July 20, 2016
Improving RocksDB’s Write Scalability & Counting Things at Smyte
Heavybit member company RainforestQA recently hosted the RocksDB meetup in our San Francisco Clubhouse, where speakers Nathan Bronson from F...
Hey, guys. Great to be here tonight. Thank you so much for coming. It really means a lot to me. It's nice to see so many friends and familiar faces here. I especially want to thank Tim, Jason, James, and the Heavybit crew, not just for putting this together, but for everything they do in creating this really great space, and environment, and community.
I haven't given this talk before, and it's about 120 slides, so we'll see how long it goes and either it'll take like ten minutes or we'll be here until tomorrow morning. We'll see how it goes.
That's my contact information. I am slowly getting older, so I've worked at a bunch of different companies now. Most importantly and most recently, a new startup that we're doing upstairs with my co-founder, Scott, who's over in the corner, called Cloud Connect, working on customer data integration. I'm happy to call many of you in this room customers. Prior to that, Dropbox, a lot of time at Salesforce, a company called GrandCentral. The first company I started in the web analytics space called Personify, and a million and a half years ago when dinosaurs ruled the earth, I worked at Stanford Research Institute.
Tonight's talk is really about kind of a combination of those things. One is really focusing on a lot of my experience and things I learned at Sales Force. I was there for about seven years doing marketing and learned lots of interesting lessons. And then I left Sales Force about three and a half years ago. I've been lucky enough to spend a lot of time working with a variety of early stage companies. Again, many of you here in the room, folks like Heroku where I was lucky enough to be involved early on as an investor and adviser. Pantheon, the Streak guys, RingDNA, Attachments.me and others, and I've had a bunch of conversations with start ups about marketing. And what I've attempted to do tonight is clumsily condense those into a couple lessons of the most frequently kind of discussed topics or kind of the marketing FAQ.
If you were here a couple weeks ago, my former Dropbox colleague, Ivan, gave a great talk on kind of bottoms up marketing and growth hacking which he really, although I think he avoids using the term, is the best practitioner of that art. Really a lot of where the consciousness of the startup community is today, really what you think of as kind of customer acquisition, everything really directly related to on the ground, working with customers.
As somebody who's been in the marketing world for a long time it's very gratifying to see so many great minds coming into that space and reinventing the techniques and all the work and attention that's focused there. I would argue though that there's probably a little too much focus on that, relative to kind of the broader scope of marketing. At least certainly since Ivan was here a couple weeks ago talking about bottoms up marketing, I thought I'd talk more about kind of top down. And the two are obviously related, but instead of just literally, how are we getting people from that one stage of the funnel to the next, little bit higher level in terms of some of the broader functions that marketing serves from pricing and positioning and PR, how you think about BD, alliances, all that kind of stuff as well.
I'm lucky enough to be a practitioner, and have been for a while, lucky enough to be a practitioner of developer marketing, which is something that all of us get to participate in.
I think we're all really uniquely lucky to be in developer marketing, because the history of developer marketing has foretold the history of marketing. If you think back to, and developer marketing goes before this, but perhaps its most important era, or initial era, 1984 with the release of the Mac and the Mac Developer Relations program, and there in the slide you see a youthful Bill Gates.
That's actually an ad from the original Mac launch where of course it was really important to recruit three companies to be Mac developers. One of course was Microsoft, the other was Lotus -- and many of you probably have Mitch as an investor in your company. You can see him there. And then there's this guy whose name of the company I can't even read, like Software Development Corporation, I don't know. Maybe somebody older than me remembers him. Clearly, his Mac product didn't do as well.
But this idea of actually finding developers to write stuff for your platform and your product has been around for a long time. And, gee, look at the techniques that were developed and were used way back then, even in 1984. To do this, w hat did we have? (Using we generously). We had online communities, right? Creating what was fundamentally a user-generated content; people creating applications, kind of participating with these platforms.
It was hugely content marketing, which again today, like all the people at the Marketo summit today, I'm sure they've heard all about content marketing. Well, this was happening in 1984, right? We had fundamentally what was bottoms up adoption. So if you think about the state of IT in the 80s and early 90s, this stuff was sold to the CIO by guys who looked like Don Draper who had business cards from IBM.
