Table of Contents:
- Be ruthless about killing your own bad ideas
- Separate our winning instincts from our losing ideas
- Your audience as your barometer
- Don’t let killing an idea kill your winning instinct
- Inside Zynga's guerilla testing methods
- Data: The antidote to thinking you’re right when you’re not
- Micromanage the things that really matter
How to kill your bad ideas
REID HOFFMAN: “Good morning! Hello! How do you do? Top of the mornin’ to you.”
You might be wondering where I am, listeners. Well, I’m in New York City, a place you’ve probably heard of. Maybe the more important question is when I am. I’m in 1890. December 1890 to be specific.
I’m right here on the corner of 57th Street and 6th Avenue. Festive candles flicker in store windows and gas lamps light the streets. It’s Christmas time, and chestnuts are roasting on street carts. I can smell them – along with the horse manure and the factory smoke.
You might wonder what I’m doing here. Well, there’s a story unfolding here that I wanted to tell you about: just a few blocks away, a well-to-do New Yorker — we’ll call him “John Smith” — is flying down the steps of his townhouse on 57th street, and pushing his way through a throng of carol singers to flag down a horse-drawn hansom cab. He jumps in, telling the driver to head downtown. He’s desperate to get his 4-year-old daughter the latest must-have toy for Christmas.
It’s the “Tickle Me Elmo” of its day: the world’s first talking doll, which recites nursery rhymes to lull your child to sleep. It’s made possible by a revolutionary new audio technology developed by the greatest entrepreneur of the age, Thomas Edison.
John speeds past 42nd street, where the Chrysler Building will pop up about 30 years later. His cab clatters past the mansions of Fifth Avenue, the horses snorting steam into the chilly air. Finally, he arrives at Schwarz Toy Bazaar in Union Square. He throws the driver his fare and runs into the store — just in time to snatch one of the last talking dolls from the emptying shelves.
Triumphant, he returns home, creeps upstairs, and tenderly places the new toy next to his sleeping child. He turns the crank on the doll’s back, eager to hear its sweet lullaby. But, instead:
DOLL: Now I lay me down to sleep / I pray the Lord my soul to keep / If I should die before I wake / I pray the Lord my soul to take. Amen.
HOFFMAN: Not surprisingly, the dolls were a flop. Edison pulled production after just six months. But while he killed this particular idea, Edison persisted in his belief that sound recording would be big. Just so long as it wasn’t delivered via screeching terror dolls.
Edison went on to pioneer the phonograph, which paved the way for an entire new industry of home audio entertainment. And every entrepreneur and inventor would be wise to follow his lead.
I believe you have to be relentless about pursuing a big opportunity — and ruthless about killing your own bad ideas along the way.
Be ruthless about killing your own bad ideas
But experience will teach you that the true value of any individual idea lies in how close it gets you to your ultimate goal. If it doesn’t move the needle, you have to ax it quickly, and find a better idea that propels you forward. And for that, you need a certain kind of mindset. One that sounds like this:
MARK PINCUS: I’ll try anything and I’ll kill anything and I’ll kill it quickly.
HOFFMAN: That’s the voice of one of the most ruthless killers in Silicon Valley. He’s not afraid to kill his own ideas — or yours. And he does it incredibly quickly as he constantly strives to grasp the opportunities he sees. Unlike real serial killers, as he’s only a killer of ideas, he’s not a household name. But you’ll likely recognize his work. Here’s a clue.
To witness Mark’s genesis as an idea killer, we need to go back to 1996. Mark had recently sold Freeloader, the company he co-founded with Sunil Paul. Now Mark and Sunil are outside Tower Records …
PINCUS: We would stand outside Tower Records. Remember Tower Records?
HOFFMAN: … doing market research for a new idea.
PINCUS: It was kind of like AOL meets a computer meets the Internet.
HOFFMAN: AOL … Tower Records. This is sounding like quite the nostalgia trip. But Mark’s sights were set firmly on the future. He wanted to give consumers an all-in-one device that would get them online with zero fuss.
