Jason Shah

What I Learned About Failing Fast and Winning in the Startup World

Tomorrow will mark one year from my college graduation. The first year after graduating from college is a uniquely challenging time. You are no longer constantly surrounded by friends. You lose control of your schedule to some employer or the demands of your startup. And chances are that you became a ‘regular person’ – even if you’re in the startup world.

Last year I graduated from Harvard with a degree in sociology and computer science. I had built and sold a company during college. I wanted to do it again – at a much bigger scale and it seemed like there was no better time.

—

Please take a sec to spread this post: Vote on Hacker News and discuss.

—

So I moved out to SF in June, set up shop in a sublet in The Sunset, and started coding a product that I thought would be the company that made me. In the year that has followed, I ended up abandoning that idea, moving onto something else that proved to be entirely unprofitable the more it scaled, learned why you should leave your apartment at least once a week, considered doing consulting on the side, and now am a Product Manager at Yammer.

I didn’t expect to be here 12 months ago. If you’re graduating and thinking about joining or launching a startup, there are some non-obvious things to know. Here they are:

Learn the fundamentals, and don’t act like you know them.

In college, mistakes weren’t cumulative. A paper gets a grade. A semester ends. And you graduate. There are fresh starts all the time. But now the game is different. You will accumulate life debt the more you try to bullshit your way through. The embarrassment you think you will experience is nothing compared to outsized intellectual and experiential debt.

If you don’t know how to sell, learn. Now. If you think you can ship buggy code forever, you’re wrong. You need to realize that there is no more “Oh, I’ll learn that later.” Later is now, so stop procrastinating. This means asking questions and letting yourself be wrong until you finally understand. Even when someone says, “So you get how this works, right?”, do not nod your head unless you really truly get it. Because class is over, and now is the time to learn.

Experience really does matter, but only if you pursue it aggressively and build cumulatively.

I used to think everyone who touted the benefits of working at another startup to gain experience was just bullshitting me. I thought they didn’t have the guts to go do their own thing and were just rationalizing cowardice. Since joining Yammer, my understanding of data-driven product decisions, the ins and outs of an acquisition, and how to build a meritocratic culture makes me laugh at my former self. Sure, I could have figured those things along the way; many entrepreneurs do. But learning these sort of things is not binary: it’s not an “either you get it or you don’t.” thing. This learning is cumulative, and no one ever has life or entrepreneurship nailed down pat. It’s not that simple. Instead, you need some mix of good intuition and useful experience to navigate any given situation in the best manner possible. We get better at exercising these skills with more experience, so ensure that you pursue a wide range of experiences aggressively and build on top of previous experiences. When you don’t build on previous experiences, you end up repeating learning processes.

Time is finite, so experiencing the same things means you’re occupying time that could otherwise be useful for learning new things. Failing to learn cumulatively means wasting time. Wasted time means time not spent growing. You need to grow to compete well. Remember to reflect and connect the dots every few months.

Your relationships will determine whether you see some doors open or not, and networks aren’t built overnight.

When you are working somewhere, it becomes easy to forget the rest of the world. When you’re doing your own startup, it seems like any time that is not building or selling is a waste. But you never know where that random connection at a Hackers and Founders lunch may lead. And you never know which other engineer sitting across the room that you might end up starting a company with eight years from now when the Valley is back on the rise yet again.

But then again, you don’t want to waste your time at the wrong events. You will never be perfect at guessing what you should or shouldn’t go to. I would say to do your best to investigate: look at the speakers, look at the guestlist, look at the sponsors. If an event is only sponsored by law firms, run. They are there to develop relationships; you don’t get as much out of them. If it is a panel, check the moderator out on Twitter. Do they ask insightful questions? And what is the title of the event? This actually is half the battle. If it is “Finding a Co-Founder 101” - run. You can get this from blog posts in half the time. If you’re going to hear entrepreneurs speak, to try to get a sense for whether they will be sharing unique personal stories vs. giving condescending boiler-plate advice.

Keep in touch with your college friends, go to events that aren’t bullshit, and keep meeting interesting people.

Do not let your age help you rationalize a lack of responsibility. Make impact always.

You’re an adult. You have been for years, so act like one. If you start saying, “Oh, I’m the youngest guy here. I’ll pay my dues for a couple years, and then they’ll let me be in the driver’s seat.” Wrong. Think like that, and you’ll always be in the backseat. Don’t settle for pushing widgets around or some ridiculous investor terms because you’re “an inexperienced founder.” Listen to what your investors are saying are weaknesses or what your company’s problems are, and solve them. Logic will trump seniority if you’re good at what you do, and conduct yourself respectfully.

Your time is the most abundant that it has ever been. But it’s also the most expensive.

