

Strong skill candidate: an "unfair advantage auditor" that inventories personal + company advantages, maps each to the four classes, and flags the red rungs on the Ladder of Proof.

---

From God of Skills: a curated, hand-tested directory of AI skills, prompts, templates and image style guides. Source: https://godofskills.com/skills/strategy-fundamentals?ref=claude-skill