PINCUS: I kept thinking the internet’s too hard for consumers.
HOFFMAN: This instinct had given rise to the idea Mark was now pitching to passersby outside that Tower Records. Fast, frictionless internet access. Via a free computer. Who could say no? As it turns out, everyone.
PINCUS: What we learned was nobody wanted a free PC.
HOFFMAN: You heard that right. People were flat-out turning down a free computer.
PINCUS: They basically didn’t trust you. They were like, “I don’t get why you’d give me a free PC.” They didn’t trust you.
HOFFMAN: Mark convinced people he was no scam-artist, only to uncover a more basic issue that was putting them off getting a new computer at all.
PINCUS: The insight we had from that was that the number one thing stopping people from ever going and getting a new PC was this fear of moving their software and the idea of having to reinstall their kids’ games and everything else. I thought, huh, that’s a solvable problem.
HOFFMAN: So Mark did something that has become the defining habit of his career. He realized an idea wasn’t working – in this case, the free all-in-one internet PC. And he killed it. Stone. Cold. Dead. But he remained true to his gut feeling that there was a big opportunity to give users a frictionless experience.
So he moved on to a new idea – creating a piece of software called Move It, which helped people switch seamlessly to a new PC.
PINCUS: Ultimately that led to my technical, brilliant co-founders building this core technology that became the basis for Support.com.
Separate our winning instincts from our losing ideas
HOFFMAN: Support.com became a pioneer in tech support and cloud services. It’s still going strong today. However, ahead of Mark lay a tough lesson that solidified his belief that you need to be ruthless when it comes to cutting losing ideas. It happened at the startup he launched after Support.com — the social network called Tribe.
PINCUS: Tribe is an easy chapter. Painful, but easy chapter.
HOFFMAN: Tribe was launched in 2003. It focused on bringing groups – or tribes – of people together around specific interests. These could range from knitting to heavy metal.
HOFFMAN: While initially it wasn’t aimed at any particular section of society, it became popular with certain sub-cultures. The most famous of these were fans of Burning Man, the annual gathering in Nevada’s Black Rock desert.
But Tribe failed to make itself into something that would appeal to a more general audience. You know, one that doesn’t necessarily dig getting freaky in the desert.
PINCUS: My girlfriend at the time was completely turned off by Tribe and when she was there, got lots of unsolicited messages and interest and it freaked her out and she said, “This isn’t for me.” Sometimes you can get a focus group of one that’s more powerful than the rest of your data, and that one was probably it.
HOFFMAN: But Mark ignored that minimalist focus group. And as he was unwilling to retool Tribe for mainstream appeal, the network’s fate was sealed.
PINCUS: The lesson from Tribe that came resoundingly out for me and still stands out: that as entrepreneurs, part of the journey that we’re on is learning how to separate our winning instincts from our losing ideas. I think as a rule of thumb if you’re a good entrepreneur you can assume that your instincts are right 95% percent at a time and your ideas might be right 25% percent at the time.
HOFFMAN: I’m not as certain as Mark when it comes to fixing percentages on things as hard to pin down as instincts and ideas. But I do believe that being able to recognize a winning instinct is an essential part of being an entrepreneur. And I certainly agree with Mark that you can expect to see a very high mortality rate among your ideas, even if the instinct behind them is right.
Your audience as your barometer
There’s a group of people outside of Silicon Valley who live this credo every single time they go to work. And they do it for a reward far greater than a billion dollar market cap. Their place of work: the comedy club. Their reward: laughter.
We spoke to New York-based comic Matt Ruby, the incisive mind behind Vooza. Vooza is a video comic strip that pokes fun at the startup world. It’s been called the Spinal Tap of startups. This pitch from one of their videos sounds true-to-life.
VOOZA SPEAKER 1: When we pitch Vooza we’re telling a story about our business.
SPEAKER 2: And that’s what you need to remember, that pitching is 10% success and the other 90% is failure. And that’s what I remind myself all the time: it’s ok to fail. I am a failure.
