March 3, 2016
DX and Open Sourcing at Netflix
In this Speaker Series presentation, former cloud architect for Netflix and now Battery Ventures technology fellow Adrian Cockcroft discusse...
Thank you, so great to be here. I'm Andy Raskin, and I work with leadership teams on strategic messaging and positioning, which is another way of saying I basically help them get the story straight. Which can be really valuable for fundraising, scaling, marketing and sales, even building product and recruiting a great team.
I began my career as a developer, and through my own experience raising VC money I became fascinated with the power of the story. I had this idea about launching a consulting practice about five years ago around helping entrepreneurs craft their story. I didn't do it, because I just felt like, what entrepreneur has a line item for storytelling?
But one day I get this call from a CEO of a Series B company, and he says, "Listen, I just spoke to Oren Michels." Oren was the CEO and founder of Mashery where I ran product marketing. Oren says, "You can help me with 'boom, boom, boom, boom, boom,'" and he basically is describing the crafting the story stuff that I wanted to do. Except he didn't call it that. He said, "Could you help us with strategic messaging and positioning?"
I launched my strategic messaging and positioning consulting firm, and that first engagement went really well and generated some really good word of mouth. Over the last couple of years I've been privileged to work for over 40 leadership teams on getting their stories straight. Many of them are funded by folks like Andreessen Horowitz, First Round Capital, True Ventures. In fact, I see more and more of my engagements are actually starting from a referral from a VC, usually in preparation for a round or an acquisition, or in some cases, after the round and where the team is about to scale up and they're looking for a really clear story.
I think part of the reason for that is that there is much more understanding now and appreciation for the value of telling the company's story really well. I think you really can't go anywhere these days without somebody telling you, "You have to improve your storytelling skills as an entrepreneur." But my interest around storytelling, and really the focus of my work, is on structure. Because I think if you get the elements of the story right and you order them correctly, and you have a shared understanding of those among your whole team, then it's not just the CEO who's going to be a great storyteller and telling the right story.
Everybody's going to be telling that story, and it's going to be consistent across all the interactions that people have with you. So let's talk about how to do that effectively. Now, I think most of us approach storytelling this way. We start with some strategy, some way we want to change the world. We build a product to do that, and then the story is some kind of packaging, a kind of wrapping paper that goes around that. And I think this is why a lot of strategic messaging either fails or is kind of uninspiring. It's really just an explanation of the product.
In that context, I think this quote by Ben Horowitz is really profound. If your story is your strategy, then it's about something far more elemental, far more dynamic than just your product. It's about that change that you want to make happen in the world.
As a former engineer, I know this can sometimes be hard to digest because it sounds at first glance like maybe I'm saying "Don't talk about your product." That's not what I'm saying.
What I'm saying is when you do talk about your product, do it within the context of a larger story, a story about the change that you believe in.
That's how you build strategic messaging that helps you raise money, helps you close deals, helps you build a great team.
So, how do you tell that story? Well, the not-so-great news is that, of course, there's really no one-size-fits-all template for telling a great strategic story, and the reason for that, of course, is you have to pitch to so many different kinds of audiences and so many different kinds of formats.
I mean, there's full-on investor pitches, there are customer sales pitches, there's your homepage on your website, that's a pitch. Every time somebody asks you, "What do you guys do?" They're basically inviting you to pitch. So there's no one template that's going to address all of that. But my approach instead is to identify the core elements and then arrange those as needed, as appropriate for the audience, and the format that you're going to have to pitch to.
The first thing I do when I work with a team is I get them in a room and I try to facilitate consensus around these five elements. Number one: name the enemy. Who is our customer, and what is standing in the way of their happiness? I think we'd all agree that the first thing we have to do is make people care.
In Breaking Bad, you know the main character Walter White, he's a man who's a meth dealer, he regularly murders people to gain more power. Yet we find ourselves caring about him, even rooting for him, at least for the first few seasons. Why? My friend Jennine Lanouette, who's a screenwriting scholar, says that we'll pretty much have sympathy for anybody as long as we see them at a power disadvantage.
With Walter White, he's first battling cancer, and then there's this series of ever nastier villains that he's doing battle with. So we root for him because, in some way, we appreciate his struggle, and we empathize with it.
Now, a lot of pitch advice will tell you to start with the problem, and that's not wrong, but better if you can show how the customer is experiencing the problem, if you can show that struggle in emotional terms. About a year ago, Elon Musk gave this keynote for the Tesla Powerwall. It's a battery that stores power collected by solar panels.
