a16z Podcast - a16z Podcast: Bending Every Pixel to Your Will -- Optimizely and the Next Wave of Internet Tools

Episode Date: May 20, 2014

Optimizely is a superb example of the democratization of software development. You don't need an engineering degree to fire up Optimizely and start testing how design changes on your website -- down to the pixel level -- affect things like time on site, closing sales, navigation, etc. The a/b testing Optimizely offers is just one example of a new wave of tools born of the internet, and designed for how people work, shop, research and entertain themselves online. Andreessen Horowitz's Scott Weiss, who is taking a seat on the Optimizely board following a16z's recent investment, a16z Partner Tom Rikert, and Optimizely co-founder Dan Siroker discuss the next wave of internet tools, where entrepreneurs are headed next, and how virtually anyone can avail themselves of this technical brawn.

Hi, this is Scott Weiss. I'm a general partner here at Indreason Horowitz, and I'm here today with Dan Serocher, the CEO of Optimizely. I'm also here with Tom Reichert, another partner here at Indreason Horowitz, who is our enterprise SaaS expert. So we're gathered here today to talk a little bit about Optimizely. We just completed a $57 million round, and since I have the founder here, I've always I've always wondered when companies have such a founder market fit where this idea for Optimizely came from him and his co-founder Pete while they were at Google. I'd like to just go back and capture that moment at Google.

When did you know that you had to make a company out of the testing and optimization that you guys are doing as product managers there at Google? Well, the seed was first planted after I graduated. I was a 23-year-old kid straight out of school at Google, and I was working with a bunch of engineers who are twice my age, significantly smarter than me. And one Friday afternoon,

I found myself going in to have to pitch a controversial product idea to the founders of Google, Larry, and Sergey. And before I went in, I asked one of my mentors for advice about how I should pitch this controversial product idea.

And he gave me some advice that sticks with me to this day and was actually the first seed that we planted to start optimizing. And his advice was to go in and tell them we just want to run an experiment. Just run an experiment. And those magic words transformed the conversation

with the founders of Google from a controversial idea to a data-driven idea. It says, hey, let's run an experiment, let's see what the data shows, and if it's a good idea, it'll prove it out. And to this day... Who could say no to let's run an experiment? That's right.

And actually, then that seed actually really grew into a blossom when I left Google to go to the Obama campaign, and that's where I saw firsthand, And the distinction between an organization that's a data-driven meritocracy of ideas and an organization like a political campaign, which, believe it or not, there's politics and politics. And even as much as the Obama campaign was trying to buck the trend of a lot of the political things that go into campaigns, the decisions were often made based off of the hippo,

the highest paid person's opinion. And so, you know, David Axelrod and David Plough, they would have smart ideas. But one of the things that I was able to introduce into the campaign was this idea of experimentation. And really, that's where this initial seed that was planted. Google grew into this much bigger opportunity of A-B testing and website optimization. But how did you know you wanted to build a whole company around it? Like, I think it was a great idea, and then you took it to the Obama campaign. But what was the, we've got to make this accessible for everybody? How did that come about? Well, it was really actually a narrow focus on just building the product I wanted. And in fact,

I never started the company thinking this is a ubiquitous problem. In fact, we got more lucky than good there, realizing that the problem we face in the campaign of wanting to do A-B testing, enabling non-technical people to do A-B testing, was actually a problem that many organizations face. Even the biggest organizations, you know, companies like Starbucks, a Fortune 500 company that wasn't doing A-B testing before they came to optimizely, fundamentally can do this because it now enables non-technical business users to do A-B testing. So, Tom, I wanted to hit you with this one because when he just talked about bringing AB testing to non-technical people,

you know, as you've looked at at enterprise SaaS companies, this whole notion of disruption by, you know, just something that is usable by non-techies, not needing an engineer to implement these experiments. We've seen this across a number of companies, and it's really accentuated here with optimizely. Yeah, I think there's been a few interesting trends, which have really helped to make this kind of revolution happen. For one, enterprise software has gotten so much easier to use, right? Like in the old days, enterprise grade was like better. But I think today, consumer grade, the consumer grade user experiences are actually the superior experience, and that's what companies are starting to build. So even the most technical tools, whether it be for data analysis or A-B testing, are things that are more familiar to someone who might be, you know, someone coming out of a university program who's never coded, but they've used Facebook.