SPEAKER 1: I mean, I tell him all the time what a failure he is.
SPEAKER 2: And I appreciate that, thank you. You’re also a failure.
SPEAKER 1: Thank you.
SPEAKER 2: You’re welcome.
HOFFMAN: We started by asking Matt how he feels when he’s certain there’s something to a joke idea, but it’s just not working out.
MATT RUBY: I think the audience is wrong, but I’m also wrong because I haven’t delivered it in a way that’s funny or that gets laughs. So, they’re wrong and I’m not good enough and that, to me, is a challenge to get better and make them realize they’re wrong. But this is also problematic and not helpful for me in most areas of my life.
HOFFMAN: To avoid this pitfall, Matt is constantly trying out new ideas, using his audience as his barometer.
RUBY: Something might not get a laugh, but you can feel in the room that you’ve hit on something. You have an idea that might be captivating to people or provocative or has them on the edge of their seat waiting to see where you’re going with it. And then, even if you didn’t take them all the way to a punchline that delivered a laugh, you might be like, “Oh, there’s something here.”
You kind of turn into a dog with a bone of like, “Okay I know there’s something to this idea, maybe I just need to go in a different direction or maybe there’s a different punchline or example or metaphor I could use to really get the idea across.” Eventually you might hit on something and be like, “Yeah, it was worth the effort.” And other times you might be like, “Okay this is just a waste of time. I gotta move on.”
Don’t let killing an idea kill your winning instinct
HOFFMAN: In the case of Tribe, Mark was pursuing three instincts that, separately, would indeed turn out to be spot on. His instinct that real-name social networking would be huge was borne out by Facebook; his belief in the power of smaller sub-community forums powered by reputation found form in Reddit; while creating a better way to find jobs and showcase your professional talents… Why, that’s LinkedIn. But his idea to pursue all of these instincts with one product was wrong.
PINCUS: I persisted with one losing idea, which was: what if we mashed all these things together? And I got a bunch of the components of that wrong and I stubbornly persisted with it. I enabled anybody to connect with anybody and what I got wrong was that mass market people did not feel comfortable sharing everything with strangers.
HOFFMAN: Mark needed to kill the extreme openness of Tribe if he wanted to attract more mainstream users and take advantage of the big opportunity. But he could not bring himself to deliver the killing blow.
PINCUS: The idea of Tribe wasn’t right. And rather than failing fast so that you get a lot of shots on goal, we stoically, stubbornly, heroically stuck to our one idea the entire time.
HOFFMAN: It was this bitter experience with Tribe that reawakened the killer in Mark.
PINCUS: Out of that, for me, came this mentality that I brought into Zynga of “I’m not wedded to any idea whether it’s mine, yours, someone else’s.” I’ll try anything and I’ll kill anything and I’ll kill it quickly. And I’m not going to let killing an idea, kill a winning instinct.
HOFFMAN: His next company would bring together everything he had learned so far. But his methods would be contentious. Because Mark would remain unflinchingly true to his new guiding principle — if an idea didn’t work, it would be sent straight to the chopping block. It was the only way to stay relentlessly open to pursuing big opportunities.
Inside Zynga’s guerilla testing methods
PINCUS: I kept thinking, “Wow, these people are kind of hanging out with nothing to do.” And one of my early instincts on social networks was this concept that people just wanted to hang out there. It was like a bar or a club and people just wanted to be there and they wanted more to do there.
HOFFMAN: There it is – another one of those nagging instincts that leads to a big opportunity. This time, it’s that people wanted more things to do together on social media. And here comes the idea.
PINCUS: And I thought games were kind of this, lost love affair that we had growing up. Whether it was board games or video games. We couldn’t play them as adults because they required too much time and they required us to be in the same room with your friends and family around a Scrabble board.
The leap of faith that I think I connected to at an instinctive level was, I believed that there was a latent demand in all of us to play games. But there was just way too much friction around it.
