Sachin Agarwal: We actually have another fireside chat coming up next, it is Nicolas, I promised him I wouldn't try to do his last name, who is the CEO of Algolia joining us on stage right now. So Nicolas welcome to Empower.
SA: Good to have you, please have a seat.
Nicolas: Good. Hey. How are you doing?
SA: It's good to have you here thanks for coming and making time.
SA: Well, let's start off, I guess, with the big question, like the story of Algolia, how you guys got started? 'Cause it's... We were talking in the back and it's actually fascinating.
Nicolas: Indeed, we are about five years old today, and we started...
Nicolas: Alright so starting again. Yeah, we are about five years old today, and actually when we started, we started with a different product, so today if you don't know us yet, we are a search API. So you probably chatted a lot about API today, but we would help websites apps a way to deliver a great search experience, like people like Medium or Twitch are using us. But when we started five years ago, we actually started with a very different product search yes, but on mobile. So our idea was to develop an SDK that developers could use directly in their apps on mobile locally on the device. It was not possible yet, we saw a lot of unanswered questions from app developers on Stack Overflow for example, and we thought we had big market there, interesting market with the experience to do that, it didn't turn out that well, the market was not there and so we decided to pivot to what we have today which is a search API as a service which means we pivoted from local to the device to a SaaS-based service.
SA: Yeah and one of the things that's interesting about that is that if you had started off there you probably wouldn't have had the company you had now, so you had to go through that pivot.
Nicolas: That's right. Had we started SaaS, then we would have probably used open-source elastic search directly. We would then have done better. The thing is that the constraints we had on mobile first you only develop something different to kind of reinvent search for these constraints and when we switch to mobile, when we pivoted, this mobile experience became our differentiation. We realized that what we'd done on mobile was such a perfect fit for most of the user facing use cases consumer grade search all over the web and that's probably one of the main reasons for our success today.
SA: Excellent and pivots can be hard. One of the things we talked about is how is the culture of the company able to successfully guide you through that pivot?
Nicolas: So at that time, it was still the two of us, so we are close to 150 today, we were still two people at the time so I wouldn't say that at that time yet that culture was overwhelming or like had a big impact, but it was already the... We already knew what we wanted to build in terms of culture, we had that notion of ownership. So the idea for us was to be the company where everyone would be owners empowered to make decisions at that level so that this kind of pivot or these kind of key questions they could address them every day. And I'm very happy to dive in [chuckle] [04:00] ____, yeah.
SA: Yeah let's go a little bit into that, so you had this culture of ownership. You've also had... What are the other parts of that culture with other tenants that you have, the key values in the company.
Nicolas: So today of course the culture evolved a lot along the years, we formalized it and today what we have built is a set of five core values that define who we are. And on the other side, the way we work. The way we work should change every single day as the company scales. Who we are should not change much. And so the way we defined that are five core values that wholly are sustaining that ownership culture that our grit, so grit [04:49] ____ perseverance. [04:50] ____ getting out of your comfort zone and if you fail that's okay, you learn from it, you step up again. Trust, if you want to empower people to make their decisions, you need to trust them with as much data as you can. At Algolia everything, pretty much everything is transparent. Everyone knows how much cash we have in bank. I share my investors' update with the team, even the compensations of the team are transparent. Everyone in the company knows how much I make.
Nicolas: Care is our third value. Could be on many [05:25] ____, the idea here is really [05:27] ____... We want people to care about customers but also like about each other. One of the thing I'm the proudest of actually is to see sales and engineers working together which is pretty rare as companies scale like maybe a quick anecdote on this one in we're soon going to change office in Paris. We have like close to 90 people there. We're going to move to much bigger office with several floors. The obvious way to do that was to put sales and engineering on two separate floors. We decided especially not to do that because that would have been creating culture of death. In one year, they would not speak to each other anymore. So that's really kind of very important in who we are.
Nicolas: The fourth value is probably the most unusual. It's candor. We try to be really candid with each other because we care. So if you want to help people grow, help people really grow professionally, you need to provide feedback to them 'cause it's the only way they are going to learn. And we do that because we care. So the idea is of course to be very nice in your feedback but not to hold it back. And the last one is, humility. That goes also with candor, you cannot be candid if you do not know your feedback can be wrong. But that goes also with our audience. We speak to developers all day long. I don't like very aggressive marketing for example, doesn't work. Being very direct and honest, and humble about what you do, what you don't do is what's really appreciated, what generates trust. So that's really who we are.
SA: Yeah. That's really interesting. And then, you've codified the sense of ownership with these five key values.