They've used Gmail to use these different applications, which has trained them to become excellent. enterprise users. I think that's something that's really exciting, but it takes a real focused effort from companies. It takes a lot of discipline to build really good user experiences. In fact, you know, A-B testing, you can even do that with your own product, even if it's an enterprise product. So it kind of comes full circle to what optimizely is focused on. And to add to that, you know, one of the things that I've seen over the last four years is, you know, there's two things happening. Our customers are getting less patient and their expectations are going up. And that combination of things means that they expect to just work.

They're used to, you know, consumer products that actually just work. They don't need to have a training manual. They don't need to go through an onboarding. They can just go in and from the very beginning understand exactly what the product does and are delighted in that experience. And we focused on building a product that just works. And it happens to be a product that businesses use, but the expectations in our standards are those that we would hope to be ubiquitous for all kinds of products. You know, one of my favorite engineering expressions is RTFM, which means read the fucking manual. And it's always sent to the newbie when they ask a stupid question of another engineer.

They just get a one line sentence that says RTFM. I agree with you, what you just said, if it needs a manual, it's probably not a mod, doesn't have a modern UI. It probably doesn't have, it's not a consumerization of IT product. And this notion of designing for, you know, kind of an average person, I remember like the old days of enterprise software that Tom was just talking about, you know, specialists had to enter just enter simple things in. You just think of what like an old reservation agent used to do to sit on the other end of the line and enter your data in that now we enter in an Expedia with

drop-down boxes. And I think the whole world is headed towards that way. If you look at the new applications like Box, like Evernote, you know, a lot of these new applications, Applications are built for mobile. They're built for kind of an HTML user interface, and they're intuitive in the way that Tom was talking about like Facebook is. So everybody can enter in their own data. We don't need this specialist in the middle, and we can get more things done. So I want to switch gears a little bit. One of the things that has really excited me about this investment is this new wave of enterprise companies. Like I honestly don't think that the old line companies are going to make this shift, make this change. And one of the reasons I contend

is they can't hire the design talent. You know, if you think about the designers that are necessary to come in and make an interface that's that easy to use, they come from consumer land. And by and large, they're the hardest ones to hire across our portfolio. Now, you walk into optimizely, the culture, the look and feel is reminiscent of a consumer company. In fact, it feels like Google. It's got ping pong tables and, you know, open seating. Like you walk into IBM or Oracle or any of those old companies, like, I don't, I think a designer walks in there and gets hives. So, like, what is your prediction? I actually believe that there's going to be new franchises that are made on top of these franchises. And when even the oracles

come knocking, like the cultures are going to be so different and distinguished, like, I don't think they're going to take the buyout offers. It's just like, what am I going to be doing? What sentence am I putting myself to go work for this company? How fun is that? Yeah, and I think that's a great point. And I think it's a trend that we've seen and it affects the entire process of the consumption of our product. And I think one way to frame that is, you know, today the best products are bought. They're not sold. People come to optimizely and they try our product and they convince themselves of the product before they even talk to a salesperson on our team. And, you know, we do things like, you know, in our ads, we actually put a little form field that you can enter a URL directly into

our ad, click, test it out. And what we sell you on is the product, it's the product experience. And world-class designers want to build products that are bought, not sold. And I think we have a huge opportunity to transform the mentality of what an enterprise software product looks like and feels like to be able to be just as delightful as a consumer product. And that's one of the things I loved about Optimizely is the time to actually do something interesting in Optimizely. It's about two seconds. You go to Optimizely.com and you can type in a URL. Instantly, you're editing a page being able to run an experiment. So the old days, getting up to speed on an enterprise application was,

if not a manual, at least a training class or a tutorial or watching a video. But optimizely, I mean, it's literally a click. And it's immediate gratification to see the changes and actually put them alive to start to see data coming in. And that's something that is magical. I remember implementing Oracle Financial, and I had EDS at Ironport for like three months training everybody how to use it. And it was just training the finance staff.

It wasn't like anybody else could possibly. use it. And so I think this wave is going to help everybody kind of really understand and get benefit out of enterprise software. Hey, there's something that I brought up a minute ago that I'd like to touch on, and that's the culture optimizely. From the minute you walk in the door to, you know, every interaction I have with anybody on the staff, you can tell that you've got happy employees there. And I just like, what are some of the secrets? I know there's a lot of entrepreneurs out there that are thinking about culture, but just don't know the underpinnings, how to build it. What were some of the things that you and Pete did early on to make sure this is going to be an

enduring place? Well, it's something that, I mean, first of all, we really thought about. I think that's a very important step to take to say that we are going to be purposeful about our culture. It's not just going to be the organic outgrowth of a bunch of people, but there's going to be somebody, in this case, who was me and my co-founder, really thinking about the kind of culture we wanted to build. And one of the litmus tests I use was, is optimising the kind of place I would want to work at as an employee?