What Microsoft, and I'll give Microsoft particular credit here, did that was very radical is they actually went to the developer first and did this kind of very bottoms up style marketing. This whole kind of consumerization of IT really owes a lot of its heritage and roots to developer marketing.
All of which is to say that this art that we are practicing in developer marketing today will continue to lead a lot of marketing practices and innovation.
So you should consider yourselves to be on the forefront as opposed to followers, which led me to one of two Andy Warhol references I'll make in this presentation. And this is something I used to say a lot at Sales Force, which is:
"In the future everyone will be a developer for 15 minutes."
These kinds of practices and how we relate and communicate with developers in developer marketing is something that again you see being increasingly common if you look at anything from YouTube to Facebook. So, that's my feel good about being developer marketing part of the presentation, and kind of what we're really here to talk about.
There are Seven of Them
This is one of my favorite ads of all time. This is from 1988, This is from 1988, I believe it's '88, might be '89. This is the original Next computer launch ad, which I managed to find online. It's kind of slightly yellowed out and as you can read the premise of the ad is that there are going to be ten great things about computers in the '90s and here were seven of them being introduced. I actually used this format once for a deck at Salesforce, that unfortunately for the company everybody had to be trained on. Y ou remember Future Software, are there any Salesforce employees? OK, no. So, I love all of the Apple and Next stuff, this is great stuff. You should look at it, get inspiration from it.
The reason why I put it here is I'm going to talk about my seven principles or ideas that I would frequently find useful in talking to different startups, so you guys will get the structure of the talk.
OK, number one, and if there's one picture that you guys remember from this presentation, I would argue that it should be this one. And for those of you who have had the misfortune of being in a marketing meeting with me before, you'll recognize this picture at least slightly from some scrawled version that I put on the whiteboard.
But, I keep going back to this framework, because for me it really answers a lot of the questions of fundamentally, what are we trying to accomplish with marketing.
What this slide is trying to say is there are really two dimensions to thinking about your company and its position and what you're trying to achieve, and those are essentially how strategic you are and how emotive you are.
Those, at the essence for me, are kind of the two characteristics that matter most. And of course when I think about what is the most strategic and emotive company out there, and I think this is probably a view that's widely shared, it's something like Apple. You'd call that kind of in contrast, you know, when people talk about Apple they'll sometimes talk about, oh, and Harley Davidson. People really love those products. Well, yes, that's a really emotive brand, but that's not a strategic brand. There's nothing strategic about being a motorcycle manufacturer. But certainly Apple is strategic. It has this unique combination of both of these things, because its not just this incredibly emotive experience, but they also are the most important player in mobile, which is the future of computing, and TV, and music, and media, right across all these different domains. cross all these different domains.
So that's the end state. That's who you want to be. And guess where you are? We all start here. Even Apple was there. You're at the bottom right hand corner or what my co-founder Scott refers to affectionately as the kiddie pool.
Our job collectively is to kind of plot this migration path up to the upper right-hand corner. You can kind of think about different companies and where they sit on this dimension.
You take a company like Oracle, even being a little bit long in the tooth, it's still a hugely strategic company. It is the most important company probably in enterprise software. It shapes that market. You couldn't imagine a lower emotive brand than Oracle. You see their ads in the airport and it was one guy in Photoshop for two minutes or something. They just don't care, so you can imagine how they sit kind of low on the emotive scale but really high on the strategic scale. And that's kind of one set of companies, right? Companies that are truly strategic.
You have another set that are, and this is the worst, companies that think they're strategic, but in fact are commodities. We've certainly seen examples of that and we see companies that are kind of slowly shifting back from being strategic into being commodities. That's kind of the Dell, HP, arguably Intel kind of slide, as you can argue where their position sits in the universe. And there's kind of another random example that I find fascinating, which is companies that think they're commodities, but are really strategic. And they use that kind of language. I think Amazon is certainly in that category; and a lot of what you see happening with some of the Chinese contract manufacturers.