Nicolas: That's right.
SA: How is that reflected in the product?
Nicolas: So, the product is something much larger than purely the culture. But the culture is helping of course make better decisions. I'm trying to find an example here. One of the outcome of that culture is to have bottom-up new ideas. The team can come up with the ideas and speak up. We expect them to speak up when they have things to say or things to suggest. And one of these things that happened a couple of years ago was one of our PHP developer came to us and said, "Hey, there is this new awesome framework called Laravel that is very trendy in the PHP world. We should do an integration." "Sure. Go for it." And don't even have to ask, he did it. He also implemented a search of the documentation, did [08:26] ____ request. They were super happy, accepted it. And that became one of the biggest key generator we have because he created a very strong partnership with them. More than that, we directed product out of that because we saw that this documentation search was so useful for them. We created what we call the doc search where we power the search of many developer frameworks. We act like Stripe [08:56] ____, like so many projects. I think it's more like 50 projects now. All of that, came from that first initiative from one team member. So it's probably getting them to help us.
SA: So, enabling one developer to have that ownership, to make that decision had all these downstream effects that you never even imagined.
SA: That's fascinating. And then, there are obviously other parts of the organization more than just development. There's customer service. There's obviously you have a sales team now. There's other parts that... How else does ownership influence the way that you've intentionally built this company more than just making sure people aren't on segregated floors?
Nicolas: Yeah, so one of the key things here is that from the beginning, we wanted to have everyone customer-facing, speaking with our customers, with our users. And one of the way that impacted the company was actually customer support. We don't have any support team. The key thing here is that developers hate to speak to support people because they seem they know better. What we did here is that all support requests are handled by the tech team, by our developers. It's kind of like a win-win situation because for customers, they love it. They love to have access to the people who develop the product. And for us, that's so important that our developers get to speak to customers to see their pain, to see their needs, but also to feel the satisfaction of solving them. And it's a very good incentive to build a better product, to build a better documentation so they don't have so much report to do. It's a very good incentive and that really helped shape the company as a whole and not only... It was not only a small part like support. Actually, that's one of the implication when I was speaking about sales and engineers working together, it's one of these other impacts of that culture.
SA: Yeah. And so, the engineers do support. And so, obviously you're building new features. So, how do you balance having the same engineering team? Do you rotate? Does everyone does everything?
Nicolas: They develop [11:24] ____ in time because in the very beginning, every single person in the company was supposed to do support. Today, we get, scale that differently to make it efficient. So what happens today is that they would rotate. They have some slots where they're covering support. And then, the last thing we did was to actually have someone dispatching. So what happened is that, as we're scaling, the product was getting bigger, more and more features. So not all developers were able to answer all questions. So the people that are kind of on duty are going to answer all questions they can and simply dispatch, redirect the more expert questions to the team. And even if you are not on support, you are going to handle these questions. And that's kind of the priorities come from the care. You want to make sure the users are going to have the best experience ever, because that's so important. It's kind of like business where word of mouth, referrals are so key to your success, you want to make sure your customers are super happy. I don't know if some of you follow, have an NPS score, like Net Promoter Score, that's something we follow very closely. Making sure that our product community is really supportive and has a great experience with us.
SA: Yeah. And so just to, do one last question on that is, if you're not on the rotation, and you get dispatched a question, do you interrupt the new feature building to solve that first? Is that core value of care or is it handled a little different.
Nicolas: Yeah, it's not, except if it's urgent, but then you're going to be pinged anyway like on Slack or anything, some thing's going to happen. You're gonna be woken up. No, you are not expected to answer right away, but it's [13:16] ____ like hours, day maximum kind of timeframe. Then you are expected to answer. But these questions are not going to one individual developer, so that's usually going to a team, so in that team usually there is someone that's going to be available to answer. If it's questions on the specific [13:35] ____, I was speaking with PHP before for example, so if we have a customer that is implementing in PHP and has a complex question related to the language, it's going to be dispatched to a team that is expert in PHP. And that's going to be [13:54] ____ four people today. It's not like a one person.
SA: Wonderful. So we already mentioned you have a Paris office with 90 people. You have an office here in San Francisco, also Atlanta, New York, anywhere else?
Nicolas: Yes. Not yet.
SA: Not yet. Where else are you going?
Nicolas: Next is going to be London. And we're trying... We're thinking about opening an office to Australia in the coming months. Next season, not this month.
SA: Not this month, yeah sure. And then how do you do that? How do you open a new office? How do you make sure everyone is connected and up to date? 'Cause remote is hard. Remote offices are a little better but it's still difficult. How do you handle that communication? How do you make sure that the offices don't drift apart from each other?