And as a very entrepreneurial person, that's a very high bar. And the biggest part of our culture that I think resolves or solves that is our culture of ownership. You know, we really focus on empowering people to own their successes and failures and to do that for their team as well. And it starts, I think, with me and Pete, we view our role as working for our team, and they view their roles working for their team. In fact, you know, we codify that in some of the symbols.

One of the things we do is we action, every org chart at Optimizely, the managers at the bottom of the org chart, not at the top of the org chart. And that's a reminder that not only do we actually work for our teams, and we do what we can to make them successful, but we also have a hiring philosophy of hiring above the mean. We want to hire better than the average of the other people at Optimizely. And, you know, those are small symbols that get translated over and over again in the actual culture that materializes.

We've thought long and hard, and we have a cultural acronym that we use internally called Optify. It stands for the cultural values we care about. It's ownership, passion, trust, integrity, fearlessness, and transparency with a why. And those are all things that we, and I think the culture is only as effective as your ability to actually live up to that culture. And, you know, Pete and I both, when we communicate with a company, we'll say not only what we think we should do or what our decision is, but more importantly, why we think it's important. And that will say, you know, I think the reason why our, you our travel expense policy should be X is because we have a culture of ownership and of trust.

And those are things that we've codified and then we live up to in our actions and deeds. I want to talk a little bit about transparency because it's something that I've always held dear. You know, I would go over the board slides with the whole company. And I think one of the things about being transparent is to not be afraid to admit where you're wrong. In fact, I think when the leadership stands up and says, hey, we really screwed this up, it kind of enables and encourages everybody else to be very open about what the problems are. And that was one of the things that really stood out

when I first got to optimizely and I started talking to each one of the executives. Everyone was completely consistent about where the weak areas were as well as where the strengths were. And these are some executives that had only been on for a couple of weeks. So it must be a lot of you and Pete

just kind of going open kimono about what's really going on at the company and you get more brains on the table that way. Yeah. I think that's partly just our nature. You know, Pete and I both will be the first to admit that we don't know what we don't know. And we also are really willing to say and be honest about the mistakes we make. And, you know, we have a couple of small things we do that I think translate really well into that effort.

One is every meeting at Optimizing, the notes from that meeting, including the board meeting we had last week, get sent to the entire company. You know, we have a separate mailing list for that so you're not spammed with notes. but that small symbol of every meeting within the company being transparently sent and the notes getting sent, and everybody knowing that really has a lot of great unintended consequences. One is that any time there is a meeting, everybody in that room knows, hey, let's not talk about building a kingdom or politics. We know the notes from this meeting, regardless of what I want to or not, are going to get sent to the rest of the company. So you get that cross-pollination and serendipity of communication, but you also, you know, transparency is the best disinfectant. And I think that focus on not just the good, but the bad, that's really when transparency is tested,

is are we willing to talk about the biggest and worst problems we have? So I want to switch gears a little bit, and I want to talk about the future of optimizely. One of the things that got us really excited when we were evaluating the company was the demo that you had for the iOS and mobile, because, you know, as Tom talks about a lot internally, mobile is becoming the kind of the most important part of enterprise. and I don't think many enterprise customers have really optimized for mobile.

If you look at most of the enterprise, even the enterprise SaaS companies, their mobile products are crap. And so we saw what you guys were doing with Optimizley's iOS product. It was beautiful. It was easy to use. And I want to hear it from you. What was important about moving to mobile? And where do you see the opportunity?

Well, it starts with our vision. And our vision is to enable the world to turn data into action. And we think that's a long journey. The first step was website optimization and enabling you to take the data you have about your customers and your visitors on your website and turning that into action through AB testing. Mobile is the natural next step on that journey.

And we see that not only because that's where the people are, but because there's such an opportunity with limited screen real estate to make the right choices about what you show and you don't show in a mobile app. And I think that's really an opportunity where AB testing can play a huge role. Because on a website, you can put a bunch of stuff, and some of it may be effective, some of it may not.

When a mobile device, when people have limited amount of time, limited attention, and unlimited screen real estate, you're able to actually, those hard choices become much clearer through optimization. And I think related to that, the stakes of your mobile application succeeding, it's a higher stakes game in that if you don't have a good response

in the first few days, your app falls below the fold, and then you may never make up that traction. So getting it right the first time, being able to test, being able to make changes on the fly, you know, as you enable, once the app's even out in the market, like, that's essential, especially in that first critical week or two of the launch.