In general what's happening with the contract manufacturing is very interesting. These hardcore commodity players which are now potentially disrupting some of the mainstream consumer electronic brands, all of which has nothing to do with anything, except to set up my honor for the worst company positioned in this graph, which is a company that thought they were both strategic and emotive, but in fact was neither which is Sun. And if you guys remember, JavaWorld 2000, 2001. Remember when they hired Britney Spears to be the spokesperson for Java.com? Like the world's worst effort to attempt to be kind of this emotive brand. It was awful. OK, so we'll actually make some of this more useful now.
So, your job is to kind of plot in some series of moves, or paths, your direction through this space, and I'd argue that there are kind of fundamentally two directions.
You have this direction that goes from strategic to emotive -- that's kind of the traditional enterprise path, and you have this direction that's kind of emotive to strategic -- that typically is more of a consumer path.
And you can think about different companies and the techniques they use through this lens, and I find this useful. So you can imagine where Salesforce started. It's hard to remember this far, especially if you weren't in the industry back then, but 2002, 2003, there were a million contact management websites. That whole space was littered with undifferentiated, unstrategic competitors.
In fact the whole space was considered so unstrategic that Gartner's official recommendation was you deploy Salesforce while you're waiting for your SAP or Sable system to come online. It was just so pejorative and dismissive from an entire industry framing and of course Salesforce managed to become both strategic, and I would argue one of the things that really Mark deserves an enormous amount of credit for among other things, is being emotive.
If you've ever been to a Salesforce event, what makes a Salesforce event different from a traditional VMware industry event, they've got that sassy character who's played by the guy who's the character mascot for the Giants.
They've put a lot of production value into a character. Why? Because they're trying to increase the emotive quality of the experience and I think they've successfully done that.
A couple other interesting examples, I'll give. On the consumer dimension you can kind of think of somebody like Instagram. And imagine that kind of, really highly emotive experience you have the first time you apply that filter, and it's like, wow, this is really cool and it's pictures of my friends. There's nothing more emotive than a picture of a human face. We're wired to think in emotive terms about that. At the time that Instagram launched, if you had the same reaction I did, you're like that's great. It's a freaking picture taking application.
There are 10 billion companies that hang out in that high emotive, low strategic value, like half the crap that comes out of YC or whatever. And then all of a sudden this kind of magical thing happens. There's this transformation that all of a sudden Instagram went from being this random photo taking app to being potentially disruptive to Facebook and like an industry shaper.
I don't know, if you're like me there was a moment where that whole framing of how you think about that company changed. And probably the same thing's true of something like Twitter, right? When it started you're like, what the hell is this thing for, I don't get it. It's stupid, I'm never going to touch it. And then it's like, oh my god, this is like the future of media, right?
And I offer that as kind of an imprecise way of thinking about the same transformation that you want to effect for all of your own companies.
Where, today, because you're in the bottom left- hand corner, like the rest of us, nobody think about you being special in that way, and what that shift feels like when you've made that kind of transformation. I'll give Box some credit here. When I was at Dropbox, Box was like podunky, super undifferentiated.
If you even think about that whole kind of file space, there's nothing more low rent and kiddie pool than software that you download to your PC. Because this is the world of like malware and anti- virus and browser toolbars. This is the dregs, there is nothing strategic or emotive about these products. And obviously Dropbox has done a great job in going to the upper right-hand corner, particularly some of the work that John Ying's done on the brand, great, amazing emotional qualities.
What Box, I'll give them credit for, has done is taken the Salesforce playbook and literally done it verbatim to move into the strategic quadrant. And I think that those of you who watch the industry can probably relate to when you remember Box as being a crappy way to share mp3s on Twitter, to now being this kind of fundamental and important cloud infrastructure thing. And they very much actively did that, right? That didn't happen just passively, that happened because they actively projected it. It requires this kind of internal force to make it happen.
I think Heroku's actually an interesting exception in the sense that I think if you were to plot Heroku's path, because the brand was so emotive to begin with and because the nature of the rails community.
One could argue that they kind of did the emotive first path, in that through the positioning and then later the acquisition it kind of became more strategic. But I don't honestly know if that's a common thing.
Or you can almost imagine a kind of straight line, but obviously they did a great job on that front too.