He starts his pitch by showing this horrible image of the burning of fossil fuels. If he had just started talking about batteries, probably most people I think don't care that much. But when you see this, you don't just understand that there's a problem; you almost experience why it matters. One company that I think does this really well, this naming the enemy thing, is Dollar Shave Club, and I'm going to show you one of their advertisements.
"Hey, one pack of razors please." "Comes to $19.65." "What? That's a lot." "Yeah. It comes with a free gift." "Oh, free gift? That's cool." "Tired of getting beat up by high razor prices? Dollarshaveclub.com ships amazing razors for just a few bucks."
Okay, now, they use humor to make me aware of something that I think I always felt, but didn't quite consciously acknowledge, that when I go to a drugstore and I'm paying these exorbitantly high prices for razor blades, it's almost like I'm being abused. And so I really do feel, at some level now, that, yes, these drug chains are kind of like my enemy.
Life isn't so great for your customer because there's this enemy, but the enemy has probably been around for quite a long time. So, what's happening now in the world that's raising the stakes, that's making it so that your customer more likely to act now? In every successful story, there's this moment when the hero is prodded into action. So in "The Martian," he's stranded on Mars by a sandstorm. In "Bridesmaids," she's tapped to be maid of honor by her best friend. In "Frozen," her older sister freezes the whole kingdom under a sheet of ice.
In the original "Star Wars," it starts out with Luke bellyaching about how he's really bored and he wants more adventure. So Obi-Wan shows him this holographic video of Princess Leia in distress, and he says, "Hey, Luke, let's go and rescue Princess Leia on the planet Alderaan." And what does Luke say? He basically sounds like a prospect you've been trying to sign who just keeps coming up with more excuses not to.
When Luke finally does buy, when he goes with Obi-Wan to Alderaan and they start going on their adventures, it's only after the Empire destroys Luke's home, kills his aunt and uncle and leaves him with nothing, which is to say George Lucas is showing us not only why Luke might want to act, but why he has to act now.
Similarly, Elon Musk knows that people are going to be skeptical that the world is suddenly going to act to rid itself of fossil fuels. So he shows this chart, this graph, that's called the Keeling Curve. It shows the rise of atmospheric CO2 concentration, and it's been rising steadily for some time. But it's about to hockey stick. It's about to get a lot worse, and he says, "If we don't act now, we're going to see levels that we don't even see in the fossil record."
Here's Steve Jobs introducing the first iMac in 1998. Why now? Because now there's this thing called the Internet, and more and more people, including people who are not necessarily computer experts, want to get on it. You've shown your customer struggling without you and why there's urgency to change. The temptation is then going to be to introduce your product. But if you remember one thing from today, it's that before you do that, instead give a glimpse of what your customer's life will look like if they go with you, if they use your product.
I call this "teasing the Promised Land."
This is especially important for developer-focused companies, because the developers who are using your tools, they may need to get buy-in from their colleagues or other people on their team who don't necessarily understand the Promised Land, can't visualize it just from looking at your tool. So, in a sense, you're giving them a tool to sell internally when you're not there.
With Musk, I think a lesser pitcher might have gone straight to the problems that a battery has to solve like cost and power output. But before Musk goes there, he teases a future in which everything is powered by the sun, and the enemy, fossil fuels, is totally defeated. You can see what he's doing here. By the time he introduces his battery, it's not going to be a battery anymore. It's going to have this other significance as almost a kind of savior of mankind.
Steve Jobs was a master at teasing the Promised Land. You know, most people would show a chart like this after they introduce their new product, but Jobs gets us imagining a world with this super easy to use, super smart phone before he even shows it to us. And then when he finally does show the phone, kind of like Musk's battery, it's already imbued with this richer, higher-level meaning.
Notice, too, that Jobs simultaneously uses this Promised Land as a basis for differentiation. Instead of getting into "We're going to have these features that are better than their features," what he's saying is, "We're going to get you to this Promised Land. If you want that Promised Land, come with us. "If you want a different Promised Land, go with them."
Strategic messaging always works best when it differentiates this way, based on the Promised Land, not based on features.
One last thing I want to say about the Promised Land, and I hope it's obvious by now, is that it's one of the most important strategic decisions that you'll make. Because if story equals strategy, as Ben Horowitz said, then high-level story choices equal high-level strategy choices. So it's really worth devoting some time to defining your Promised Land and honing that definition.
There's a great YouTube video of Kurt Vonnegut sketching the shapes of stories on two axes, and I'm going to borrow those axes here. We have happy and unhappy on the Y, and time on the X. Let's say that we're "fairygodmothers.com," and we help tormented young adults attend balls. Yes, our beta release breaks at midnight. We're working on it.