When you’re 21, it feels like you have all the time in the world. I remember thinking “Oh, I have 3 shots at this before I’ll probably be married and can’t roll the dice in the same ways I can now.” Time feels abundant when you’re 20 something. No matter how busy you are, a vast and potentially awesome future lies ahead. But that time is more expensive with every day that passes. You don’t know what will come up that will stop you from starting a company next year. You don’t know whether equity at a startup or perks at a place like Google will make you never want to leave. So treat your time to build a company like the precious asset it is, because starting a company may not always be an option.

—

Joining a company and building a company are vastly different. But certain tenets hold true when you’re young and excited to make a mark on the world regardless of where you end up in the startup world. So fail fast and learn with an open mind. It’s a long game. Don’t expect to win after the first inning, but play as if you can.

Thanks to Dan Shipper, Chris McCann, Brandon Liu, Peter Boyce, Vinay Trivedi, Ben Lang, and Jason Freedman for reading drafts of this post and sharing insightful feedback.

There is no easy way to know what to charge for. Possible problems with charging:

- No one values your service (enough to pay for it)

- Charging violates your users’ trust (or so you think) 

- Once you set a price, you can’t raise it (much)

- Charging will reduce the rate at which new users join, even w/ free trial

—

Upvote on Hacker News here or just discuss.

—

These are all valid. But I have seen some successful attempts at charging. It seems like identifying the right feature set to charge for is make-or-break.

Take Instapaper. The service helps you save and access content to consume when it’s convenient for you. Imagine you come across a fascinating article in The Atlantic and don’t want to forget to read it. You could email yourself the link or tweet it out, but saving it in your own personal Digg type of account is more natural.

Instapaper could charge for saving a certain number of items. It could charge for saving certain types of media. It could charge for exporting or sharing your links.

But instead, Marco, the creator of Instapaper, charges $4.99 for the iPhone app and variable amounts for a range of subscriptions. Of course.

Don’t limit use until you have some (sticky) critical mass. Capping the number of links would just make me not wanting to use the app for fear of hitting my limit.  

Charge for luxuries — not the things that make your app functional. Technically I can still use the app. I can open up the links on the web and read. But it would be a luxury to have this content at my fingertips when I’m on a train or something. 

Leverage the recurring subscription payment model. Nowadays, you don’t need to just charge $0.99 for your app. Subscription pricing allows you to charge a low fee for the app, but continue getting users to pay low but substantial (over time) fees that apply a meaningful multiplier to your lifetime value.

-

What are some other great examples of choosing the right angle for charging your users?

We are in what feels like an unprecedented age of innovation. Many companies being formed. Many products being built.

You cannot possibly know about every startup out there working on what you’re working on. You cannot know every big company’s product roadmap.

2 guys in Palo Alto, or better yet, New Jersey, could be building a product that eats your lunch. Facebook could release your “company” as a new feature.

How do you find out what companies are your competition?

• Read tech blogs obsessively, so you will know any competitor that has launched publicly
• Network heavily, so you’ll have your ear to the ground on potential upstarts
• Google the hell out of keywords that could lead to your competition

But none of these tactics contribute to product development. Therefore, researching competition only stalls building product — the one weapon you have against the very competition you were trying to stave off.

-

People think building a product for which there are no competitors means it will be easier to sell. Because you won’t have to steal business away from that competitor, or be the 2nd or 3rd pitch to that prospect. But it’s much harder to convince someone to pay for something that they never paid for in the first place.

And consumer applications competing for users who aren’t paying? You, too, are much better off getting consumers off of a platform than making someone begin a behavior that they never did before. 

Make no mistake about it. Google took eyeballs away from Lycos, AltaVista, and Yahoo. Facebook got people who already had Friendster and MySpace accounts. Mint came after Wesabe and competed against Quickbooks (the differences may seem more apparent now with the benefit of hindsight, but they were competitors, no doubt). And Instagram took users away from using the native iPhone application. 

So do yourself a favor: stop doing “competitive analysis” that feeds into a confirmation bias and a blackhole of reasons not to compete.

Just build the best product anyone’s ever seen.

—

Discuss and vote on “Startup Hopefuls: Competition Can’t Matter” on Hacker News.

—

Why? I don’t know for sure. I do know for sure that I have seen a flood of posts today, thanking moms around the world and declaring one’s own love for their very own mother.

Some interesting questions and observations:

• What proportion of people saying Happy Mother’s Day on FB are actually friends with their mothers on Facebook?
• When in history have people said HMD to their friends/public INSTEAD of their own mom (they do that, too, obviously)?
• When before in history were average people broadcasters of sorts, talking about all the moms in the world and the goodness they represent?
• Will a Facebook post and an @mention of your mom (if you’re FB friends) eventually cannibalize some greeting card sales? I’m not sure if this has happened much for birthdays. I do think people send cards less to ‘fringe’ friends now instead opting for FB/email/text.

