Viral Loop
A viral loop is a self-reinforcing cycle where existing users bring in new users, who in turn invite others. It lets a product grow through its own usage rather than paid acquisition alone.
Key takeaways
- A loop closes only when the recipient reaches the same shareable moment themselves.
- Cycle time compounds; shortening the delay between hops often beats raising acceptance.
- Inherent loops close most reliably; incentivised loops attract reward-seekers rather than users.
- Every hop reaches a more saturated audience, so loops decay without new entry points.
- Count generations, not shares; a strong first generation with no second is a launch.
In depth
A viral loop is a closed sequence: a user reaches a moment worth passing on, a share surface makes passing it on easy, the recipient lands on something that makes sense without context, and the recipient eventually reaches the same moment themselves. The loop only closes at that last step. If recipients arrive but never reach the shareable moment, what you have is a one-hop broadcast that decays, not a loop that repeats. Each step has its own drop-off, and they multiply.
Loops differ by what motivates the hop. Inherent loops, where the product only works once you bring someone in, close most reliably but only fit collaborative products. Incentivised loops buy reach with rewards and attract people who want the reward rather than the product. Content loops depend on the shared artefact being interesting to strangers. Cycle time is the second lever and the underrated one: halving the delay between hops doubles how many hops fit in a quarter, which usually beats a small lift per hop.
In practice, each step gets its own event and every share link carries an identifier so hops can be chained into generations. A scorecard quiz is a convenient loop because the shareable moment and the entry point are the same page: a personal, comparative result is worth passing on, and the recipient can reach an identical result in one sitting without installing or configuring anything. That short path from landing to shareable moment is why assessment loops close faster than loops requiring onboarding.
Loops run inside finite networks, so each generation reaches people who are closer to those already reached, and the loop slows even when nothing about the product changed. They also break for reasons outside your control: a platform that changes its sharing affordance can flatten a loop overnight. A loop with a large first generation and a negligible second is a launch, not a loop, and reporting it as virality invites a repeat that never comes. Loops amplify a product; they do not repair one.
Example in practice
How to measure it
Measure the loop as a chain, not a number. For one entry cohort, count how many people reached the shareable moment, how many shared, how many recipients arrived, and how many of those recipients reached the shareable moment themselves. Those four counts give you the drop-off at every step. The ratio of second-generation entrants to first-generation entrants is the loop's real output.
Then measure elapsed time between the share and the recipient's own share, taking the median rather than the mean because a few late hops distort the average badly. Plot that median alongside the generation ratio over successive cohorts. A stable ratio with a lengthening median means the loop is saturating; a falling ratio with a stable median means the shared artefact stopped being worth passing on.
Common mistakes
Teams instrument the share button and stop there. Share counts tell you the loop started, not that it closed, so a dashboard can show growing shares while every generation gets smaller. Tag each share link with the sharer and the generation number, then report how many second- and third-generation entrants each cohort produced. If generation two is near zero, the share button is not the problem, the landing experience is.
The other failure is adding a reward before the loop works mechanically. Incentives raise the number of invitations sent and leave the acceptance rate untouched, so you pay more per hop for the same growth and import people who churn once the reward clears. Fix the recipient's first thirty seconds first: make the shared link land on something immediately intelligible, then decide whether an incentive is still needed at all.
Frequently asked questions
What is the viral coefficient?
The viral coefficient, often called k, is the average number of new users each existing user successfully invites. When k is greater than 1, the loop grows on its own without paid acquisition.
Do you need a viral coefficient above 1 to benefit?
No, most products stay below 1 yet still gain a lot. Even a coefficient of 0.3 to 0.5 meaningfully lowers blended CAC by supplementing paid and organic channels.
Why does cycle time matter in a viral loop?
Cycle time is how long one full invite-and-join loop takes, and shorter cycles compound faster. Reducing friction in sharing and signup can boost growth as much as raising the viral coefficient.
What is the difference between a viral loop and a referral program?
A viral loop is the mechanic; a referral program is one way to power it. The loop describes the full cycle from a user's shareable moment through the recipient reaching that same moment. A referral program adds an explicit reward and tracking layer on top. Loops can run with no reward at all when the shared artefact is interesting enough on its own.
How long does it take to know whether a viral loop is working?
At least three full cycle times, because you need a first, second and third generation before the trend is real. If a hop takes a week, that is roughly a month of data. Launching and reading the numbers after five days measures the launch push, not the loop, and almost always produces an encouraging figure that later collapses.
Can a viral loop work for B2B products?
Yes, though it usually runs inside organisations rather than across the open internet. The loop hops from one colleague to another, so the network is a company rather than a social graph, which caps each loop at the size of the account but makes acceptance rates far higher. Benchmarks, assessments and shared workspaces are the artefacts that travel best in that setting.
What breaks a viral loop most often?
The recipient's landing step. A share arrives with context the sender had and the receiver does not, so the page has to explain itself in seconds or the hop ends there. The second most common break is a shareable moment that arrives too late in the product, after most users have already left. Both show up as high shares with a tiny second generation.
Should I shorten the cycle or improve the acceptance rate?
Shorten the cycle first if the delay is measured in weeks, because compounding is exponential in the number of hops and only linear in each hop's yield. Once a hop takes a day or two, further speed gains matter little and acceptance becomes the binding lever. Look at your median hop time before deciding; the answer follows from it.
Does a viral loop replace paid acquisition?
Almost never on its own. A loop below one hop per user amplifies whatever enters it rather than generating growth independently, so paid and organic entry points still have to fill the top. What the loop changes is the effective cost of each entrant: every paid signup that produces a fraction of a further signup lowers the blended cost of the cohort.