One possible Promised Land for our customer is we help you attend balls, but if you attend the ball, you might meet and marry royalty. And if you do that, you might eventually die happy. So leaving aside objections about whether that's true or even desirable, which is the right Promised Land here to, say, emphasize on the front page of fairygodmothers.com's website?
Absent any other information, I like "marry royalty," because I feel like "attend balls" is not motivational enough, and "die happy" is a little too vague. However, the right story really depends on your strategy, and we see this in the dating app category. So we see some dating services touting Promised Lands that are kind of close to "die happy," like eHarmony, for instance. And then there are others that are promising a Promised Land that's, we'll say, a little more short-term.
Neither one is right or wrong. They're right for their strategy. In this context, it's been incredibly interesting to watch Uber as its Promised Land has kind of up-leveled in this way. Here we are in 2011; what's the Promised Land? Basically, that you're going to have a chauffeur, a kind of luxurious life. UberX is released; this is 2014. Now what's the Promised Land? It really looks more like it's something about getting around town. She's still kind of glamorous, but the car is a Prius, and it's really about her getting where she needs to go easily.
More recently, Uber just released some new strategic messaging, and again, the Promised Land, I think, has changed. Now it's even a kind of higher-order thing that you're going to have some kind of control over your life.
Okay, so finally, it's time to talk about your product, but by setting it up this way in the way that we can do it in a much more impactful way. Specifically, we can start with this question, "You've defined that great Promised Land, so why can't your customer get there without you?" And "How are you going to help overcome those hurdles?"
There had better be some pretty big hurdles, because if there aren't, why are you even in business? So I'm sure there are.
There's a great lesson from the epic stories, because often there's this wizard character who's kind of doing what you're doing. He's trying to sell the hero on a vision of the future. We saw Obi-Wan doing that earlier, but he also helps the hero along with some special gifts: lightsabers, Jedi mind tricks, spells, that kind of stuff.
By the way, it's not always a bearded old man. In one of my favorite films, "500 Days of Summer," it's a precocious 12-year-old girl who gives her older brother very sage dating advice. But it can help to think of yourself as this helper, this kind of wizard character. I saw that Michael Dearing was here a few months ago, and he talked about Barbara Minto's approach to executive communication.
One of her rules is to make arguments in lists of threes. With Facebook, we see this at the top, the Promised Land. How are they going to get you there, and what are going to be the challenges to getting you there? Well, you're going to need to get updates from your friends. You're going to need a way to broadcast what's new about yourself, and you're going to need a way to discover new people and new friends.
We don't see the obstacles here, because the gifts kind of imply them, and that's very common. I grab these from the website of Rainforest, the Heavybit company, and they're really doing this I think very nicely. I'm sure Rainforest has hundreds of features and capabilities, but they've organized them down into three gifts that they can use to structure all of those things and make sense of all of those things, and those three gifts are all building up to what they have as their Promised Land: deploy faster without breaking things.
By the way, there may be some formats where it's appropriate to go into a lot more detail. I'm not saying that it always has to be this short, but even there, even if you're doing a 10-page white paper, you can still organize the material, organize the capabilities and features in this way.
We've explained the transformation in the world that we are committed to making happen. We've explained how we're going to make that happen, but audiences are skeptical.
What evidence can you offer that the story you're telling is going to come true? And I'd say hit them with the strongest stuff you've got, which will kind of depend on your stage.
At the end of Elon Musk's Powerwall keynote, he muses about how great it would be if keynotes, like the one he's doing in auditoriums where he's presenting, were all powered by the Tesla Powerwall and solar energy that was stored in batteries. And then he says, "Actually, that's exactly what's happening right now," and he has a camera go down to the basement. Then it shows this video image where it's showing that there's actually zero power coming in from the grid to the building, and it's all being supplied by Tesla Powerwall batteries.
This is essentially a product demo, and product demos are one kind of evidence that you can offer. Of course, Steve Jobs was famous for product demos as his evidence after introducing all the other pieces of a story. If you've got customers, talk about those. Evidence.
Your team is the evidence part of the story. If you've got traction and sales, all that stuff is evidence that the story that you're telling has a likelihood of coming true.
This is the core story, and you're going to have to modify it a bit, add to it, subtract from it for different audiences and different formats. For example, how would you modify this, what would you have to add if you were going to pitch to an investor, for instance?
The way I like to think about this is that when you pitch to the investor, you're still pitching the same core story. You're still Gandalf helping Frodo put the ring into the fire. But there are these helper characters, like the elves, who are supporting the journey. But they have interests of their own, and so you have to show these characters what's in it for them. And the classic way to do that, of course, is show them what the Promised Land's going to look like.