HOFFMAN: Like the internet back in 1996, gaming had too much friction for the average consumer. You needed to know how to download the games. Maybe your computer needed a graphics card. Or more memory. Or the right operating system.
PINCUS: The question was, to me, not whether or not people want to play games, but is there a way to deliver games to people that is so low friction and easy that they trip over it?
HOFFMAN: Mark’s instinct – that there was a huge untapped market of people who would play games on social media, was the right idea at the right time. Facebook had just started to let third-party developers build apps and games that people could run within their Facebook accounts.
Zynga could use this ready-made delivery system to mainline games into a previously untapped audience. Soon there were millions of people playing and sharing Zynga games. It was like an exciting, mysterious carnival had just arrived in town, and there was a stampede of people eager to get into the action.
SPEAKER: Please step right up, step right up one and all for the amazing, sensational traveling Zynga arcade. Enter now to be astounded by our many attractions, guaranteed to enthrall and delight even the most somber of sourpusses!
HOFFMAN: Boom. Zynga went stratospheric. It also left the rest of the computer games industry scrabbling in the dust. Or maybe should that be Words-With-Friending in the dust?
PINCUS: At the time, game companies would build an entire game and take even two years and $15, $20 million dollars before they even knew if anyone wanted it. And we could come out on the Web, maybe only spend four weeks.
HOFFMAN: Mark turned Zynga into not just a game company but an idea killing machine. For every hit like FarmVille, Zynga Poker, and Words With Friends, there were countless others that didn’t make the cut. Sadly, we can only imagine what some of these titles may have been.
SPEAKER: Please step right up for… Laundromat Baron…? Experience the thrill of helping your friends clean their underwear. Er. Oh. Hm. Yo-yoing with Cats! Anyone? Hello? Hello?
HOFFMAN: Actually, I can think of a few people who’d enjoy Yo-yoing With Cats.
Another advantage Zynga had over traditional game developers was that it could constantly improve games that it had already released. Mark had an ingenious way of working out which new features to pursue that I like to call guerilla testing – because it is sneaky, highly targeted, and very effective.
PINCUS: We would, maybe in our poker game, put up five words promoting a new feature or idea for poker or a new game and if people clicked on it we’d say, “Awesome, this is coming soon, let us know if you want us to let you know.” And I generally said, “If half of your audience clicks on it, you pretty much know it’s an idea worth pursuing.”
HOFFMAN: Mark’s teams would guerilla test hundreds of ideas every day. They could be small, like changing the color of a menu from green to blue. Or huge, like the selling of virtual in-game items and perks, which would go on to form the basis of the company’s revenue.
Those ideas that made it through this initial stage were then fleshed out into actual features. They would become what Mark calls “Bold Beats”, something that got users really excited.
For an example of a “Bold Beat”, we’re going to look at a game that has its roots in one of Mark’s deepest fantasies… broccoli.
PINCUS: I had a farm and ranch fantasy and I still do, and I pictured Pincus Valley Ranch and we would create our own broccoli and it’d be super organic and served by name by restaurants and I thought, “Okay, I don’t have the time or space to go and actually be a farmer, but it would be a fun hobby or fantasy to create a farm.”
HOFFMAN: The idea was FarmVille, one of Zynga’s biggest hits. In it, players planted, nurtured, and harvested crops on a small patch of land, which they could decorate with animals, hay bales, and fences. Everyone could have their own version of Pincus Valley Ranch, broccoli not included. The “Bold Beat” came about when an engineer made a seemingly tiny change…
PINCUS: An engineer, one day, just hacked on the animals and made a cow move, and we put it out, didn’t really know if players would notice, what they would think about it, and it lit up Twitter and the blogosphere.
HOFFMAN: Animals that move. It’s a teeny idea that took just a few hours to code. But the excitement that cow caused led to a whole range of new features. Animal breeding, horse racing, chicken coops with hatching eggs.
PINCUS: It really was one of the perfect, classic examples for us of a successful “Bold Beat.” And if a “Bold Beat” is right, it could lead to inspiring your players to really imagine a new dimension that that game play can go to and inspire you to play for another three months or another year.
