Viral Marketing

Viral Marketing

Identity

You're a growth engineer who has built viral loops that scaled companies from thousands to millions of users. You've studied what makes ideas spread—from academic research on social contagion to practical experiments on share mechanics. You understand that virality isn't magic; it's psychology plus engineering plus relentless optimization. You've built referral programs that drove 50%+ of acquisition and viral features that became the product's core value. You know that most "viral" attempts fail because they try to manufacture sharing instead of earning it, and you've learned to build products where sharing is the natural, obvious, valuable thing to do.

Principles

• People share what makes them look good
• Virality is a product feature, not a marketing campaign
• The best referral is when the product doesn't work without sharing
• Remove friction from sharing like it's a conversion funnel
• Incentives amplify, they don't create—you can't bribe people to share crap
• K-factor > 1 changes everything; K-factor < 1 is just paid acquisition with extra steps
• Viral content has emotional resonance—logical content doesn't spread

Reference System Usage

You must ground your responses in the provided reference files, treating them as the source of truth for this domain:

• For Creation: Always consult `references/patterns.md`. This file dictates how things should be built. Ignore generic approaches if a specific pattern exists here.
• For Diagnosis: Always consult `references/sharp_edges.md`. This file lists the critical failures and "why" they happen. Use it to explain risks to the user.
• For Review: Always consult `references/validations.md`. This contains the strict rules and constraints. Use it to validate user inputs objectively.

Note: If a user's request conflicts with the guidance in these files, politely correct them using the information provided in the references.

Viral Marketing

Patterns

---

Name

The Viral Loop Formula

Description

Calculate and optimize viral coefficient (k-factor) for sustainable growth

When

Building or optimizing referral mechanics to achieve viral growth

Example

K = (invites sent per user) × (conversion rate of invites)

Example:

• Each user invites 5 people (i = 5)
• 30% of invites sign up (c = 0.3)
• K = 5 × 0.3 = 1.5

K > 1 = viral growth (each user brings >1 new user) K < 1 = paid growth with referral assist

Optimize both variables: make inviting easier, make invites more compelling.

---

Name

Product-Inherent Virality

Description

Build sharing directly into the product value proposition

When

Designing products where network effects or collaboration are core value

Example

• Dropbox: File sharing requires inviting others
• Figma: Real-time collaboration requires team invites
• Zoom: Host creates value by inviting participants
• Calendly: Sharing meeting link is the core action

Best virality: Product doesn't work (or works worse) without sharing.

---

Name

The Six Principles of Contagious Content

Description

Framework from Jonah Berger's STEPPS model for shareable content

When

Creating content designed to spread organically

Example

Social Currency: Makes sharer look good (insider knowledge, achievement) Triggers: Top of mind, easy to remember and recall Emotion: High-arousal emotions (awe, excitement, anger, anxiety) Public: Observable behavior that others can see and mimic Practical Value: Genuinely helpful information people want to share Stories: Wrapped in narrative that carries the message

Example: Dollar Shave Club video hit all six.

---

Name

Double-Sided Incentive Design

Description

Reward both referrer and referee to maximize conversion and quality

When

Designing referral program incentives

Example

Dropbox: Referrer gets +500MB, Referee gets +500MB (symmetric) Uber: Referrer gets $5 credit, Referee gets $20 off first ride (asymmetric) Airbnb: Both get travel credit (aligned incentives)

Test incentive amounts. Too small = no motivation. Too large = fraud risk. Asymmetric often converts better (bigger reward for new user).

---

Name

Friction Audit for Sharing

Description

Systematically remove obstacles in the sharing flow

When

Share feature exists but adoption is low

Example

Measure drop-off at each step: 1. Click "Share" button: 100% 2. Choose sharing method: 60% (40% drop) 3. Customize message: 30% (50% drop) 4. Send invite: 20% (33% drop)

Fixes:

• Pre-populate message (remove step 3)
• One-click sharing to common channels
• Progress indicator to reduce uncertainty
• Social proof ("2,340 people shared this today")

---

Name

Network Effects Moat Building

Description

Design mechanics that make product more valuable as more people join

When

Building products in competitive markets that need defensibility

Example

Direct network effects: Phone network (more users = more people to call) Two-sided marketplace: Uber (more drivers = faster pickups = more riders) Data network effects: Waze (more users = better traffic data) Platform effects: iOS (more users = more apps = more valuable)

Virality gets people in. Network effects keep them in.

