The Difference Between Virality And Network Effects
Virality generates rapid growth in word of mouth but it cannot replicate the defensibility associated with network effects — Zoom and Slack’s business models are ideal examples of this dichotomy
Over the last couple of months, Covid-19 and subsequent lockdowns have led to a dramatic shift towards remote work tools. Zoom has been one of the biggest beneficiaries of this shift, with daily active users growing from 10M in Q4 2019 to a staggering 200M in Q1 2020. This has also led to a surge of investor and startup activity in the remote workspace. As a result, I have received numerous questions about Zoom, from both investors and entrepreneurs. These questions largely boil down to “How strong are Zoom’s network effects?”. While this is an interesting topic to explore, it is the wrong question. The right questions are: “Does Zoom have network effects? And if not, what makes it defensible?”.
Before we can answer this, we first need to address another, related point of confusion. Zoom’s business model is often misunderstood and conflated with Slack, even though they are very distinct. That said, it is also easier to comprehend the strengths and weaknesses of Zoom’s business model if we understand how Slack works. So let’s begin with that.
Slack: Dual Network Effects
As I have explained previously, Slack’s business model has a lot of similarities with Carta and Github in the way it combines SaaS and network effects. One area in which it differs is that Slack has two different forms of network effects that come into play at different stages of their customer adoption cycle.
Slack’s adoption cycle begins with individuals within a team, who use the free version to collaborate with each other. Slack becomes more useful to them as more collaborators sign up (meeting the definition of network effects). As adoption continues to grow, it eventually reaches collaborators who work cross-functionally. These collaborators act like “network bridges” within the organization, helping Slack spread organically across teams. As more teams begin to use Slack, the more useful it is for all Slack users within the organization. At this point, the structure of Slack’s network effects resembles social networks like Facebook but restricted to the boundaries of the organization. This effectively locks out Slack’s competitors while its adoption continues to grow.
Growing adoption of Slack’s free product within the organization increases the incentive to upgrade to a paid plan. This also triggers the end of intra-organization network effects. At this point, Slack morphs into a pure SaaS product. Since all employees in the organization are mandated to use it, the addition of new users (new employees) no longer adds any value to the product, i.e. it becomes a tool, not a network.
Until recently, the defensibility of Slack’s paid product relied on creating switching costs (much like traditional SaaS products). Paid customers would lose message archives and existing integrations if they moved away from Slack. Then in 2017, Slack announced a beta program to allow collaboration between organizations. This feature, called shared channels, was exclusive to paid plans, i.e. all collaborating organizations had to be paid, Slack customers. As more organizations signed up for Slack’s paid plans, the more useful it became for all of them. In other words, shared channels extended network effects to Slack’s paid product as well, without any organizational boundaries. This should result in rapid adoption, which unsurprisingly is what happened. Adoption of shared channels has grown from 15% of paid Slack customers in January 2019 to nearly 30% in January 2020, even as its base of paid customers increased by 55%. Adoption of shared channels among its largest customers (those paying >$100K) is now a staggering 89%.
To summarize, Slack’s free product has network effects between individual collaborators, but these are limited to the boundaries of the organization. When organizations become paid customers, network effects within the organization end and a new network effect between organizations kicks in. At this point, Slack’s network effects are no longer constrained by any boundaries but continue to be deeply reliant on user identity. As I have previously explained, this combination results in very strong and defensible network effects. While competitors like Microsoft may attempt to clone Slack’s feature set and UX to limit churn among their customers, there is very little they can do to steal Slack’s customers.
Doesn’t Zoom work the same way? Not really.
Zoom: Virality, not Network Effects
Unlike Slack, Zoom’s primary value proposition has always been external facing. A free (or paid) Zoom user can initiate a video call by simply sharing a link, without requiring an account or download at the other end. This makes it especially useful for sales calls (or any meetings with external stakeholders), removing the need to exchange personal account information (e.g. Google Hangouts or Skype). This not only removes a key element of friction but also embeds virality into the product, i.e. users spread the word about Zoom in the process of using it.
Zoom’s virality and product quality have helped it grow rapidly. However, it had to sacrifice network effects in order to accomplish this. The definition of a network effect is that the value of a product increases as more users are added to the network. Network effects are the strongest form of defensibility because the presence of other users on the network prevents churn. But in Zoom’s case, its USP is built on enabling frictionless communication with users who are not on the network. In other words, the value of Zoom’s product is the same irrespective of whether a user is communicating with another user or a non-user. As a result, Zoom has no network effects whatsoever, i.e. it is a tool, not a network.
Crucially, this remains the same whether Zoom is used as a free product or a paid one. Zoom’s primary triggers for free-to-paid conversion involve the limitations of free meetings, such as the number of meeting participants or removing time limits on group video calls. While this follows the same bottom-up sales motion used by Slack, it relies on the virality of its free product and not network effects. In-built virality has certainly helped Zoom scale its base of free users, i.e. its sales pipeline. However, this does not lock out competitors and free-to-paid conversion becomes a function of product preference.
In another contrast with Slack, Zoom’s core value proposition also prevents it from creating a cross-organization network effect. This is because functionality on Zoom’s paid plans is defined simply by the meeting host, not all attendees. In other words, a paid Zoom customer can initiate a video call, without any restrictions, with any number of non-customers. This is a feature, not a bug — Zoom has willingly sacrificed network effects to unlock virality at this scale. This gives Zoom a major UX advantage over competitors, but at the cost of long-term defensibility. If the presence of other users or customers does not influence product value, then it makes sense to keep using Zoom as long as it retains that UX advantage. Despite the current hype, this makes Zoom more vulnerable to cloning attempts from competitors — this is exactly what Microsoft, Google, and Facebook have been working on.
This isn’t to say that Zoom has no defensibility whatsoever. In the absence of network effects, Zoom’s defensibility is limited to creating switching costs via third-party integrations. Customers who use Zoom integrations extensively should have higher switching costs because they would need to set up their integrations all over again. That said, this form of defensibility is linked to extensive, not sporadic, use of integrations. Evidence of this is limited as Zoom’s app marketplace is still in its early days.
What can entrepreneurs learn from this? Sacrificing network effects at the altar of virality can certainly result in near-term growth, as Zoom’s performance has shown. Zoom has proved that you can scale and take a company public purely based on product superiority and virality. However, the likelihood and sustainability of this growth is not entirely in your control. Rather, it is a function of how long your competition remains ham-handed, even after you show them the roadmap to a great customer experience. If you want to control your own destiny, then it would be prudent to follow Slack’s path and bake defensibility into your model from the start.
A version of this article originally appeared on Marker.