Level 3 of the Hierarchy: Creating Virtuous Loops and Network Effects

[network_effects_nerd]

The pinnacle of Sarah Tavel’s Hierarchy of Engagement is Level 3: Self-Perpetuating. This is where you convert user engagement into fuel that powers your company forward through virtuous loops.

What Are Virtuous Loops?

Virtuous loops are flywheels where user activity makes the product better, which attracts more users, which generates more activity. The strongest form is the network effect—where more users directly make the product more valuable for everyone.

Types of Virtuous Loops

1. Direct Network Effects

More users = more value for each user.

• Social networks: Facebook, LinkedIn—your friends are there
• Marketplaces: Uber, Airbnb—more supply and demand
• Communication: Slack, WhatsApp—everyone you need to reach is there

2. Data Network Effects

More usage = better product for everyone.

• Waze: More drivers = better traffic data = better routes for all
• Google Search: More queries = better understanding of intent = better results
• Chainalysis: More blockchain data analyzed = better pattern detection for all clients

3. Content/UGC Network Effects

User-generated content attracts more users.

• Pinterest: More pins = better discovery = more users pinning
• YouTube: More videos = more viewers = more creators
• Reddit: More posts/comments = more readers = more contributors

4. Growth Loops (Not Network Effects, But Still Virtuous)

User activity generates distribution.

• SEO loops: User content gets indexed, brings organic traffic
• Sharing loops: Users invite others or share content externally
• Re-engagement loops: Notifications bring users back

The Pinterest Example

Pinterest’s virtuous loop is elegant:

1. User pins content they like
2. Pinterest learns their interests better
3. Discovery feed improves
4. User finds more relevant content to pin
5. More pins = more content for others to discover
6. More users attracted by rich content library
7. Cycle repeats

The product literally improves just by users using it.

Why Level 3 Matters So Much

Without Level 3: Evernote’s Struggle

Evernote has strong Level 1 (core action) and Level 2 (accruing benefits, mounting losses). But there’s no Level 3—your notes don’t help anyone else. Result: growth plateaued. They acquired users through marketing spend but had no organic flywheel.

With Level 3: Defensible Moats

Companies with strong virtuous loops become increasingly hard to compete with over time. Each new user widens the moat. This is why investors prize network effects so highly.

The Challenge

Not every product can achieve network effects. Some categories are inherently single-player (note-taking, personal finance, individual productivity).

For these, focus on other virtuous loops:

• SEO from user-generated content
• Sharing and referral mechanics
• Platform/ecosystem plays

But be honest: if your product doesn’t have a path to Level 3, growth will require continuous marketing investment. That’s not necessarily bad—many great businesses work this way—but it’s a different model.

What virtuous loops have you seen work particularly well? And how do you think about products that can’t achieve traditional network effects?

[crypto_builder]

Great breakdown! I want to add token economics as a unique form of network effect in web3.

Token-Based Network Effects

In crypto/web3, tokens can create network effects that traditional products can’t:

1. Financial Network Effects

• Token holders are incentivized to promote the network (their holdings appreciate with adoption)
• Early adopters become evangelists with skin in the game
• Example: Bitcoin, Ethereum—holders benefit from network growth

2. Governance Network Effects

• More token holders = more diverse governance = better decisions
• DAOs leverage collective intelligence at scale
• Example: MakerDAO, Uniswap governance

3. Liquidity Network Effects

• More liquidity = better prices = more traders = more liquidity
• DeFi protocols compete intensely on this
• Example: Uniswap, Curve

The Double-Edged Sword

Token network effects can bootstrap growth incredibly fast (look at DeFi summer 2020), but they can also unwind quickly:

• Mercenary capital leaves when incentives dry up
• Token price drops → less incentive to participate → price drops more

Sustainable token network effects need genuine utility beyond speculation. Chainalysis (mentioned in the post) is interesting because their network effect is data-based, not token-based—but they serve the crypto ecosystem.

Question: Do you think token-based network effects are as durable as traditional ones? Or are they inherently more fragile?

[seo_specialist]

SEO loops are underrated as a virtuous loop. Let me break down how to actually build one:

The SEO Virtuous Loop

1. Users create content on your platform
2. Content gets indexed by search engines
3. New users discover you via search
4. New users create more content
5. More content = more indexable pages = more search traffic

Tactics for Building SEO Loops

Make Content Indexable

• Each piece of user content needs its own URL (no infinite scroll without pagination)
• Clean URL structure: yoursite.com/topic/keyword-rich-title
• Proper meta tags, schema markup
• Fast page loads (Core Web Vitals)

Encourage SEO-Friendly Content

• Prompt users to write descriptive titles
• Allow long-form content (not just tweets)
• Support internal linking between content

Examples Done Well

• Pinterest: Every pin has its own page optimized for image search
• Quora: Every question is a landing page targeting long-tail queries
• Reddit: Subreddits rank for topic keywords, posts for specific questions
• Stack Overflow: The canonical example—dominates programming search results

Common Mistakes

Putting all content behind login walls (can’t be indexed)
JavaScript-rendered content without SSR (Googlebot struggles)
Thin content that doesn’t satisfy search intent
Duplicate content across pages

Measuring the Loop

Track: Organic traffic → Sign-ups → Content created → Pages indexed → Organic traffic

If any step breaks down, the loop stalls. Most teams only measure organic traffic but don’t track the full cycle.

SEO loops take time to spin up (6-12 months to see results) but compound beautifully once established.

[notifications_pm]

Re-engagement loops via notifications are powerful but easy to get wrong. Here’s what I’ve learned:

The Re-Engagement Loop

1. User takes action (posts, comments, follows)
2. Activity generates notifications for others
3. Others engage with the content
4. Original user gets notified
5. User returns to respond
6. More activity = more notifications = more engagement

Best Practices

Personalization is Everything

Generic notifications (“See what’s new!”) get ignored. Personalized ones (“Sarah replied to your comment”) drive 5-10x higher engagement.

Hierarchy of notification value:

1. Direct responses to your content (highest)
2. Activity from people you follow
3. Content matching your interests
4. Generic platform updates (lowest)

Timing Optimization

• Send when the user is likely to be available
• Learn individual patterns (some check mornings, others evenings)
• Batch notifications to avoid overwhelming
• Pinterest was famous for send-time optimization

Frequency Capping

• Set maximum notifications per day/week
• Let users control granularity
• Quality > quantity—one high-value notification beats ten spammy ones

The Permission Problem

• iOS and Android permission prompts are scary
• Prime users before asking (“Would you like to know when Sarah responds?”)
• Explain the value clearly
• Honor “no” gracefully—in-app notifications can still work

Anti-Patterns

Manufactured urgency: “You’re missing out!” when nothing is actually happening
Zombie resurrection: Emailing users who’ve clearly churned
Dark patterns: Making it hard to unsubscribe
Notification spam: Training users to ignore you

The goal is building a habit loop, not annoying users into uninstalling.