So, the question is, how do you become strategic? What is the key ingredient to being a strategic company? That is number two, having an industry transformation narrative. Every company needs one.
This is perhaps the most important slide I ever built. It's just about to celebrate its 10 year anniversary.
If you have been to any Salesforce presentation or Dreamforce or go to any one in the future you will see some much nicer version of this slide in 16 x 9 HD. But the core message is the same. And if you think about all the things that Salesforce could have chosen to talk about ten years ago, and they talked about this before too, but in particular in this way.
They could have talked about CRM. They could have talked about content management. They could have talked about generating more revenue for your customers. What did they choose to talk about? Something that's a little bit unexpected, right?
We now view it as obvious, but it was not obvious at the time. They talked about industry transformation. Because at the end of the day there's nothing more strategic than industry transformation. That was what separated Salesforce from being a company or being a product into being a stand-in for a movement.
The same way that things like Facebook are today. And it actively positioned itself there. Salesforce chose to talk about its product that way, it chose to develop and reinforce this narrative over and over and over again. Implicitly you can see what was happening here, where Salesforce is positioning itself with other at the time cool internet brands, like eBay and Amazon. That's a very, very explicit decision about how the company is asking to be perceived.
You can think about some of these other shifts. Obviously, the industry shift from PC to mobile, that's like a core force that's driving a lot of implicit positioning. If you think about, even a company like Zynga, that's doing something as silly as having people tip your cows for you, or whatever the hell's happening in Farmville. Well, there's a strategic narrative about the shift from console to casual gaming, that was really propelling the Zynga narrative.
There was the Facebook narrative of, it wasn't just about sharing photos. What the strategic narrative of Facebook was, it's the transition from search to social, and the billions of dollars of advertising revenue that's at stake there. And even with something like Heroku, the transition from hosting to pass, and kind of the distinction there. All examples of narratives.
So you need a narrative. I don't know what your industry narrative is, but I can promise you that it's based on this fundamental idea, which is that,
"The internet changes everything."
It's just some derivation of that. The internet is the most important cultural force of our lifetime. That is already true and indisputably will be true in the future. All of those things I talked about were just kind of local instantiations of that core fact. So, if you don't know what it is, just start with that.
Literally when you write your press release or your blog posts, just put the title, The Internet Changes Everything, and kind of roll from there and you'll come up with a better resolution version of the future. But that is the fundamental truth, and it's funny, if you go back and look at the press releases I've written. Gordon Evans would relate to this. When I worked with him at Salesforce, literally you would see the line, the internet changes everything in there, and I think I worked with Jesse on an Attachments.me release not too long ago and I had that line in there too. It's just kind of the core narrative of our time, and it's so just emotionally true, and intellectually true. It's like, how can you dispute that?
That's kind of implicitly not strategic, and I don't think commoditization, and maybe I was unclear about that before, is something that's implicitly not strategic or bad. Certainly, in kind of a Clayton Christensen formal sense, something like Salesforce is a lower cost replacement for its competitors and a lot of the time that's the case. Or as I used the examples of contract manufacturers who are absolutely commodity driven who I think are hugely strategic.
It's just really a question of your business model is aligned to leverage that, and in fact the most interesting and fascinating developments in our industry are those who are able to exploit commoditization well.
And even look at the dramatic changes, like when Apple/ Steve Jobs would fret about building these factories and [inaudible] that had these millimeter precisions for milling cubes and it all had to be owned by him. It was essential, vertically integrated, as important as owning the OS. And of course Apple built new Apple on contract manufacturings. Anyway, we could talk about that for a long time.
OK, number three, differentiation. This is another one of my favorite Saleforce quotes that I would use all the time.
"It's better to be different than it is to be right."
And I say this because as developers we have a natural tendency to want to speak in very precise and accurate terms.
But there's nothing more important than just having your fundamental differentiation. If you're not differentiated, you're dead. I mean, there is no essence of you.
And I use the Salesforce, they don't use this logo much anymore, but this for many years was the logo, and the logo was the big, fat 'No Software' bug. And their core logo, their core identity, was the negation of something. How could you be more implicitly differentiated than to have a negative symbol be your identity? And I use this example of "It's better to be different than be right."