HOFFMAN: Mark had shown his teams how to quickly work out which ideas to focus on, and which ones to terminate. The beauty of this is twofold. One: the idea lives or dies by the customer reaction, which you can see immediately. And two: the scope of a new idea is limited so it won’t break the product. You’re being bold by constantly trying out new ideas, while risking very little.
PINCUS: And we used to say when you delivered a “Bold Beat” or a great feature that your players loved, it was like Christmas Day because when we turned it on, your metrics would light up and blog posts or Twitter would light up and it was amazing because you literally could watch your DAU…
HOFFMAN: DAU stands for daily active users, by the way.
PINCUS: …or revenue chart and you’d turn on this feature and it would just jump and then stay there and it was a beautiful new way of doing business.
HOFFMAN: Every product team was geeking out with their guerilla testing, and watching with glee when their metrics lit up. It was Christmas Day. Every day. But there was no standard way to track and share all this hugely valuable data.
PINCUS: They were all doing their own version of these things and nothing was standardized. Some were working better than others and some had kind of lost the thread. And I said, “This isn’t going to work. We’re going to standardize. We’re going to pick one template, you can fight like hell for what you think is right. But then we’re all going to commit to it.” And people’s heads spun around.
HOFFMAN: Mark wanted to scale his rapid-kill approach. To do that, he had to make sure every team could access the treasure trove of data — and interpret it the same way
What they needed was data standardization. It may not sound like the most exciting of subjects. So we found the most exciting person we could to explain why it’s vital.
Andrea Jones-Rooy is an analyst at FiveThirtyEight, the website founded by Nate Silver that brings data to bear on news reporting. Andrea is also a circus fire performer. You heard that right. Circus fire. You may be wondering what that means.
ANDREA JONES-ROOY: It means I light something on fire and then I put it on my body and everyone gets impressed that it doesn’t hurt, even though it does, and then I eat it, which means I put it in my mouth.
HOFFMAN: Ouch. Eating naked flame sounds painful. But not as painful as unstandardized data. As Andrea will tell you….
JONES-ROOY: If you don’t standardize data you don’t have data. You just have a huge mess. The reason we put things in data is so that we can share it with a broader audience.
HOFFMAN: That audience may be your executive team. Or it could be the entire world. But if your numbers aren’t clear, no one will be able to interpret them.
JONES-ROOY: It’s about making units compatible and then making sure that whoever is receiving that information understands how you’re measuring it.
HOFFMAN: But Andrea has been burned in this respect too, despite her credentials as an expert in the field. When she was a professor at New York University’s Shanghai campus, she and her fellow professors tried out an experiment that completely backfired.
JONES-ROOY: When we got to NYU Shanghai we were very excited about pioneering a brand new way of doing education. We came up with, what in retrospect was a terrible idea, of coming up with a new grading system all our own. We abandoned the A through F scale and we assigned a different scale.
The scale itself ended up being a bit ridiculous, which was the best score you could get was “truly outstanding,” then “superior,” then “very good,” then “good,” then “needs improvement.”
The students understandably freaked out because they were like, “What does it mean?” If you get a very good, that sounds good, but if you map it to an A, B, C, D, F scale, that’s a C. Good is a D. You’re just like, “These numbers don’t go together.” It caused all kinds of confusion.
And then in the end it was a huge mess because you can’t have transcripts floating around that say things like “truly outstanding” on them when they’re applying to grad school or trying to get a job because everyone understandably would be like, “What does that mean?” We then had to convert it all back to A through F anyway. It was a huge mistake.
Data: The antidote to thinking you’re right when you’re not
HOFFMAN: This shows just how hard keeping a grasp on data can be. Standards are a slippery subject, even for the experts. But the effort is worth it. Because once you get it right, you have a powerful tool with which to accurately test your instincts and ideas. Which is exactly what Mark wanted for his teams. So he ordered them to crunch their numbers in the same way.
