Viral Loop

What is a viral loop?

Viral loop (German: virale Schleife) denotes a repeating cycle in which use of a product itself brings in new users: a user invites others or shares a result, some of the recipients register, become active and continue the cycle. The strength of the loop is measured by the viral coefficient (K-factor). It states how many new active users an existing user wins on average. This is comparable to a multiplayer board game: anyone who wants to play it needs fellow players, and some of them then buy it themselves and in turn invite others.

Viral loops are a tool of customer acquisition, above all for software products that customers try and buy themselves without a sales conversation (product-led growth, PLG). In funding rounds they play a role because they lower the average cost of customer acquisition and thereby improve the unit economics, that is, the profitability per customer. Investors therefore examine closely whether claimed viral effects are backed up by data.

The viral coefficient

The viral coefficient, usually called k or K-factor, is the product of two quantities. The first is the average number of invitations or shared items per user. The second is the share of recipients who register and become active (conversion rate). As a formula: k = invitations per user × conversion rate. Anyone who sends four invitations on average, of which 10% lead to active users, reaches a k of 0.4. Both quantities can be influenced: the number of invitations through the product itself, the conversion rate above all through the onboarding of new users and its design (UX/UI).

Each wave of invitations produces a new group of users, a generation. If k is above 1, each generation wins more new users than it comprises itself, and the user base grows of its own accord until the market is saturated. If k is below 1, each wave fizzles out. That is not self-sustaining growth, but it multiplies every user won through paid channels: one purchased user becomes a total of 1 divided by (1 minus k) users.

Besides k, the cycle time counts, that is, the time from a user's registration to the registration of those invited by that user. A loop with a short cycle time runs through more often in the same period and can therefore bring more users than a loop with a higher k and a long cycle time.

Types of viral loops

• Built-in virality: The product delivers its benefit only with others, for example in joint editing of documents, scheduling or payments between people. The invitation is part of the use itself. This form is the most stable.
• Visible results: Work results that the product generates carry a reference to the provider, such as shared reports, forms or presentations. Recipients thereby get to know the product.
• Referral programmes: Users receive a benefit for successful referrals, often the person referred as well. Such programmes work quickly but cause costs and partly attract users who come only for the benefit.

Legal limits in Germany

Viral mechanisms touch competition law and data protection law. Under German competition law, advertising by email is in principle permitted only with the recipient's prior express consent. According to the case law of the Federal Court of Justice (Bundesgerichtshof), referral emails that a product sends to third parties at a user's instigation count as advertising by the company. They therefore also require the recipient's consent; the consent of the inviting user is not sufficient. Uploading address books also raises data protection questions, because personal data of third parties who have not consented is processed.

For investors and buyers this is not a peripheral topic. If a substantial part of growth rests on legally vulnerable mechanisms, cease-and-desist letters, fines or being forced to switch off the function are risks. In due diligence, the examination before a financing or purchase, it is therefore checked how invitations are designed technically and legally.

Significance for planning and valuation

For valuation, what counts is whether viral effects sustainably lower customer acquisition cost (CAC). Investors compare the cost per customer won through paid channels (paid CAC) with the average cost across all customers including those won virally (blended CAC). The larger the gap, the more the loop contributes. k is measured per cohort (cohort analysis), that is, per group of users won in the same period. An overall average conceals that early cohorts often show different values from later ones.

In robust planning, the viral effect lowers the average acquisition cost rather than appearing as an independent growth driver. The financial model therefore works with users won through paid channels and a measured uplift for those won virally, rather than extrapolating the user count from k alone.

Worked example

A provider of scheduling software wins 2,000 users through paid advertising at a cost of EUR 60 per user, EUR 120,000 in total. Each new user sends on average 2.5 invitations, of which 24% lead to an active user. The cycle time is 14 days.

• Viral coefficient: 2.5 × 0.24 = 0.6
• First generation: 2,000 × 0.6 = 1,200 new users
• Second generation: 1,200 × 0.6 = 720, third: 720 × 0.6 = 432, fourth: 432 × 0.6 = around 259
• Users after four cycles, that is, 56 days: 2,000 + 1,200 + 720 + 432 + 259 = 4,611
• Average cost per user: EUR 120,000 divided by 4,611 = around EUR 26.02 instead of EUR 60
• Theoretical upper limit: 2,000 divided by (1 minus 0.6) = 5,000 users, that is, EUR 24 per user

Even a k well below 1 is therefore valuable: each paid user theoretically brings 1.5 more, and the average acquisition cost falls by more than half. With a k of 0.3, there would be only 2,000 divided by 0.7 = around 2,857 users. The cycle time decides the pace: at 28 instead of 14 days, only two cycles would have run after 56 days and 2,000 + 1,200 + 720 = 3,920 users would have been reached.

Distinctions from related terms

• Network effects: The product becomes more valuable to each user the more others use it. A viral loop, by contrast, describes how new users are won. Both often occur together but are independent of each other.
• Word of mouth: Recommendations outside the product, for example in conversation. It works similarly but can hardly be measured and cannot be steered in a targeted way.
• Freemium: a pricing model with a free basic version. It lowers the barrier for those invited and thereby strengthens viral loops, but is not one itself.

Common mistakes

• Looking at k without cycle time. A high coefficient with a long cycle time may bring fewer users in the planning period than a lower one with a short cycle time.
• Counting invitations instead of active users. What counts are registrations that lead to active use, not messages sent.
• Assuming k above 1 permanently. As market penetration grows, k falls, because more and more of those invited already use the product.
• Incentives without a convincing product. Referral rewards bring users who quickly leave again without real benefit.
• Legal design neglected. Invitation emails without the recipients' consent and unchecked address book imports are open to challenge.

Frequently asked questions

What viral coefficient is realistic?

That depends on the product. Values permanently above 1 are rare and mostly tied to products that are useful only together with others. Values well below 1 are also economically valuable, because they noticeably lower acquisition costs.

How is the viral coefficient measured?

Per cohort: for a group of users won in the same period, the number of active users it has won through invitations or shared content is counted. A prerequisite is that the origin of each registration is recorded technically.

Do viral loops work in business with companies?

Yes, but differently from consumer business. Spread usually takes place within an organisation and through work results that are shared with customers and partners. The purchase is then often made centrally, after use has already spread.

Does a viral loop raise the company valuation?

Indirectly. Lower acquisition costs improve the ratio of customer value to acquisition cost (LTV/CAC) and the capital efficiency of growth. Investors reward this if the effects are shown to be stable across several cohorts.

The Roemer Capital view

We include a viral coefficient in a client's financial model only if it has been measured per cohort, and never permanently above 1. A model that rests on self-sustaining growth would not survive due diligence and would damage the credibility of the other assumptions.

If a client, on the other hand, can show that each customer won through paid channels brings further customers on average, that is a verifiable argument. It belongs in the equity story, the common thread of investor outreach. For this we set the cost per paid customer and per customer won overall side by side and test them against an acquisition cost payback period (CAC payback period) of under twelve months.