When you reach the Promised Land, what's it going to look like for them? This is Airbnb's famous early pitch they did where they just did this one very simple slide that says, "Hey, if we can attract this kind of user base, here's the kind of commissions we're going to get for every lodging. And here's what the revenue's going to look like over four years."
As I said, most of the work I do is in preparation for a big investment round, acquisition, or a large scaling-up of the team. But once you have this framework in place, it can really drive everything: sales decks, call scripts. Product roadmap, in my opinion, should actually be driven by a story like this, and to see an example of that, I'd like to introduce Andrea Echstenkamper, the very talented Head of Marketing at the Heavybit member company LaunchDarkly. Andrea and I know each other from our days back at Terracotta, and she's going to talk about how she used the framework that I talked about in a recent marketing project of hers.
Andrea Echstenkamper: Hi, everyone. I'm Andrea Echstenkamper, Head of Marketing at LaunchDarkly, and I'm going to tell you about how our company narrative came to be. I was looking to make an explainer video, and while we have messaging in place that's working well, what I really wanted to do was create a story end-to-end, and I was having trouble figuring out how to get started.
I know Andy from working with him at Terracotta, and I thought perhaps I could use his framework to try and make this story. I got about three people from my company to sit down in a room and go through, step-by-step, each of these five points. And as you can imagine, there were a lot of ideas, and it was a little bit disjointed. What happened is I worked with Andy, and we kind of culled it down so that there was really these narrative elements standing out, the five of them.
What I'm going to do is kind of go through those five that I came up with. The first one: name the enemy. For developers and product managers, feature release is really too slow for them. It's difficult to de-couple them, and the difficulty there became our enemy.
Why now? So, the reason is customers. You want to get new features to customers as fast as possible, and especially if you're a B2B company, you may have a particular marquee customer that is ready for a feature yesterday. They want to go ahead and see it, and you want to go ahead and release it directly to them. So tease the Promised Land. Here we want the potential customer to really think about how awesome it could be if you really could release as fast as you've dreamed of and you could do that without risk.
The three obstacles and three gifts where you really want the customer to realize how they can't get to the Promised Land without your product, and this is where you really start going into more specific features. Specifically for us, the first one is that it's difficult to get early feedback, and that really informs your release plan, so we give you the mechanism for tying feature releases to goals and slow percentage rollouts.
The second is that it's difficult to control subscription plans and beta plans, and we let you control that with just a few clicks. And then third is that our customers originally managed their feature flags in config files and databases, and at one point, it gets too difficult for them, too many. They can't manage them, and that's when it starts to be a pain.
That's really where our key value proposition comes in, the fact that feature flagging is easy. But feature management is very hard, and we provide an easy-to-use dashboard with all of the enterprise features you need to manage feature flags as a team. I know Andy said three, but I kind of broke the mold here and added a fourth one, which is that we have SDKs for all major platforms as well as mobile, so you can start using it immediately across your entire stack.
Then offer evidence. For us, it was that we have a lot of big companies using us and really depending on us for their feature flag management. You know I get to this point where I have these five things, and really a story arc starts to develop. I start to see how we can put these all together and make it a real story, and I want to thank a designer that helped animate the explainer video that we made, and Edith, my CEO, who did some editing and really supported me through it, and then of course Andy who helped me get to the final product. But just wanted to show you kind of what it all looks like in an actual video.
"Is your feature release process as fast as it could be? Today, customers demand to always have the latest, greatest features of your product, whether they're on web or mobile. However, if exposing a feature means long, waterfall release cycles, there's no way you can move fast enough when rolling out new features. But imagine how much happier you could make your customers if you could separate your feature rollout from your code deployment."
"Today, major dev teams use LaunchDarkly's feature flag as a service platform to manage feature lifecycle independently of code releases, from local development through QA and production. With LaunchDarkly, you can test new features on limited segments before rolling them out, and you can get live user feedback on how a feature is performing."
"Bundle features into a new pricing or access package with just a few clicks. Instead of managing feature flags in config files or databases, LaunchDarkly gives you an easy-to-use, intuitive dashboard for making changes to live features without redeploying, and our enterprise grade platform ensures no latency or down time. LaunchDarkly has SDKs for all major platforms, including mobile, so you can start using it immediately across your entire stack. Is your feature management still tied to your code? Try LaunchDarkly today for free and get control over how you release as a team."
AR: Like as we saw, the evidence can be, especially if you don't have customers yet, a demo, a product demo. It could be sales and traction, customer testimonials or logos, they all support that. What if you don't even have a product to demo? What if you're at that stage? Then not so great, but you could show maybe some sort of indication of demands, testimonials of people who fit the target customer profile who would say, "I would buy that if it existed." Not great, but maybe the best you can do.
