This piece explores the full arc of viral loops, from the Web 2.0 golden age to the mobile era and into the AI age. It explains how viral loops are built, how they are measured, why retention matters as much as sharing, and how modern growth tactics differ from the earlier era of email invites and social graph exploitation. The broader lesson is that virality is not magic. It is a measurable product system, and like every system, it changes as platforms and user behavior change.

1. The Golden Age of Viral Loops

Between roughly 2005 and 2010, many products were intentionally designed to spread from user to user at massive scale. Social networks, collaboration tools, messaging products, and user-generated platforms all benefited from viral mechanisms that were systematic, measurable, and optimizable.

The author sees that era as a time when the industry truly learned how viral growth worked. Many of the people who mastered those loops later became founders, major tech executives, or top investors.

As the world shifted to mobile, much of that practical knowledge faded from view, but the underlying ideas still matter today, especially for:

• product-led growth,
• referral systems,
• and AI apps built around sharing output.

2. Viral Growth as an Equation

The article distinguishes true viral loops from generic buzz or word of mouth. A viral loop is designed into the product through things like:

• invitations,
• referral links,
• tags,
• or shared content that pulls new users back into the tool.

The key concept is the viral factor:

Viral Factor = number of new users created by a cohort / size of that cohort

If 100 users generate 150 new users, the viral factor is 1.5. If they generate 50, it is 0.5. Below 1, the loop eventually dies out.

To measure this well, products need to attribute new users to specific sharers, often through encoded referral links. Once that data exists, teams can track viral performance by cohort and improve the loop through product changes.

3. Why Retention Matters as Much as Virality

The piece warns against obsessing only over shares and conversion rates. A product can generate millions of signups and still fail if it does not retain people.

The deeper issue is stickiness. Viral loops can create bursts of growth, but without retention they become short-lived spikes. That is why the author highlights additional quality indicators such as:

• flattening retention curves,
• strong active-user ratios,
• habitual use,
• and sustained organic acquisition.

Web 2.0 produced many apps that spread quickly but lacked enough retention to become durable businesses. The same pattern can repeat today, especially in simple AI products.

4. Two Types of Viral Products

The article splits viral products into two categories.

4.1. Simple, Highly Viral Apps

These products center on a single, highly shareable behavior. They often spread fast because the core action is easy to create and easy to pass along. Early Instagram and some generative AI apps fit this model.

Their strengths are speed and simplicity. Their weakness is that they often depend on novelty and may collapse once attention fades.

4.2. Deeper, Stickier Products with Sharing

These products are harder to build, but once they win users, they keep them. Their viral growth is slower but more durable. Examples include tools like Slack or Figma, where retention and workflow depth create long-term compounding.

The article argues that many current AI products resemble the first category: highly viral, but potentially fragile if they cannot deepen usage over time.

5. Why Viral Performance Decays

Even successful viral loops usually weaken over time. The article gives three main reasons.

5.1. Novelty Fades

When a product category is new, people are more likely to click, try, and share. As the category matures, the novelty premium disappears and the bar rises.

5.2. Markets Saturate

The more successful a viral loop becomes, the more it runs into people who have already seen the product or have already decided they do not care. Conversion naturally falls.

5.3. Platforms Fight Back

Viral loops almost always operate on top of another platform such as email, Facebook, YouTube, or TikTok. Once a host platform sees behavior as spammy or threatening, it can suppress or block it. That is part of why older invitation-based growth tactics became less effective over time.

6. The Mobile Shift and the End of Web 2.0-Style Virality

One of the strongest arguments in the article is that mobile largely killed classic Web 2.0 virality.

Email-based growth once worked because uploading an address book and blasting invitations to hundreds of people was normal enough and technically feasible. On mobile, that flow became clunkier, less socially acceptable, and more tightly regulated. SMS-based invites quickly ran into spam problems and platform resistance.

As a result, the old world of supercharged viral loops weakened dramatically.

7. Modern Virality and AI

Today, many people confuse virality with launch trailers, ragebait posts, influencer clips, or founder-led social media. The author does not dismiss those tactics, but argues that they are usually one-off attention mechanisms, not durable viral systems.

What matters more is whether the product can keep expanding the ratio of new users to active users over time. That requires loops embedded in ongoing behavior, not isolated marketing bursts.

This is why AI products are so interesting. They often generate outputs that are inherently shareable, making them ideal for simple viral loops. But they also risk becoming short-lived phenomena if they do not build retention, product depth, or network effects around those outputs.

Conclusion

The piece's broader message is that virality is not a mystical property. It is a product system shaped by:

• measurable loop design,
• user behavior,
• retention,
• market saturation,
• and platform dynamics.

The Web 2.0 era revealed how powerful viral loops could be. Mobile changed the rules. AI may be opening a new chapter, especially around shareable generative output. But the old lesson still holds: virality without retention is fragile, and growth without product depth rarely lasts.