Yeah, and, you know, on mobile and an iOS in particular, a couple of things that we've built into our product that I think are much, much bigger solutions and pains that people have are the ability to slowly roll out new functionality. That's another, something on the web, you can quickly add something, quickly remove things, but on a mobile app, you know, you break your app, you're going to get a bunch of one-star reviews, and people are going to, you know, really give you a hard time. So that slow roll-out, and slow deploy, I think, is a key feature on our mobile iOS optimization product, and also the ability to actually make changes without going through the app review process.

And that, the ability to send different configuration data through optimise the tweak and change things directly in our interface, and those go live immediately in your app, I think are two huge solutions to problems people have had in mobile development. I don't think that most people that are winning in mobile, if I think about all the gaming, all of the applications like Evernote and Dropbox and Box, they've all gone freemium. And freemium, just by its nature, is I'm presented with a, you know, do I pull out my credit card here or there? It just stands to reason that this is going to be such a big business on mobile as companies try to figure out, you know, how do I turn my one to five percent of buyers

out of my 90-ish percent of users? And moving that by a point or two points, is very material. And I think that feels like a big promise for optimizely on mobile. Yeah, I agree completely. You know, as a firm, we've looked at so many different enterprise companies. And we see like this whole new cast of companies that were born from the Internet. And I'm thinking of companies like Zendesk, like Aptio. There are companies that are SaaS companies that are performing a completely different function.

oftentimes these functions are born of new technology. Optimizing feels like one of these new franchises. And it hasn't shaken out yet. Like I believe that, you know, somebody asked me the other day, hey, what's going to happen with the old incumbents? You know, doesn't there seem like a lot of SaaS companies out there? And I think there's going to be consolidation. Like I think the same consolidation that happened where IBM and Oracle and even Salesforce

and others were buying these companies. and then they ended becoming these, you know, hundreds of billions of dollars or tens of billions of dollar behemoths, I think that consolidation is going to happen again. But, you know, if you think about like BEA and PeopleSoft and all of those acquisitions that happened, those were standalone independent franchises.

They were bought for multi-billion dollars in stock by these other companies. I think the enterprise wants one throat to choke in one sense, but they also, with the departmentalization of IT, maybe each department wants to have, you know, three or four companies. So there won't be as much consolidation. Just to want to get your guys' thoughts on, you know, how do you see these new, let's say there's 50 new franchises. I'm certain that Optimizely is one of them. How do you see it consolidating? Do you see it kind of staying the course? And, you know, because of departmentalization,

like each department will have five or six vendors and that'll be cool. Well, you know, the way we think about it is we want to be best of breed at what we do. And so we've been initially very narrowly focused on website A-B testing. Now that we've become the number one most adopted solution in that field, we've moved to mobile optimization, and we will go beyond that. And the reason why we think that approach will work is because we also foster an ecosystem. We know that it's not just a standalone product, and that's not all it's going to take to make somebody successful. We really need a product that integrates seamlessly with every other products you use. And the way I see the next few years playing out is that in markets where

there is a best-of-breed product that is a delightful product that practitioners and departments will decide to use and evangelize for, as long as those products integrate seamlessly with the Zen desks, with the other products that they're using, then it's inevitable you'll find a best-of-breed product in many of these categories. They work well together, and I think that's a choice that business is going to bet on versus a frank and suite of products that have been glued together through acquisition that don't really work well together. So it's kind of like if you think about the SaaS world, because of the preponderance of APIs that make this data sharing very simple between applications, there may not be a need to have that same salesperson sell a bag of tricks like they had in the past.

Yeah, and I think, frankly, if you look at what an Oracle salesperson or IBM salesperson is trying to keep in their head, like all these different products and ways of bringing together the professional services usually to integrate them, it's going to be much simpler if you have some best-in-breed companies who come and they do one thing really well but as you're talking about they have an open platform and I'm also excited because I see increasingly a way of gluing in a good way these different point solutions together in a way that even non-programmers can do it so Zapier or if this then that are two companies where they're allowing non-programmers to plug the output of one company or system into the input of another and so ultimately I think a lot of what used to be kind of expensive, long-term system integration work, and now that can be done by non-programmers, just the way non-programmers are using tools like Optimizing.

Well, this has been a great conversation. I want to thank Dan and Tom for joining me. Really looking forward to building Optimizely into another one of these giant enterprise software franchises.