I've ridden the N Judah for many, many years, and if any of you have ridden the N Judah you live in the Cole Valley like me, you'll know Craig Newmark, who of course of Craigslist, who's a wonderful guy; and many, many years ago, I'm on the N Judah with Craig and we're just catching up and he's saying,
"What are you up to?"
I'm like, "I'm working at the Sales Force thing."
He's like, "Oh yeah, that no software. Yeah, I don't get it, you know. I don't really understand. Of course they're software. You guys write code, they're software."
It's like, well, that's not really the point. This isn't designed to be like a literal interpretation.
What Salesforce was trying to do fundamentally of course was position against a model. We weren't trying to position against a company, we were trying to position against a model.
There's another random story I'll share here. When I was at Dropbox I got a call from the Economist. And normally you go, oh great, the Economist is calling. We can do a great story with the Economist. And they wanted to do a story about how a lot of the internet giants weren't doing that well, or they weren't as innovative.
You know, eBay, or Yahoo!, or Google, they weren't innovating as much as they should, and like hot new start-ups out of the YC world were doing well. And I'm like, we do not want this story written. The last thing we want is to be positioned against companies like eBay or Google. Nobody wants that, right? And what we wanted is to be positioned against Microsoft, against Windows.
And in one of my greater PR accomplishments, if you read the article, at the bottom of it, it talks about what Dropbox is really competing with is the Windows 3.1 model of file sharing.
And that was a very, very explicit kind of differentiation that we were trying to achieve and kind of speaks to how important it is how you pick who you're really going after.
Does Dropbox compete with Windows in the literal sense? Does it sit in the shelf next to them? Does Dell decide which to put in OEM boxes? Of course not. Conceptually those are the kind of things we're thinking about. The same way that Byron famously would talk about how what Heroku competes with, right, is servers. You can think about that literally or a little bit more large.
OK, number four, ingredient branding. This is one that I'm a little bit on a tear on recently, because I think it's a really important one and one that people can do a better job on. I was in the Ferry building this weekend and there's a stand that sells ultra-Portlandia, organic high- end versions of classic, crappy treats. So they have, like Ho-Hos that are made from Petaluma milk or something. They literally have Pop-Tarts that are hand, artisanal crafted Pop-Tarts. And I thought, this is a great example of ingredient branding, because the only thing that matters is the ingredients.
In terms of our industry, and stuff that's a little more relevant, things that should be obvious are like, the mouse, right? There's the whole concept of the Mac, but the mouse is like the thing that stands out as the key ingredient that physically manifests the differentiation in the message. So you can point to it and say, that's what makes it different. It's a way of taking some of these abstract things and kind of putting a box around it. My friend, Ken Norton, used to use the example of the pause button on the TiVo, and say, why is that the biggest button on the remote?
Well, it's the most important feature, cause it's the most differentiated feature. And so like that was an instantiation of the ingredient branding.
Ones that I think are more relevant to us in the developer community, and probably some of the really great ones, probably the most important word in IT created in the past 15 years is the word virtual machine. If you think about the amount of money and the architectural disruption and the size of that category, for something that didn't exist before.
And it was important that VMware had this idea and others had this idea of a virtual machine that you could talk to and point about and infuse with all these benefits and differentiation. They weren't just out there talking generically about virtualization. They were saying, we have this thing called virtual machine.
One of the words that we introduced, and this only came later, this was like '04, '05, if you look at the history, and a word that became essential I think for Salesforce's positioning is the word multi-tenancy. And again, it was a very active decision to invest in that word, to repeat that word, to use that word, to summarize a set of technical differences that were essential for carrying the value and benefits of the product.
I'll use two Heroku examples here, and this is where I'll talk about my friends in Heroku Postgres as an example of a challenge. So, clearly like Heroku's popularized the idea of a Dyno, which helps communicate some of the value, although there's some pieces of the value that are out there.
I'd argue, as an example, and this might be too esoteric for some of you, I don't know if you guys know the Heroku Postgres product. It's a freaking awesome, awesome product. It does all these magical, wonderful things like, you can automatically shard and replicate. Well, not shard per se, but do multi-backups of your database. It's a really great thing.