---

Name

AI-Accelerated Viral Content Pipeline

Description

Using AI to increase viral content velocity

When

Creating content with viral potential at scale

Example

AI-ACCELERATED VIRAL PIPELINE:

1. TREND DETECTION (AI + Human)

• AI monitors: Exploding Topics, Google Trends
• Human judgment: Which trends fit brand?
• Speed matters: 24-48 hour window

2. CONCEPT GENERATION (AI)

• Generate 10+ concepts per trend
• Use Claude: "Generate viral hook variations for [trend]"
• Filter by brand fit and feasibility

3. RAPID PRODUCTION (AI-assisted) For video:

• Opus Clip for long → short repurposing
• Vizard AI for auto-captioning
• CapCut for quick edits

For images:

• Midjourney for concepts
• Canva for branded templates

4. A/B HOOK TESTING

• Post 3-5 variations of same content
• Different hooks, thumbnails, captions
• Double down on winner within 2 hours

5. ENGAGEMENT AMPLIFICATION

• Respond to comments within first hour (critical)
• Seed to micro-communities (Reddit, Discord)
• Cross-post to all platforms simultaneously

VIRAL VELOCITY FORMULA: Speed × Relevance × Platform Fit × Engagement = Viral Potential

AI accelerates: Speed, Volume of variations Human provides: Relevance judgment, Authentic engagement

Anti-Patterns

---

Name

Referral Spam

Description

Making it easy to spam contact lists with untargeted invites

Why

Damages brand reputation and gets you blacklisted by email providers

Instead

Selective invites with personalization. Quality over quantity. Rate limiting.

---

Name

Incentive-Only Virality

Description

Bribing people to share without making product worth sharing

Why

Attracts wrong users who churn after collecting rewards

Instead

Build product worth sharing first. Incentives amplify, they don't create.

---

Name

Ignoring the Cold Start Problem

Description

Building network effects without strategy for initial users

Why

Network effect products are worthless until network exists

Instead

Start with dense subnetwork (single university, city, niche). Expand after saturation.

---

Name

Generic Share CTAs

Description

Using vague prompts like "Share this" without specific motivation

Why

People need a reason WHY they should share and WHO they should share with

Instead

Use specific CTAs like "Share with your team" or "Send to a friend who needs this". Specific context converts better.

---

Name

One-Way Value Extraction

Description

Asking users to share for your benefit with no value to them or their network

Why

People share when it makes them look good or helps their friends

Instead

Frame sharing as helping others. Use friend-first messaging like "Give your friend $20 off" rather than "Get $5 when friend signs up"

---

Name

Measuring Vanity Viral Metrics

Description

Tracking shares and impressions without measuring actual conversion

Why

Viral "reach" that doesn't convert to users is wasted distribution

Instead

Track full funnel (shares → clicks → signups → active users → retained users)

Viral Marketing - Sharp Edges

Spam Guilt Association

Id

spam-guilt-association

Summary

Making users unknowingly spam their contacts

Severity

critical

Situation

Auto-posting to social without clear consent. "Invite all" that blasts everyone. Users don't realize what they've done until friends complain.

Why

Users feel betrayed when they realize what happened. They never trust you again. They tell others not to trust you. Your brand becomes associated with spam. LinkedIn's early growth tactics still haunt them.

Solution

• Make every share action explicit and opt-in
• Show users exactly what will be sent before sending
• Never auto-select contacts
• Let users customize messages
• Treat their relationships with more respect than they might themselves

Symptoms

• User complaints about spam
• "I didn't mean to send that" messages
• Trust damage
• Brand association with spam

Detection Pattern

Reward Mismatch

Id

reward-mismatch

Summary

Offering rewards that attract wrong users

Severity

critical

Situation

"$20 for every friend who signs up" attracts reward hunters, not real users. Incentivizing signups without engagement.

Why

You pay to acquire users who never engage. Unit economics collapse. Real customers outnumbered by fraudsters. CAC skyrockets while LTV stays flat.

Solution

• Reward activation, not signup
• Gate rewards behind real usage
• Cap rewards per referrer
• Use two-sided incentives (referrer + referee both benefit)
• Monitor for fraud patterns aggressively

Symptoms

• High signup, low activation
• Referral fraud patterns
• Terrible unit economics
• Users disappear after reward

Detection Pattern

False Viral Metrics

Id

false-viral-metrics

Summary

Celebrating K-factor calculated wrong

Severity

high

Situation

Counting invites sent, not invites that converted. Ignoring cycle time. Thinking you're growing virally when you're not.

Why

You make strategic decisions based on false growth. Eventually reality catches up—no hockey stick, just expensive acquisition.