Some guesses at the why —

• It’s only natural. We say everything else on Facebook. 
• “Richer” expression. Now we can use the Instagram/Facebook photos in our posts, and tailor just what we want to say in the medium of expression we use most frequently.
• Bigger audiences make Mom blush even more (if she sees it). Saying it on FB means a much larger audience, so if our mothers see it there, then it’s as though we have stood up before thousands of people and said “I love you.” instead of just scribbling it onto a nondescript holiday card.
• We want social confirmation. Never before could people Like and Reply to our Mother’s Day words and cards. If people Like my Mother’s Day statement, that’s nicely reinforcing (and maybe my mom would see all the people who enjoyed this declaration of love). If people reply to it, now we can talk publicly about how much we love our moms. I didn’t think this meant that much initially. But on second thought, if your mom hears you ‘naturally’ saying how awesome she is rather than hearing it in a contrived delivery or card, that actually DOES say a lot.
• Historical record. Now cards won’t get filed away in a drawer alongside stamps, mail openers (anyone else?), and mom’s address book. They will be living breathing things on here Timeline that you can reference any time in the future and she’ll be able to pull up quickly whenever she misses you.

—

How are you saying Happy Mother’s Day today?

I noticed something subtle about Foursquare’s language when you share a check in. 

I was not with 4 others. In fact, when they told me (in app) that 4 others were at Philz at the time, I clicked through and those were really just people that had checked in within the past few hours (and were no longer really there).

By…

1. extending the period of time that someone is considered “there” and checked in somewhere

2. broadcasting a check in as if I was “with” others in the sense that I was actually hanging out with those people (which I wasn’t; I was by myself when I checked in)

Foursquare made itself seem even more popular than it is, to me the user AND to the people to whom this update was broadcast.

If the time period to be “checked in” were only, say, 30 minutes, then I would not have seen the exciting “6 people are here” facepile.

Instead it would have looked more like this, if only 1 person were there that recently. 

Blank squares? Sad. Would have been even sadder if the element were hidden and there were no faces, as it is when there are no recent checkins at all.

—

Also, to people reading my update, it seems like I’m super social. So most Foursquare updates make it look like people are hanging out with others, and there’s some element of social proof. If it didn’t say “w/ 4 others”, it wouldn’t be a huge psychological loss. But seeing “w/ 4 others” helps convince others that this app is in heavy use and is fair to use in a real-life social context with other people IMHO.

A lot of applications have downtime or otherwise fail at times. Twitter is certainly one of them, including the notorious Fail Whale.

But tonight I noticed them owning things.

I tried to share a tweet. It didn’t go through. Some people would show some variation of one of the below errors.

Instead, this is what Twitter did:

“Sorry! We did something wrong. Try sending your Tweet again in a minute.”    X

1. Apologize. Right away. Unequivocally. Cheerfully.

It’s the first thing said. They say it with an exclamation mark. Done.

2. “Mea Culpa” - It’s your fault. Don’t blame me (your users).

They clearly say: We did something wrong. They don’t twist words, giving the passive tense an even worse bad name with something like “Something went wrong.” or “There was an error (with something you did)”.

3. Clear instructions on what to do next.

They don’t waste time in saying what you should do next. Wait a minute and give it a go. They also underpromise and overdeliver because it almost always is working right after this error happens; not a minute or more, but just a sec.

Nice job, Twitter. Now just keep the fail whale in the water and off my screen, and we’re good to go. And let me Quote Tweets from the web and not just mobile, too.

An OS that remembers wireless networks, even when you don’t expect it to (MiFi connections)

Google Chrome relaunching previously open tabs when you (accidentally) quit the application

iOS letting you double tap the home button to view currently open applications

Facebook navigating you where you last interacted when you click away to a story and then return to the feed

—

These aren’t the ones I am most thankful for, or the most innovative things. Just things that came to mind when I was pleasantly surprised that my MiFi network was remembered and my wireless connectivity “just worked”.

I am sure you have “better” ones. What are they?

What’s SoundCloud?

SoundCloud is a service for independent artists to share their music and build a following. I think it’s broader than that, but this is how I see it today.

They just released a revamped product and have a custom landing page for introducing it: http://next.soundcloud.com/. 

(Upvote on HN if you like this post http://news.ycombinator.com/item?id=3955306)

The UX Lesson

Here’s the default view.

Notice the Try the Next SoundCloud orange call-to-action button. They want you to click it.

So once you scroll past their primary “Try the Next SoundCloud” button, they place a new one on the fixed header.

Zoomed In -

This keeps the call-to-action (CTA) in view no matter where you are on the page. It also hits users who have expressed interest (measured by scrolling) with a UI change that will catch their attention. This may push them to actually click (rather than getting sucked into your *hopefully* awesome landing page content). Nice job, SoundCloud.