The problem is, like, how do you talk about where all that special stuff lives? Because there's the database, which people understand, and people understand virtual machines. But there's this magic layer in between that is doing all this management of the database automatically for you, and I use that as one example of something I probably could use, an ingredient brand to really carry the differentiation.
If you don't have it, it's hard to talk about all those special properties, because otherwise they're being attributed up into the database, which people already understand, or it's down into the VM, which people already have kind of a cognitive model for. So that's again, kind of creating that additional layer.
I'll use another example of something that I think should have happened in terms of ingredient branding. How many of you know what Google calls their database? Big Table, you guys know? You guys are generally familiar, right, that is a concept. How many of you know what Salesforce's database is? It doesn't have a brand. It was never explicitly branded.
So you understand that Google has this really special, magical database that they built that has all these special, magical Google properties that represents a lot of value. Salesforce hasn't really explicitly done t hat ingredient brand on its database. In a similar way I think it makes it hard to communicate the value of something, it is really quite interesting. OK. Bouncing around a lot here, hopefully this is right. We're good? We're good, OK.
Number five, this is an important one, and this is a little bit more on the customer acquisition side, a little more tactical. Leveraged acquisition. Man, this is basically what I did for ten years. So the idea is, if you've got a developer product, how do you get a lot of developers? Well you go to where the developers are. So, what did we do at Salesforce?
We very explicitly thought about, what were the largest developer communities and how do we align ourselves with them so that we can communicate our message directly to where those people already are?
And, remember, Salesforce didn't have a natural affinity to some of these developer communities, the way that Heroku and Rails does. These had to be very actively forged. So, one of the Pantheon guys actually came up with this, which I kind of like, which is the idea of minimal viable marketing. And what we would explicitly do is create a product that was kind of the MVP that was necessary to carry a marketing message and then launch and use that and then plug behind that all the kinds of traditional marketing campaigns to target that community.
So, the question is, how do you do that?
Well, you identify what your company can do for that larger developer community, that larger platform, that's something unique. And when we were doing this at Salesforce ten years ago, we weren't that big. It's not like you have to be super large to be that special. Certainly later, it was a little bit different. I remember, of course, Microsoft back in the day was probably one of the larger developer communities, still is.
What did we do? We created a Win 32 wrapper around our API, and at the time Microsoft was very interested in getting people to build apps around Office. And they were spending a lot of money on this whole like Office developer conference and programs and stuff like that, and we said, hey Microsoft, let's launch this thing together called the Salesforce for Office developer platform, whatever the hell it's called, which was just a Win 32 wrapper.
It was nothing magical, and tell a story about how you can build apps for Office that communicate with Salesforce. You know, this thing was money, this was gold, it got picked up everywhere. Now we have a reason to talk to Microsoft developers. They have an entry point to learn about this crazy web services thing.
And there were a bunch of these we did, I'll call out a couple that were highlights. A couple years later, the Adobe community was important, especially as the Salesforce platform was evolving to the point where you could do more robust things in the presentation layer. We wanted to get web developers. Where are all the web developers, creative people that were in Adobe?
So we essentially did the same thing for Flex. We became one of the great stand out enterprise stories for how you use Flex in a business, which a company like Adobe, that's certainly useful for them. We met James Ward in the process, who was at Adobe and later went to Heroku. That actually evolved so far that there ended up being a special version of Flash Builder, which was what later Flex became, specifically for Force.com. So it kind of emerged over time.
The key thing I want to emphasize here is that these aren't like really complicated large scale BD projects.
These were two guys and a compiler and a Word document. None of these, except until they were much later along, had contracts associated with them. These were not BD, like let's go forge an alliance.
This is, go find the right product person who has a problem, who will support your marketing efforts when you launch a product with them.
And that kind of stuff is great.
But the message was great. The message was about the future of software and the future of telecommunications and the thing had real value. People used it a lot. So, it doesn't have to be complicated.
I was working with another company I won't mention and there was a larger IT company that has a private on- premise pass platform and they wanted them to be involved and they were thinking, like, we're going to have port our whole product to this thing. Like, how are we going to think about this from an engineering point of view. It's like, no, they don't care, right? Just build a shim. The minimal viable thing in order to align with them, and you can reap all the marketing benefits. The last one I'll point out because it's the one that I'm most proud of.