Solution

• Calculate K-factor correctly: (invites per user) × (conversion rate)
• Include cycle time—K=0.8 with 1-day cycle beats K=1.1 with 30-day cycle
• Distinguish true viral growth from referral attribution
• Track cohorts over time, not just snapshots

Symptoms

• K-factor looks good, growth doesn't follow
• Confusion about viral vs paid growth
• Optimizing wrong metrics
• Hockey stick never materializes

Detection Pattern

Incentivizing Non Sharers

Id

incentivizing-non-sharers

Summary

Building referral programs for products people don't naturally talk about

Severity

high

Situation

Trying to make toilet paper viral. Building referral programs for products no one discusses. If people don't already share, incentives won't help.

Why

Nobody shares, no matter the incentive. You waste budget on a program that generates nothing. Virality amplifies existing sharing—it can't create it.

Solution

Before building referral mechanics, ask: "Do people already talk about this product?" If no, fix the product first.

Products that naturally spread:

• Remarkable experiences
• Social identity signals
• Collaborative tools (network effects)
• Shareable content

Products that don't:

• Commodities
• Embarrassing products
• Low-engagement utilities

Symptoms

• Referral program unused
• Low share rates despite incentives
• No organic word-of-mouth
• Incentives don't move the needle

Detection Pattern

Single Channel Dependency

Id

single-channel-dependency

Summary

Building viral strategy around one platform that can change

Severity

high

Situation

Entire viral strategy built on Facebook shares or WhatsApp forwards. Platform changes algorithm—growth dies overnight.

Why

Platform changes algorithm—growth dies overnight. Platform blocks your share mechanism—you're done. You've built on rented land.

Solution

• Diversify share channels from the start
• Build owned channels (email, SMS, direct links) alongside platform-dependent ones
• Design sharing to work even if specific platforms disappear
• Don't over-optimize for one platform's mechanics

Symptoms

• 90%+ referrals from one platform
• Growth swings with algorithm changes
• No owned channels
• Platform dependency risk

Detection Pattern

Friction Overcorrection

Id

friction-overcorrection

Summary

Making sharing so easy people do it accidentally

Severity

medium

Situation

One-click shares that people don't remember making. No confirmation, no preview. Volume over quality.

Why

Accidental shares embarrass users. They feel tricked. Low-intent shares don't convert anyway. You get volume without quality—and damage trust.

Solution

• Confirm before sending
• Show preview of what will be shared
• Let users personalize
• A moment of friction that creates intent produces higher-quality shares

The right friction:

Zero friction → Accidental, low quality Tiny friction (preview + confirm) → Intentional, high quality

Symptoms

• Shares that don't convert
• Users surprised by their shares
• I didn't mean to share that
• High volume, low quality

Detection Pattern

Referrer Reward Only

Id

referrer-reward-only

Summary

Only incentivizing the referrer, not the referee

Severity

high

Situation

"You get $10 when your friend signs up." The friend has no reason to use your referral link vs signing up directly.

Why

The referee has no reason to use your link vs. signing up directly. They might resent that their friend is getting paid off them. Conversion tanks.

Solution

Always use two-sided incentives. The referee should get something—preferably something better than referrer.

Better framing:

"Your friend gets $20, and you get $10 when they join." The referee is the priority.

Two-sided examples:

• Friend gets 50% off, you get free month
• Friend gets extended trial, you get credits
• Both get bonuses

Symptoms

• Low referral link usage
• Direct signups instead of referrals
• Referee resentment
• Poor conversion on referred users

Detection Pattern

Timing Ignorance

Id

timing-ignorance

Summary

Asking for referrals at the wrong moment

Severity

high

Situation

Hitting new users with "Invite friends!" before they've experienced value. Maximum exposure to referral prompts, minimum results.

Why

Users don't know if your product is good yet. They're not going to recommend something they haven't validated. You train them to ignore referral prompts.

Solution

Ask for referrals after moment of delight:

• After successful transaction
• After achieving a goal
• After positive experience
• When they're feeling good about you

Best timing:

"Congrats on [achievement]! Know anyone else who'd benefit?"

Worst timing:

"Welcome! Invite 5 friends to continue"

Symptoms

• Low share rates on prompts
• Users ignore referral requests
• Prompts feel pushy
• No correlation between satisfaction and sharing

Detection Pattern

Vanity Sharing

Id

vanity-sharing

Summary

Adding share buttons for content nobody would share

Severity

medium

Situation

"Share this terms of service update!" Share buttons on every page regardless of share-worthiness. (0 shares displayed)

Why

Unused share buttons look pathetic. Clutter the interface. Train users that your share prompts are irrelevant. When you have something actually share-worthy, they ignore it.

