One of the bigger stories from Mike Isaac's Travis Kalanick profile had nothing to do with Uber at all. The relevant text:

With Mr. Kalanick setting the tone at Uber, employees acted to ensure the ride-hailing service would win no matter what. They spent much of their energy one-upping rivals like Lyft. Uber devoted teams to so-called competitive intelligence, purchasing data from an analytics service called Slice Intelligence. Using an email digest service it owns named Unroll.me, Slice collected its customers’ emailed Lyft receipts from their inboxes and sold the anonymized data to Uber. Uber used the data as a proxy for the health of Lyft’s business. (Lyft, too, operates a competitive intelligence team.)

Reactions to this have ranged wildly, from the paranoiac to the flatly cynical:

But the classic "if you're not paying for it, you're the product" refrain is a bit wrong in this instance. There are plenty of products in the freemium universe that do not sell your personal information directly to the highest bidder. There are also many different ways that products can monetize user data, each at a different point along the scale of acceptable/ethical. 

Some monetize through advertising, as Unroll.me did prior to their acquisition. Some function as lead-generation tools. Some take your data and bundle it up into packages amenable to marketers (see: Gmail Sponsored Promotions). Some don't even sell your data, even though they're commonly cited as doing so. They buy it.

There are lots of ways to skin a cat. Unroll.me chose the unsophisticated and back-handed sale of email data without taking necessary safeguards we should expect from "freemium" products monetizing on our information (from HN):

I worked for a company that nearly acquired unroll.me. At the time, which was over three years ago, they had kept a copy of every single email of yours that you sent or received while a part of their service. Those emails were kept in a series of poorly secured S3 buckets. A large part of Slice buying unroll.me was for access to those email archives. Specifically, they wanted to look for keyword trends and for receipts from online purchases.

Lastly, incentives.

Facebook, Google, and Twitter have a lot riding on the fact that they can properly "anonymize" and protect their users' sensitive personal information. They have to keep users around, or they will die. 

Slice Intelligence, not so much.

quitting smoking is easy

I'm a few days shy of 100 days off cigarettes, and the most interesting or notable thing I've learned since quitting is that it's really very easy.

You hear all the time that quitting smoking is difficult. Not just difficult, the hardest thing in the world. Harder than quitting heroin. From that NYT article:

[The addiction's] hooks go deep, involving complex physiological and psychological mechanisms that drive and maintain smoking behavior.

There's definitely truth to this. 

Physically, maybe not. Current consensus says that nicotine is almost entirely expelled from your body after just a few days. 

Psychologically, yes. There are social triggers, environmental triggers, behavioral triggers.

The most powerful psychological mechanism keeping smokers from quitting, however, might just be the perception of its difficulty.

To this day, thinking about quitting as a struggle or a lifechanging event gives me anxiety. It makes me worried that I'll trip up. It's what drives the (now very occasional) dream of lighting up another cigarette. It's the same kind of fear that, before I quit, led me to put off quitting endlessly, to defer it to some calmer, more peaceful time in my life when I would be more ready.

When I'm most comfortable is when I recognize just how painless and easy it truly is to not smoke. From Allen Carr's Easyway:

The main reason that smokers find it difficult to quit is that they believe that they are giving up a genuine pleasure or crutch. It is absolutely essential to understand that there is nothing to ‘give up’.

Quitting smoking, after a few days or so, can be either easy or hard. You can choose to buck and thrash against the idea of not smoking, or you can accept it as normal. Pretending that cigarettes are like an albatross circling your boat, constantly preying on you, about to inflict mental terror on you every time you finish a meal or have a drink -- that's a good way to stack the deck against yourself.

A question I think is worth considering. Would RJ Reynolds and others, if forced to pick one "anti-smoking" message and amplify it, prefer to have people think that quitting is hard, or easy?

blog post or syllabus

When it comes to startups, very little in the way of "traditional education" is going to help you succeed.

MBAs can be of some help, as can attending a school like Stanford that's already embedded into a local community of entrepreneurs and venture capitalists.

But there is an inextricable divide between school and the startup, a blood barrier, and it rarely gets crossed because there is a misalignment of values.

  • Academia favors the experimental and the theoretical -- either reproducible scientific evidence, or philosophical frameworks that have been socially validated, through peer review.
  • Startups favor learning on the job, persistence over consistence, documenting what you've done rather than doing it according to the documentation.

This last point -- documenting what you've done -- is why the vast library of blog posts on the internet is actually such a strong place to learn about entrepreneurship.

From Paul Graham's site to patio11's longer Hacker News comments, these articles (which blur the line between "content," it should be noted) are the primary method that the startup community uses to scale wisdom and experience and convey it to others. It's no substitute for working at or founding a startup and learning from direct experience, but it is a kind of syllabus. Studied well, it can be a lot more powerful than a traditional education in business.

must love group projects

Mike Loomis gives an eloquent explanation of how burnout happens in a blog post entitled "Why Do Young Superstars Suffer Job Burnout?":

As the workload expands, her days get longer and her performance suffers. She might get some critical feedback. It could be that the young hotshot behind her isn’t getting the opportunity to really contribute, and is unhappy. Her boss lets her know it’s her responsibility to train the new talent. Or, maybe she’s missed some deadlines and her boss is second-guessing her. There’s just too much work to do and she’s not very good at delegating. It’s a quick slide from here...