What was the most important developer platform?
In 2007, it was the release of the first iOS SDK, and if you remember, and that's actually a screenshot from the iOS launch. And you can see it says Salesforce iOS SDK launch. You can see it says Salesforce on there. Salesforce was one of five companies to participate in that. So, we worked hard to make Salesforce relevant to iPhone developers, even pre-SDK, and that we worked hard to endorse the message of iPhone for business.
And that ultimately led to a really great relationship, we were invited to be a part of that. You can imagine, if you're trying to build a developer community, developer brand, you're at the premier developer event, one of the most important developer events in the history of platforms. That was pretty special.
OK. We're getting close to done. My second Andy Warhol reference, and my homage -- this quote is from Bruce Francis.
"Repetition is the soul of modernity."
And this is another thing I wanted to really make sure I highlight especially for developers. So we're talking about doing all these things, industry narrative, transformation, yadda, yadda, yadda.
The part that I want you to feel aesthetically comfortable with is the idea that you're going to repeat your message so much and so frequently and so consistently, it's literally going to hurt when it comes out of your mouth.
When your mouth is like dry for water, that repeating your message so frequently, that's when you're getting started. That's why the Warhol slide's there. This is how modern communication works. Don't feel encumbered or afraid to just repeat yourself again and again and again. You should imagine yourself like a classic rock band that had like one hit and you are going to play that song again and again and again and again.
Once you've launched a product, find a way to launch it again. And then when you're done, launch it again.My Heroku example on this is the release you guys did a couple months ago with the Heroku Enterprise Java. Well, what's the point of that? It's to remind people that Heroku is multi-language. It's obvious to everybody in this room that's the case. 99.99999% of the universe has no idea that you can use Heroku for anything other than rails. And how are you going to solve that? Via repetition. Just repeat it, repeat, and just feel comfortable, feel great about it. You're Lynyrd Skynyrd. It's Freebird. Just go for it, embrace it, love it. I can't emphasize that enough.
OK, number seven, last one. Getting a little random. I end up having a lot of conversations with companies about what kind of PR and what should they launch and how should they think about launch. This is my, how to think about PR for developers?
And this is to understand that media and PR have a set of APIs. And you can only invoke the APIs that they have. You cannot create your own APIs, you cannot think about creative ways of invoking APIs. They have APIs and they're listed in this order. The higher up you go, the better your message is. So the top API you can invoke is controversy.
So even in the past day, if you look at what happened with that Arrow TV service and Fox doing them the biggest favor in the world in getting up and crapping on them. What do you have? You have controversy. That is going to be covered left, right, and center. That is the catnip, that is the main, like highest value API you can invoke. Now, it's challenging to do that in a positive way, but it's there.
Number two is acquisition, and I remember working with a very, very small company that acquired an infinitesimally smaller company, and it was really just one person that was hired. But when we did the release we positioned it as an acquisition and man, we spent months working on product releases that didn't get more coverage. I t's just that the press has an API and the API is called acquisition and when you invoke that API it invokes a certain amount of stuff behind it.
Financial, funding announcements, partnerships, that works very well for what we were talking about before. These are built in APIs that you can invoke.
Product launching, then you get down to like hiring and metrics. What's a weak API? I don't know, it's just really, you're not getting a lot of resources for those. So think very actively.
There are fixed slots that you are going into when you're writing that release. You are not doing this for the first time, you're doing it for the millionth time. There are APIs and invoke them.
I'll throw in my obligatory Salesforce story here, which I once saw emulated I think in a Facebook deck, which I thought was hilarious. Sometimes when you look at a Salesforce deck, you'll see a little thing that says "new." And there'll be like a bullet point and it'll have a little "new" logo or bug next to it. Why do we do that? Because we were just like so trying to invoke the API, just point out exactly what it is -- this is the new thing, this is what you pay attention to.
You cannot be more obvious or straightforward or deliberate in how you interact with that kind of high level communication.
OK. In the future there are going to be ten important things about developer marketing, here are seven of them. And that's it. Thanks so much.