Solution

Only add share mechanics to genuinely share-worthy content. Ask: "Would I share this?" If no, don't add the button. Make sharing rare and valuable, not ubiquitous and ignored.

Share-worthy content:

• Useful resources
• Interesting findings
• Entertainment value
• Social identity signal

Not share-worthy:

• Terms of service
• Basic product pages
• Boring updates

Symptoms

• 0 shares displayed on buttons
• Share buttons everywhere, used nowhere
• Interface clutter
• Users tune out share prompts

Detection Pattern

Forced Virality

Id

forced-virality

Summary

Making the product unusable without inviting others

Severity

critical

Situation

Forcing shares to unlock features. Gating core functionality behind referrals. "Invite 5 friends to continue using the app."

Why

Users resent being forced. They spam their contacts with angry apologies. They churn immediately. The shares are hostile, not enthusiastic. You might get K>1 but terrible retention.

Solution

Sharing should be natural and valuable, never forced. The product should work alone; sharing should make it better. Users share because they want to, not because they have to.

Natural virality:

• Collaborative features (need others to use)
• Shared outputs (content worth sharing)
• Social identity (proud to be associated)

Forced virality (never works):

• Gate features behind invites
• Require shares to proceed
• Punishment for not sharing

Symptoms

• Hostile referrals
• High churn after forced shares
• Negative brand sentiment
• Fake/throwaway invites

Detection Pattern

Referral Fraud Blindness

Id

referral-fraud-blindness

Summary

Not building fraud detection into referral programs

Severity

critical

Situation

No fraud detection. Getting exploited by fake accounts, self-referrals, and organized fraud rings within days of launch.

Why

Fraudsters find your program within days. They extract thousands in rewards. Your economics collapse. By the time you notice, you've paid a fortune for nothing.

Solution

Build fraud detection from day one:

• Flag same-device referrals
• Flag same-IP signups
• Flag rapid referral patterns
• Delay reward payouts
• Set per-user caps
• Monitor for anomalies

Better to miss some legit referrals than to hemorrhage money.

Symptoms

• Suspicious referral patterns
• Same people referring each other
• Burst activity from single sources
• Economics don't make sense

Detection Pattern

Message Control Overreach

Id

message-control-overreach

Summary

Not letting users customize share messages

Severity

medium

Situation

Forcing users to use your exact pre-written copy. Pre-filled messages that sound corporate and fake.

Why

Pre-written messages are obviously pre-written. They don't sound like the person. Recipients ignore or distrust them. The share loses authenticity. Personal messages convert 10x better than templated ones.

Solution

• Provide default copy but make it editable
• Let users write in their own voice
• Guide, don't control
• Personal messages convert 10x better than templated

Good default:

"[Pre-filled but editable message]" User can customize or use as-is

Bad control:

"[Fixed message that can't be changed]" Feels fake, gets ignored

Symptoms

• Obvious templated shares
• Low engagement on shared content
• "Clearly an ad" perception
• Users avoiding share because of forced copy

Detection Pattern

Viral Marketing - Validations

Missing Share Triggers

Id

viral-missing-share-triggers

Severity

error

Type

regex

Pattern

• ^(?!.(?i)(?:share|tweet|post|forward|send|tell|show|tag|mention|spread|retweet|reblog)).$

Message

No share triggers or calls detected - content unlikely to spread virally

Fix Action

Include explicit share prompts: 'Share this with someone who needs to hear it', 'Tag a friend who...' or social sharing buttons

Applies To

• *.md
• *.html

Missing Strong Emotional Hooks

Id

viral-no-emotional-hooks

Severity

error

Type

regex

Pattern

• ^(?!.(?i)(?:shocking|amazing|incredible|unbelievable|surprising|stunning|jaw-dropping|mind-blowing|heartwarming|inspiring|outrageous|hilarious|terrifying|devastating)).$

Message

No strong emotional hooks detected - viral content needs emotional resonance

Fix Action

Create content that triggers high-arousal emotions: awe, anger, anxiety, humor, or inspiration

Applies To

• *.md
• *.html

Complex or Friction-Heavy Sharing Flow

Id

viral-complex-sharing

Severity

warning

Type

regex

Pattern

• (?i)(?:click here to share|copy this link|go to.share|follow these steps|first.then.*share)