The most interesting part of this post, to me, is the dichotomy that's drawn between the academic mode and the professional.

In the academic world, you get ahead by doing. Group projects are the bane of any sensible student's existence: other people are unreliable. You get ahead by putting forth your best effort and not allowing the deficiencies of others to impact your performance.

In the professional world, a focus on doing is a great way to get burnt out and make your team to hate you. After school, you have to love group projects to get ahead.


Between the celebrity accounts with a million-plus followers on one side and the 8% of all accounts estimated to be bots on the other, it can seem like there's little hope for upstarts to get any traction on Instagram. 

"Influencers" command fees that would be absurd even if their followings were not all suspect, sketchy, potentially "faked."

In my experience, it's easier, less expensive, more directly profitable and more fun to go after microinfluencers instead.

Microinfluencers have only a few thousand followers. They're more willing to try new things because 1) they're not inundated with spam all of the time, 2) they don't think of themselves as "professional influencers", and 3) they're climbing the ladder just like you.

You always want to surround yourself with people whose goals line up with your own. On social media, microinfluencers are those people -- you both have a small audience, and you both want to grow. Why bother paying $10,000 to an account with a million (possibly bot) followers when you can build up an organic following of people invested in your success?


Those who snidely dismiss adtech as manipulative and pointless miss the point completely. Better advertising technology fundamentally benefits entrepreneurs and startups more than it benefits big corporations:

  • The more personalized and targeted advertising gets, the less you need to pay to get your first few customers or validate your idea. 
  • The more niche you can get with your ads, the less big companies like Proctor & Gamble necessarily dominate the advertising landscape. 
  • The better ads can know me, the less that generic and pointless advertising will work.

It's true that advertising is essentially the process of trying to "manipulate someone into buying something they don't need/want." But why deride those who are trying to make the advertising market more efficient? The more efficient it gets, the more it benefits the little guy.

Massive companies have massive ad budgets (and are selling commodities like razors, shampoo, cars, etc.) and so they are the best positioned to eat the current inefficiencies. 

Startups are the ones that need to make people use/need "something they don't need/want," whether that's an app or a product.

stupid cancellation experiences

People who build products have a wide array of ways that they can build a cancellation flow.

Simply having a button that deletes your account when it's clicked on is pretty rare. At the bare minimum, you usually see something like Github's "type in the name of your repository if you really want to delete it."

At the other end of the spectrum, there's the "call us to cancel" school of thought ordinarily reserved for high-touch SaaS products (and the New York Times). 

That final call is generally supposed to give a retention team an opportunity to talk you out of your terrible decision, give the support/product teams insight into why people are cancelling, and throw some last-ditch coupon offers at you.

There's probably no good rule of thumb for what kinds of cancellation experiences are best. Everything depends on expectations. If I sign up for a $1,000/mo. SaaS product and you let me cancel by just hitting a button, I'm going to be pretty surprised. I'm not necessarily going to be more likely to recommend the product.

If I sign up for a $30/mo. SaaS product and you force me to send you an email, wait for a response, and reply to that email with a "Yes, please cancel" after being offered some kind of discount or opportunity to speak to the support team, then you're going down a dangerous road as far as what people are going to say about your product. Witness: one of the first Google results for "cancel instapage" is this Twitter thread.

Great customer service isn't about always doing a particular kind of cancellation experience, though. There is a time and a place for what Instapage does. The reason why it elicits such derision is the mismatch of expectations. 

When you let me instantly sign up for an account with your SaaS product using my Facebook account, offer me a cheap introductory price, and then don't let me cancel without Googling for a solution and then sending an email to your support team, I feel like I've been lured into a roach motel. Easy to enter; hard to escape.

copycat and educate

Loved this post by Hiten Shah on how some of the best companies innovate by copying.

Amplitude's mobile analytics platform serves as one of his biggest examples. They've been taking Mixpanel's customers for years, not by trying to do anything radically different, but simply by "giving away the same thing that Mixpanel charges $1,000 a month for." Where Mixpanel limited you to 25,000 events per month on the free plan, Amplitude gave you 10,000,000. So for Mixpanel customers who valued the ability to track more and more events had a really great reason to switch.

The other half of the equation, though, is what you do once you've "succeeded" at copycatting someone else's model or undercutting them. 

Mixpanel has (sometime in the last few months) totally revamped their pricing, now giving you 20,000,000 events or data points on their free plan. They're conceding, in a way, that Amplitude was right to look at events as a big value metric for customers.     

If Amplitude had done nothing but copycat Mixpanel, there would be nothing stopping their customers from all going back to Mixpanel now -- where you can now get the most events for free.

That's why Amplitude's diving into behavioral analytics and data accessibility is so important. Where Mixpanel succeeded by showing people that "active" users were more important than "registered" users, Amplitude draws deeper distinctions within the idea of activity both in how they talk about analytics and how they build their product. They're not just selling you a cheaper version of Mixpanel -- they're selling you a qualitatively different kind of outcome from your study of your analytics. They're selling you something that's smarter -- and now that Mixpanel's undercut them on price per event, that's exactly what they need to be selling.