Select Page

36 panda

Back when “viral infection” was scary…

These days, the word “viral” is strongly associated with something spreading like wildfire online – but this wasn’t always the case.

In fact, just a few years ago, folks likely would have responded in a much more fearful way when asked about viral infection as they would have associated it with the spread of – you guessed it – a virus. As obvious as this may seem, the minds of most people today immediately go straight to “something on YouTube”.

While the science and mechanics virality and viral “infection” can be tough for startup founders and business people to implement (or even understand), it’s been something that has been deeply studied at great length by virologists at the CDC and other infectious disease prevention centers and medical research organizations. These organizations make the spread of disease their main focus, so it can be helpful to look to their contagion models to help us quantify viral “infection” for the purposes of creating a user-generated viral growth engine.

That said, once you’re able to model viral infection mathematically, you should have taken most of the value from studying old-school contagion models, and it won’t help you a whole lot in a practical sense to continue to think through this lens. An actual virus has completely different variables, and it makes no sense to worry about those complexities.

You’ve got a product and you want it to grow itself. That’s why you’re here.

So – for simplicity, let’s assume you’re building a live chat tool – such as Olark or IntercomThey goal is to help brands engage prospective and existing customers in a more efficient way.

How could this tool be viral?

  • It could use viral collaboration marketing by adding a feature that lets users loop in their team members to engage the brand providing customer service through the tool.
  • It could use viral communication marketing by adding a feature that lets users email a transcript of the conversation to friends or coworkers after it’s over.
  • It almost certainly WILL use embeddable viral marketing as companies will typically embed a tool like this within their own website or app, exposing it to that site’s visitors.

intercom

There are likely a tremendous amount of other viral marketing engines that can be implemented with a tool like this. For this reason – AND because you’ve spent ample time developing it and making it legitimately valuable in contrast to the other solutions available at the moment – let’s assume you’ve build what you believe is a strong viral loop.

 

But there’s a problem. You haven’t figured out how to quantify that growth yet, which you need to do to optimize your viral loop(s) and assess how to improve.

Thankfully – you’ve discovered Viral Hero in time. We’ll fix that.

The first bit of viral math

So – you’ve worked hard on your product and your viral loop, and you’ve decided that it’s time to expose your product to the world. You start by inviting 10 of your closest friends, and since they know and love you, they all sign up as your first 10 users.

This kicks off our math with a base level of 10 users, or…

u(0) = 10

Not so hard so far, right?

Our initial 10 users seem to totally get it. They all love the experience, and you’ve made the additional value they receive for inviting others both obvious and compelling. As a result, they each send out an average of 10 invitations to their friends.

Let’s create a new variable called i, which equates to the total number of friends your users each send on average during a selected period of time. So for us…

i = 10

Still with me?

Next – for every batch of 10 invitations that get sent out, two of the people who receive those invites respond favorably and join themselves. Let’s create a new variable called conv% and calculate it like this:

conv% = 2 new users / 10 invites sent = .2 (or 20%)

So to recap, we started with a base amount of 10 users (u(0) = 10), who wound up sending 10 invites each (i = 10). Those invites had a 20% average conversion rate (conv% = .2).

The total customers at the end of the first full “cycle” (or the amount of time for all this to take place) would equate to the initial 10 users, plus the new 20 (calculated from 100 total invites * 20%).

This leaves us with 30 total customers, and a K of 2.0.

But wait…what is K?

K is a measure of potential viral magnitude. It’s helpful and necessary to quantify any meaningful viral KPIs, but it’s a bit short-sighted when used by itself. I’ll explain why later on, but for now, know that…

K = i * conv%

So for us, so far…

K = 10 * .2 = 2.0

This tells you that at this point, on average, for every user you acquire through non-viral means, they’ll bring another two users to you as well via your viral loop. (This K factor is absolutely incredible, and also very rare.)

Since we “seeded” our viral engine with 10 non-viral users, applying our K to them give us 20 MORE users, for a total of 30 users.

Piece of cake so far, yeah?

Now in all likelihood, those new 20 customers will most likely send out a similar number of invitations themselves, beginning brand new viral loops. The users they recruit will then recruit new ones themselves from their own viral loops, and so on.

The original 10 you seeded your viral engines with may sporadically send invites, but these will drastically drop to a crawl as they both max out their perceived viral value, and also run out of others they want to invite. This is called viral decay, and is something we’ll go into later in a lot more detail.

However, given this, it’s highly unlikely (scratch that – impossible) that your entire population of users will continue to send out invites during every cycle. Bank on seeing a quick spike when users initially see viral value, and then a dramatic drop to a slow trickle after that.

With a K of 2.0, you’ll see true “viral growth”. This is a compounding, exponential process that’s as rare and difficult to achieve as it is lucrative and powerful – but a substantial viral education will give you a massive edge in the probability of reaching it.

Travis Steffen
Follow

Travis Steffen

Travis Steffen is a Silicon Valley growth engineer, data scientist, and serial entrepreneur with multiple exits. He currently serves as Head of Growth at AutoLotto. He's also a crazy adrenaline junkie, is obsessed with fantasy football, and can grill a mean rack of ribs.
Travis Steffen
Follow
Viral Buttons by UP