Whatever the best you have is the best evidence to put there. If it's a format where the team makes sense, the team is evidence, too.
Yes, this team has been there, done this before, and so we're more likely to believe that they're going to make that story come true.
Yeah, I think it's all really about coming up with a Promised Land that makes sense to them. How are you changing that customer's world in a way that they will care about? If you can start there, then you have a starting point with them where they might even start to say, "Oh, well what's under the hood there?"
Editing, rewriting, that's what all writers do. That's what all filmmakers basically do, that, based on seeing how things play. Like Andrea said, when her team came up with the elements in the story, there were a bunch of ideas. Some people thought maybe the Promised Land should be this. Some people thought it should be that, and so you go with one that you feel is right, and I think you can kind of gauge the impact from people.
I wouldn't quite go as far as A/B testing, but focus group mentality around getting some feedback and refining it.
I really think the answer, again, is the same, which is what is a kind of Promised Land definition that you can use with them that's going to resonate with them, and for developers.
I think we tend to shoot low on the Promised Land. We tend to say, "If we go a little bit too high in the direction of die happy, then we risk looking non-technical, like we don't know what we're talking about."
I actually think that just going a little bit higher than is comfortable works as a framing.
And then you can be as technical as you need to be right after that. Usually, even if you're selling directly to developers, there are going to be other people who are going to be important for your success. They're going to have to understand your story, so it can't just be the technical details normally.
AE: Well, for me, it was selling the Promised Land really in a way that, once we get the story down, we'll find more traction in the company, right? But I think if you want to take two or three people, it doesn't have to be the whole company to get them in a room.
But it's fun, too. I'm a big believer in fun, and it's great having people in the room talking about these things. It gets everyone excited.
AR: I think there's a famous quote, "Marketing is too important to be left to the marketing department," which doesn't mean the marketing department shouldn't do marketing, but just that's it's sort of everyone's job. I think the story is like that in that it really is the job of the CEO, and I think there are some CEOs that, as much as I would say that, or as Bill Gurley says it in the quote I showed, or Ben Horowitz says it, they're just not going to go there.
In all of my engagements, in all the work that I do, the CEO is always in the room when we're creating the story. Because the CEO is the person who is going to have to make sure that that story's consistent, not just in marketing campaigns, but that the product supports that story. That recruiting is telling that story, that sales is telling that same story.
I feel your pain sometimes, because sometimes people don't see it that way. People basically see the story as that kind of wrap around the product. "Hey, here's our product. "Go write the story about it," and I think that usually doesn't work very well. It usually has to start from, "What is the change we want to make in the world?"
One thing that I always do is ask customers, "How has this transformed your life?" And sometimes customers will kind of laugh at you like, "It's only this thing. It didn't transform my life." But then other customers will almost get teary-eyed and tell you about how it transformed their life. Those are the ones where you can really get some good story juice, and I think once you start bringing that stuff to your team, they usually get excited about it.
I like to try to put the Promised Land as the top thing, and I think Rainforest did that very, very well. And then we're supporting that with, sometimes, a descriptor: here's the Promised Land, here's a little lower in this sort of no-nonsense what we are. And then the three gifts that are getting the customer to the Promised Land.
Andrea and I were talking about this. We didn't do a full-on sort of messaging engagement around LaunchDarkly's messaging. We just sort of talked about it in the context of the video. Andrea, do you feel that the website messaging is the Promised Land, or do you think about, sometimes, should it be something higher? Maybe that's a future thing?
AE: Yeah, I think now that we've done the video, perhaps it is de-coupling your code deployment from your feature rollout. You know, trying that as the main messaging could work well. To Andy's point before about whether it's "attend the ball," or "marry royalty, or higher up," that's something we're just kind of trying to figure out how far up the value prop we should really be going.
AR: To improve their craft, a lot of writers read other writers, filmmakers, and watch other filmmakers. Next time you're looking at websites, whether it's your own, your competitor's, other ones that you just happen to visit, ask yourself this question: who's the main character they're talking about?
Two questions: ask who's the main character, and what's the Promised Land being offered here? And see if it's clear. That's really a good way to start bringing this stuff into your daily thought process and to get better at translating the core story into the messaging.
I don't think that's a problem, and I see it very frequently. I think, ideally, they're the same story. Ideally, the story you're telling to the marketplace is kind of like the "Episode I" of Star Wars, or "IV," whatever it is. And then so it naturally fits into the rest. It's not somehow opposed to it or contradicting the other story, but I think sometimes, yeah, for the different audiences, you might have a different Promised Land, different story. Especially for investors who are looking for something longer term.