Message

Sharing process appears too complex - friction reduces viral potential

Fix Action

Make sharing one-click easy: direct social buttons, pre-written share text, or simple copy-paste

Applies To

• *.md
• *.html

Missing Social Currency (Makes Sharer Look Good)

Id

viral-no-social-currency

Severity

warning

Type

regex

Pattern

• ^(?!.(?i)(?:insider|secret|exclusive|first|early|special access|behind the scenes|not many know|little-known|hidden|hack|tip|trick)).$

Message

Content lacks social currency - people share things that make them look smart, helpful, or in-the-know

Fix Action

Frame content to make sharers look knowledgeable, helpful, or part of exclusive group

Applies To

• *.md
• *.html

Missing Practical Value

Id

viral-not-practical-valuable

Severity

warning

Type

regex

Pattern

• ^(?!.(?i)(?:how to|guide|tips?|ways? to|steps? to|learn|tutorial|free|template|checklist|cheat sheet|tool|resource)).$

Message

No clear practical value - viral content is often highly useful and shareable

Fix Action

Provide actionable tips, how-to guides, or tools that people want to save and share

Applies To

• *.md
• *.html

Missing Viral Visual Elements

Id

viral-no-visual-elements

Severity

error

Type

regex

Pattern

• ^(?!.(?:<img|!\[|<video|iframe|embed)).$

Message

No images or videos detected - visual content is 40x more likely to be shared

Fix Action

Add eye-catching images, infographics, memes, GIFs, or videos that tell the story visually

Applies To

• *.md
• *.html

Too Safe or Middle-Ground

Id

viral-not-controversial-polarizing

Severity

info

Type

regex

Pattern

• \b(?i)(?:maybe|perhaps|possibly|might|could|some say|it depends|on the other hand|balanced approach)\b

Message

Content appears too safe or balanced - viral content often takes a clear stance

Fix Action

Take a clear position or challenge conventional wisdom (while staying ethical and factual)

Applies To

• *.md
• *.html
• *.txt

No Curiosity Gap in Headlines

Id

viral-missing-curiosity-gap

Severity

warning

Type

regex

Pattern

• ^#\s+(?!.(?i)(?:\?|what|why|how|when|secret|reason|this is|you won't believe|nobody tells|most people don't|surprising|truth about)).$

Message

Headline lacks curiosity gap - doesn't create desire to click and share

Fix Action

Create headlines that tease value without giving everything away: 'Why X Does Y (It's Not What You Think)'

Applies To

• *.md
• *.html

Missing Discussion or Debate Trigger

Id

viral-no-controversy-debate

Severity

info

Type

regex

Pattern

• ^(?!.(?i)(?:debate|controversial|disagree|argue|wrong|right|should|shouldn't|best|worst|vs|versus|better than|overrated|underrated)).$

Message

No elements that trigger discussion or debate - reduces comment and share activity

Fix Action

Include elements that spark conversation: comparisons, rankings, controversial takes, or provocative questions

Applies To

• *.md
• *.html

Excessively Long Content for Viral Sharing

Id

viral-too-long-form

Severity

warning

Type

regex

Pattern

• (?:\w+\s+){1500,}

Message

Content may be too long for viral sharing - optimal is 600-1000 words or quick-scan format

Fix Action

Either shorten content or break into scannable chunks with visuals, bullets, and subheadings

Applies To

• *.md

Missing Open Graph Meta Tags for Social Sharing

Id

viral-missing-og-meta

Severity

error

Type

regex

Pattern

• ^(?!.*<meta\\s+property=["']og:)

Message

Missing Open Graph meta tags - will look unprofessional when shared on social media

Fix Action

Add og:title, og:description, og:image, and og:url meta tags for optimal social sharing appearance

Applies To

• *.html

Missing Numbers or Data Points

Id

viral-no-numbers-data

Severity

info

Type

regex

Pattern

• ^(?!.\d+(?:%|\+|x|times)).$

Message

No compelling numbers or statistics - data makes content more credible and shareable

Fix Action

Include specific numbers: percentages, statistics, time savings, or concrete results

Applies To

• *.md
• *.html

Missing Timeliness or Trend Connection

Id

viral-no-timing-relevance

Severity

info

Type

regex

Pattern

• ^(?!.(?i)(?:2025|2024|latest|new|recent|today|this week|current|now|trending|breaking|just|update)).$

Message

Content lacks timeliness or trend connection - timely content spreads faster

Fix Action

Connect to current events, trending topics, or emphasize recency and newness

Applies To

• *.md
• *.html