This post was originally published in 2012 on the Startup Grind blog, I've cleaned it up, updated it a bit, and republished it here for archival purposes (also I was pretty surprised I hadn't posted it here already)
Getting a ton of users is the hardest part of any business, that being the case, why would you choose a business model that depends on you getting a ton of users? -Ben Thompson
The answer to that question is obvious, you choose this business model because the rewards are ginormous! But when you do choose it, it's the only thing that matters. Scaling? Hiring? Funding? All that is cake compared to getting millions of people to use your product, which is the new baseline for success.
In all my internet travels, I've seen a lot of sites come and go, I've even tried my hand at one or two of my own, and one thing all content sites have in common is how they turn a random visitor into a loyal user--and how they lose that same user later down the road. I call this pattern the law of three:
- A user has to visit a site three times in order to become loyal.
- A loyal user needs to find three interesting things each visit to stay loyal.
- A loyal user is lost if they don't find three interesting things three times in a row.
Simple, right? Yet most entrepreneurs, when they talk about acquiring new users, they talk sign up flows, a/b testing, and viral coefficients. Those are all important, but they should all come after making sure their content is interesting to the users they are trying to attract. Otherwise, no matter how many users they get coming through the door, those users just turn around and go right back out again.
My favorite anecdote about greasing viral coefficients before getting the interesting content part in place (aka, putting the cart before the horse) was a startup that so effectively greased their viral coefficient, using techniques learned from Greg Tseng, that they literally increased their user sign ups from hundreds a day to tens of thousands a day but eventually had to rollback to their original--less successful--funnels because they were losing most of those new users due to a lack of compelling content.
What sites people visit to get their content fix is constantly changing (Fun question, how many sites do you visit today that you also visited five years ago?), and each new site that rises up abides by the law of three. Currently, my wife is addicted to Pinterest2 because every time she visits she's flooded with tons of new and interesting pictures. As an aside (because it made me laugh), when I asked my wife why she loves Pinterest so much, she responded:
It makes you feel artistic and crafty just by looking at the pictures even though you didn't actually do anything
In contrast, there was a time I couldn't go even a few hours without checking Digg4. Then one day I didn't find anything interesting. After a few more visits with nothing interesting catching my eye, I stopped regularly visiting Digg altogether, and I haven't really been back since. Users only have a certain number of content sites they visit regularly and so once you've lost a loyal user, it's really really hard to get them back. In fact, it might even be harder than getting them in the first place5.
Having trouble getting users to your own site? Step one is figuring out what things are interesting to the people you are trying to attract. After all, what your users get from your site is your core mechanic, and if it isn't up to snuff you'll never attract the audience you so richly deserve.
In 2012 I was using site only for website, now I'm using it as a standin for app or website. ↩
Fun fact, this was from 2012 and in 2017 my wife still visits Pinterest regularly. ↩
I've visited Ars Technica in three different decades, having first visited it in 1999. ↩
Digg was an entirely different site and company when this was originally written. ↩
As exhibit A I give you Twitter, who mentioned they've had over one billion sign-ups, but only a few hundred million active users, despite near constant media attention and exposure, those 700 million users are probably never coming back. ↩