Interview With Ed Baker About Growth

Ed Baker is known for his expertise on growth. He is the former co-founder and CEO of, a site acquired by Facebook in 2011. Ed has been working on viral growth for the past fifteen years across multiple mediums, including email, SMS, IM, and Facebook. Since the 2007 launch of the Facebook platform, he has worked on several Facebook applications that have grown to tens of millions of users, including Compare People and Send Hotness. He has an A.B. in Chemistry & Physics from Harvard, and an MBA from Stanford. Ed is now the Head of Growth at Uber.

Interview With Edward Baker About Growth

Here are my 5 questions with Ed Baker about growth and my summary of his answers:

1) What is a comprehensive way to sum up the quantitative nature and entrepreneurial meaning of “viral factor” while outlining an actionable system to achieve it?

The term viral factor is the quantitative measure of a viral loop. Here is an example of a viral loop, solving for x*y*z:

The Viral Loop
The Viral Loop as explained by Ed Baker

If you have a viral factor greater than one, then you will see exponential growth without having to spend any money on user acquisition. In order to get your viral factor above one, you must multivariate test every step of your viral loop.

Step One

Maximizing Install Rate

  • Create a catchy call to action
  • Analyze click data
  • Multivariate test invites

Step Two

Maximizing Invite-Sending Rate

  • Create incentives to invite other friends
  • Multivariate test the messaging and the invitation

Step Three

Maximizing Average Invites

  • Create incentives for inviting more users
  • Optimize the number of required invites to receive the incentive

Viral growth can also be combined with paid user acquisition. As long as Average Revenue per User rate is greater than Customer Acquisition Cost, keep tweaking and spending; you’ve created a cash machine.

2) Regarding viral campaigns, where is the line between the organic nature of a viral campaign and the ability for savvy creatives to engineer something viral? In other words, how does a viral engineer bake in the necessary mojo needed for something to go viral when one of the major ingredients of becoming viral is candidness and being genuine?

I think that “creatively engineering something” can actually mean a couple of different things. One, using your example, means creating compelling content that appeals to a wide audience so that it gets shared amongst friends. Any company that creates content and fools the user into thinking that an event actually took place, when in fact the event was staged, produced, and/or engineered, runs the risk of suffering consumer backlash. Then the question is whether the cost of the backlash is worth the gain from exposure.

When I think of “viral engineering”, I tend to think of split-testing and tuning each step of the loop I discussed in your first question. There is just as much creativity involved in engineering the loop as there is in producing the content. Start with something organic in nature, such as user statistics or user input. By design this can’t be faked by you, it is user generated content. What you can influence is the viral factor of the loop by measuring the conversion rate of each step in the loop and making adjustments to improve conversion wherever possible. That is where I believe creative engineering can have the most significant impact.

3a) With the Pareto Principle in mind (the 80/20 rule), what are viral marketers caring about, or doing too much of, that in your opinion could be time better served (the 80)?

Copycatting. I suppose there is some benefit to improving upon something that is already working but these days copycatting is overly exploited and continually dilutes the effectiveness of viral campaigns. You see this all the time on Facebook.

3b) What should they be doing more, that they’re doing not enough of (the 20)?

Not copycatting. Facebook and other social platforms provide so much data and so many interesting and new ways to reach people. Facebook Connect is used by a few in really innovative ways. It would be great to see more of this.

3c) What are they not doing at all (but should be learning) because they haven’t heard of it yet and it is coming?

Mobile. If you get ahead of the curve regarding mobile (and the future of viral activity on mobile devices) you will really be one step ahead of the game.

4) Can there be a legitimate business purpose for novelty viral initiatives like the current, “I feel sad today please LIKE me to make me happy” Facebook campaigns other than disingenuous lead farming?

No, not one I can think of.

5a) What is your favorite viral effort that didn’t find an audience (i.e. failed)?

While at Demigo we created an iPhone application that audited your calls and texts to identify your “top friends”. The application worked with Facebook and would identify the people you communicated with the most, add them to your contact list, and pull in their Facebook data (which was not being done yet at the level it is being done today). It ultimately failed for a couple of reasons. One, the iPhone was still new so the application’s audience was not wide enough to support the critical mass needed to make it successful. Two, the application also required your ATT password and requiring this led to low user compliance. It was a great application though.

5b) What is your least favorite viral effort that amazingly did find a wide audience (i.e. succeeded)?

Facebook applications that use random number generators for content, for instance, “Your lucky number for the day is X.” Who cares? This is just random meaningless data. However, people use these types of applications all the time.

5c) What is a question people should be asking about viral marketing that isn’t being asked enough today?

Entrepreneurs need to start considering the tradeoff between short-term virality and long-term engagement. There is value in keeping users around. The question that they should be asking is, “how can we truly engage our users?” They should think less about getting User X to invite Y number of friends immediately. Instead, they should think about keeping User X around as long as possible in order to get him to invite many more friends over a longer period of time.

In order for this to work, entrepreneurs should think more about making things viral by creating a great product. It seems obvious, but if you are more focused on how to reach users than on how to make a great product you’re doomed. Dropbox is an example of a great product that also has great virality as a result. I want to tell you and my friends about it because it is a great product.

A lot of new ventures lose sight of these basic fundamentals in order to chase the next big, new, exciting trend. They should really be asking themselves, “how can I make a product that people will use and enjoy so much it will inspire them to willingly broadcast their experience to their network?”