words on the internet

i'm will dennis. these are my movie reviews and thoughts. i don't proofread before publishing so forgive the editing or lack thereof

Month: July, 2013

Einstein is wrong. Here’s why.

Einstein has a famous quote:

Insanity: doing the same thing over and over again and expecting different results.

I disagree. (And I have quotes and late-night logic to back me up!)

Let’s look at another perspective on repetition from legendary music producer Rick Rubin:

I’m very much of the school of recording more than less. And I always request that artists overwrite. Write as much as possible—and then we can narrow down—because you never really know. The best song you write might be No. 25, not No. 12. For every System of a Down record, we’ve recorded probably 30 songs to get the 12 or 14 that are on the record. The same with Chili Peppers.

And another from content site Upworthy on how to write heading that makes content go viral:

1. You HAVE to crap out 25 headlines for every piece of content.

2. You WILL write some really stinky headlines.

3. Once you start getting desperate, you start thinking outside the box.


5. #24 will suck. Then #25 will be a gift from the headline gods and will make you a legend.

6. Accept that not every headline will be perfect.

7. Then write 25 headlines

And finally, an excerpt from a Vassar statistics textbook:

By the time you reach a sample size of N=30, the shape of the distribution of sample correlation coefficients is virtually identical to that of a normal distribution.

In lay-man’s terms, if you take a sample size of at least 30 from a population, the results of that sample should be representative of the population as a whole.

Repetition of at least 30 seems to be a key factor in guaranteeing representative output from a given population.

A population could be an artist’s songs (like Rick Rubin) or a journalist’s potential headlines (like Upworthy).

It becomes an interesting thought experiment when you define populations broadly.

If you date 30 people, have you statistically met as good as a sexual counterpart as possible? If you create 30 mock ups of your website, have you explored all statistically likely versions?

So is 30 repetitions the magic formula for success?

Let’s ask Thomas “I-invented-the-light-bulb-you-mofos” Edison:

I haven’t failed. I’ve just found 10,000 things that don’t work.


Unfortunately, success is not normally distributed.

The time-tested truth is that success isn’t formulaic or predictable (much to venture capitalists’ chagrin).

There is comfort to be had, however.

Assuming your chance of success is non-zero (aka you’re optimistic), all you have to do is work hard.


Success = hard work (repetition) and optimism (non-zero chance of success).

It makes sense. Your chance of success is some ratio where X number of successes occurs out of Y attempts. X is known. It’s 1. Better get to work on determining Y.

Practically speaking, you never know which investor will say yes, which spouse will accept your advances, which song you write will be a hit, or which headline will go viral.

I find this outlook especially germane in the land of start ups, where success is, on all counts, an outlier.

Despite what Einstein says, leaning on optimism and hard work is really the only sane thing to do when looking for success.

Your chances of hitting it big might still be slim, but with hard work they’re better. Lloyd has the right approach:

So you’re telling me there’s a chance!


Don’t Ask Users to Bring the Beer


Your app is a party. It’s great. You know that. Unfortunately, users don’t.

Accordingly, your app’s first-time user experience needs to act like an invite to the party. You want to drum up excitement, give guests a taste of what to expect, and provide key details so they can prepare to have as much fun as possible.

A succesful party invite gets people interested enough to commit their time to your party. An app’s onboarding experience is no different.

Imagine getting this party invite:

Dear Guest,

Please bring beer. We have a lot of awesome things planned, but to be honest this party is going to be super lame without beer.

Once you’ve bought the beer, you’ll see what I mean.


-Shitty Host

So many apps start their onboarding process with “Invite Your Friends!” – the new-user equivalent of “Please bring beer.”

This is an awful approach to product onboarding, not to mention user growth. You’re putting a lot of the responsibility on a user who can’t properly be an advocate for your service. In addition, many services aren’t significantly improved with friends. “Invite your friends” comes off as a thinly veiled effort to fuel an app’s user numbers.

If you’re looking to build your user base, don’t focus on invites from first time users. Instead, show users the core value of what you’ve built. Give them a taste of the experience. Ideally, boil down your app to the “Ah-Ha” moment. Get the user to that point as soon as possible.(Even before signup if possible).

Product walkthroughs and tutorials, though common, aren’t successful onboarding strategies. They’re boring and have no “Ah-Ha” moments.

Aside: An “Ah-Ha” moment is a single occurence of your app’s core value. It solves the user’s problem. Often it is dependant on a user taking some action. Ah-Ha moments release dopamine and should be the main cornerstone of your product’s user experience. When looking to improve product, think about increasing the speed to and frequency of Ah-Ha moments.

If your app relies heavily on network effects (and the corresponding “empty room” problem), craft a unique onboarding experience that hacks the Ah-Ha moment. Create fake users, manually generate content, make a “single-player mode.” If the onboarding process doesn’t mimic the full in-app experience 100%, that’s OK. It probably shouldn’t.

Any good app gets users to the Ah-Ha moment during onboarding just like any good party has a well crafted invite. Focus on getting users to have an Ah-Ha moment and RSVP to your party. If they’re a good guest, they’ll show up with friends and a six pack of cold ones.

If you want to see how we handle the onboarding process, sign up for early access to our app Hollerback.

Let me know if I can be helpful.

will@hollerback.co // @willydennis