Awesome content, by the way, can be a double edged sword. Write too well, without regard for UI changes to push conversion, and you’ll see good writing distract from your primary desired action.

We see this sort of persistent call to action in many places, such as blogs that want you to share their content. Don’t forget to harness it for your product (without being annoying).

—

What are some other apps that adapt their UI once you have expressed some interest? Others that ensure that there is a persistent call to action?

iHeartRadio (IHR) is an application that lets you listen to any radio station from the web, mobile, etc. In a world now owned by Spotify, Pandora, and a number of other leading music applications, IHR is trying to win on its exclusive radio content — straight from Clear Channel Radio stations (Clear Channel owns IHR). 

—

If you like this post, please share your vote on Hacker News with just one click here.

—

The UX has to be awesome for . this to win. It has commercials that are longer and less tailored than competitors like Pandora and Spotify. There’s a different type of control for users: control stations and not specific songsAnd on mobile, how do you make an experience that can overcome the crowded places in which people need to operate and listen to this application?

As the post title says, the focus here should be on expediting listening. That’s the purpose people seek out when they open a music app. So let’s see how IHR does it.

—

1. The most desired action should be one tap away.

Radio stations last listened. Only using a thumb to navigate. Every extra thumbstroke is like 4-5 keystrokes (pure speculation; there is no math behind this but me trying to count my own keystrokes and guess what is most similar to a thumbstroke)

Notice how the app remembers what I last listened to, so it’s easy to jump right in. It’s important to consider the use case: I may be at work, or walking, or riding the bus. Either way, I can’t or won’t fiddle with a lot of controls, since I’ll only have 1 or 2 fingers with which to work, and my accuracy is limited. Moreover, this (almost) mirrors any other radio: you turn it on, and the music starts playing where the station you last left off on. Mobile apps shouldn’t start right away since you probably are multi-tasking and may have audio elsewhere; a graceful, deliberate transition fits user expectations better.

2. UI elements that vastly improve web experiences can fail to look professional on mobile. Web design ≠ Mobile Design.

Don’t blindly port web design tactics (e.g. rounded corners) to mobile. On mobile devices, rounded corners merely reduce the amount of surface area available for a finger to tap.

Notice the IHR bottom navigation. It violates current standards where the bottom navigation looks full and occupies the entire screen width.

By comparison, here is the bottom navigation for Spotify. Which do you like better?

Remember on a laptop a cursor can pinpoint to a far narrower set of pixels and there is 3X or more screen real estate. Rounded corners in that sit-down open-up-your-laptop environment merely provides “shiny” styling that doesn’t harm usability. On mobile, however, it does hurt usability (and since it’s unusual, it looks weird, too). 

3. Only animate elements larger than an average thumbprint (so people actually see the animation!)

Hover/tap animation doesn’t mean anything if the user can’t see it. IHR turns each navigation option blue when tapped. But tapping means the icon is covered by the user’s finger, so why waste engineering resources on something that will barely be seen?

This is how IHR looks when you tap a bottom navigation item.

But in reality, it looks like this (because your finger covers the button).

The user will see, if anything, a seemingly partial color change. Savvy ones will know what’s going on. Not so savvy users will just think it looks bad. Even if you don’t think it looks bad, why spend limited engineering resources on making something no one will see?

Animate elements larger than the area of a user’s thumb to ensure the response is actually visible.

4. Navigation options should put any conceivable action within reach.

Notice that IHR provides navigation options that enable search, content creation, and what I consider “rescue” (and Home of course). Search is a catch-all for any content I may want to discover. Create lets me get more entrenched with the product and tailor my future experience. My Stations gives you quick access to avoid scanning through a number of stations. Can you think of anything else you would want to do?

Not really.

When you select which 3-5 icons will go at the bottom of your page, select items that will cover your bases, even if your time was largely spent building 2-3 other features that fall under the domain of just one of those buttons. Think in terms of what your users want to do, not what you want everyone to know you worked on.

5. If you have a long submenu with many options, and a couple of other submenus, above the fold you show the short submenus and only part of the long submenu. This way the user knows what remains below. It’s a cue.

The longest sub menu is “Music and Entertainment”. If that were at the top, people would never know to browse by Local or Cities. Nor would they discover Talk Radio. But by putting the smaller sub menus first, and then placing at least the beginning of the Music and Entertainment sub menu, a user at least understands the scope of what is available from one glance.

Also notice that even though IHR is a new app, they use very accepted radio language, e.g. “80s and 90s Hits”, “Alternative”, etc. It would be easy for them to slip into wanting to remake all of music and start calling genres custom names or playlist names. But they want to leverage user familiarity and really just make a new radio platform. Choose your language carefully to bridge the gap from old to new.

—

So what do you think? How else can IHR be better? What do you love about other music applications?

—