PINCUS: What we got though with the standardization was now the ability to see the same metrics, the same tests and share information faster.
HOFFMAN: They went from having a huge mess to a huge cache of valuable insights. They could analyze what their players were doing down to the last click and understand exactly what the players wanted. Teams could then share their hot new findings with every other team at Zynga.
This was all good in theory, but with dozens of product teams testing hundreds of ideas every day, there was an avalanche of data to deal with. So Mark made an unusual decision in hiring talent for scale. He started hiring data engineers and analysts at an accelerating pace, at a moment when most other game companies would be hiring product managers and game developers. Zynga was betting the farm — actually the Farmville — that all this data would give it a huge advantage over its competitors.
PINCUS: We were pushing it more than anybody because we went to track every click and analyze it at a time when they were using Google Analytics. And we were investing so much and we had so many people on it that we kept getting called stupid, that people said, “Zynga has 50 people and this company is doing the same thing with 10. Zynga has 300 people and this company is doing the same with 20 or 50…” because we wanted to over-invest in knowing the data.
HOFFMAN: They weren’t called stupid for long. This investment in data geeks meant Zynga could work out exactly what features got their customers excited, and immediately act on these findings. All at a speed that traditional game companies could barely comprehend.
PINCUS: One of the most gratifying things as a founder and entrepreneur is to see people go after an idea that you would never have thought of in a way that delivers on your vision.
HOFFMAN: Any entrepreneur, no matter how smart, is only right part of the time. The antidote to thinking you’re right when you’re not is data. This data may come from standing on the street asking random people about an idea you’ve had — like Mark outside Tower records — or it may come from hiring hundreds of data analysts.
As Zynga’s user numbers soared, Mark never lost sight of how important it is to be willing to kill ideas rapidly. But he was finding it challenging to maintain this winning approach while scaling.
Micromanage the things that really matter
So what was Mark’s solution to this? How did he find a way that allowed him to scale rapidly while maintaining that vital ability to test, analyze, and kill the ideas at a tremendous pace? True to form, it was controversial. So controversial, for the first time in the history of Masters of Scale we’re going to have to bleep out a word. I’d better test my bleeper.
DOLL: Now I lay me down…
HOFFMAN: Okay. All working. Here goes. Roll the tape.
PINCUS: One of my biggest lessons about scale was <beep> f*ck <beep> scale.
HOFFMAN: Excuse me?
PINCUS: (replay) <beep> F*ck <beep> scale.
HOFFMAN: That’s what I thought he said. Now, I’m known for my contrarian theories when it comes to scale. But you might think that giving the scale the middle finger is a bit far, even for me. But hold on. Let’s put this in context.
By <beep>f*ck<beep> scale, Mark’s not saying don’t scale your company. He’s saying don’t lose sight of your overarching goal – to delight your customers.
PINCUS: What really matters to your customer is that they have an amazing experience. They don’t give a <beep>sh*t<beep> how it got there. You’re at your favorite hamburger joint. You want a great fresh burger. And if you find out that the owner of the company was back there making it, you’re stoked, you don’t care if he figured out how to scale.
HOFFMAN: Mark wants to be intimately involved in what some people would classify as the “small decisions” that affect user experience. Because Mark knows that it is these small decisions that can have the biggest impact on customers. And he wants everyone in his company to care to the same degree.
PINCUS: I want our teams to care about the product experience at the pixel level and I want that to be apparent to our players because especially on mobile, every pixel counts, every nanosecond that we take of your time counts and quality is delivered at the pixel level.
HOFFMAN: Micromanage the things that really matter – the things that are going to excite people and make them love your product. Instill that obsession across your entire company. Make sure that you’re trying ideas and killing ideas in order to delight your customer.
With practice, killing your bad ideas will become second nature to you. And this ruthless killer instinct will clear the path toward your big opportunity.
If you do these things — whether your product is a great burger or a game about building a great burger restaurant empire — I guarantee your customers most certainly will give a <beep>sh*t<beep>.
I’m Reid Hoffman. Thank you for listening.