Nicolas: It's tough. It's difficult, challenging. If I go back a beat in time actually, the challenge started when we opened the San Francisco office. So we are headquartered in San Francisco, but maybe I can tell, we started in France. When that happened, we were a bit lucky in a way because we participated in Y Combinator. After the batch of YC, we all went back to France. But most of our customers were here in the US, and especially in the Bay area. So we had to travel back and forth a lot. Traveling back and forth is a very good way to progressively learn the challenges of working at a very different time zone. When I was spending maybe a third of my time here, and when I was here, it was so difficult to communicate and it was so painful to see... To not know what's happening there. And that kind of creative and [15:38] ____, we got used to over communicate, to prefer [15:47] ____ async messages, written messages instead of just speaking together in the office. Many things like that we build up overtime, and when we actually open the office here, we're ready. Kind of ready, much more ready. The other thing that was key there was the language. So in Paris very early on, we chose to only speak English. Very difficult at the beginning but so critical to be able to attract foreign talents. And so I think we have 17 nationalities now in Paris.
Nicolas: We have team members from all over Europe, including now we have one from Israel. We have a few from the US. It became a very, [16:31] ____, it became a big advantage for us to attract talents. Because we are not looking at the French market, we are looking at the European kind of talent market for Paris. And of course speaking only English is so important so that the team in the US is kind of very well connected. From there, opening New York, opening Atlanta, was not a big deal, because we were so used to work at the distance, that it was not such a big deal. We're discovering new challenges with these offices nonetheless. Because they are smaller. They don't, some of them don't have... Like in New York, we don't have long tenured employees. So it's creating new challenges, that we're addressing as we go. But our history of having [17:20] ____ for so long, kind of makes us, more ready to tackle these challenges.
SA: Yeah, that makes sense. And as you open these offices and you hire more people, do you train them on your culture or you train to see if they fit into the culture during the interview?
Nicolas: Both. First thing is interview process. Culture is a big big part of the interview process. We're a bit crazy on the process. It's a pretty heavy process with a, like calls to filter at beginning. Then we've got like calls with the team, assignment, then we have people coming on site, either in San Francisco or Paris. Sometimes both. We would fly people to our office and then [18:04] ____ calls and all of that. Especially during the on site. Actually most of the on site is about culture. We have now evolved a bit and now [18:14] ____ often have a key culture value to test for. We've built a set of questions to get inspired like as guidelines, like ideas of how to test cultural values. But that's something we follow very closely that we make sure... We can be wrong, we can fail, but make sure, that at least get confident that the people are going to be a good cultural fit. And once they join, I think the most important is really how the rest of the team is going to show by doing, by example. I try to think with employees after a few weeks, once they have joined, one of the questions they're asked is, "Was there any surprise, good or bad?" And the very common answer is that what was surprising is that the company was the way you described it would be.
SA: You're honesty enabled to [19:17] ____.
Nicolas: Yeah, and that comes from the cultural and from people showing care, showing trust and all the values.
SA: So you know care, trust, are values that are relatively universal. Candor is a little bit different.
Nicolas: That's the most difficult one.
SA: Candor is a little bit more difficult. And especially there are cultural implications about candor and as you think about an Australian office or maybe an Asian office or an African office, some of those values may not translate as easily. How are you approaching that problem as you look to take over the world?
Nicolas: One step at a time. Maybe to start with I would recommend all of you to read that book "Radical Candor," it's pretty easy to find. It's super good and it's super clear, exactly. What we mean by candor is exactly what's there. It's stuff, already here, already in Paris. That's [20:16] ____ actually as we speak as we scale, we realize that that's one of the most difficult values to scale because you can show by example a little but that goes so far. So we are thinking about implementing trainings, how to give feedback and so on. And that's something we are starting to explore. If we look at very different cultures, it's even more difficult. So I don't think Australia is going to be such a problem but if we want to open an office in Japan next, the local culture is... Yeah, I've lived there for some time. Being candid is definitely not in the local culture. I don't know yet the answer. We are going to try to maintain these values and probably adapt and find the right balance with the local people. I don't know what it's going to be like, but we are definitely going to try and [21:12] ____ so we need to see in couple of year and see how it went.
SA: Yeah, I'll have you back in a couple years and I'll ask you that question then. Let's take the chance to open it up to the audience for some questions. Someone up here.
Audience Q1: So it sounds like your culture has enabled some pretty unconventional ways of operation within your company like sales talking to engineering, also the fact that your engineers are doing support. I'm really curious about the latter. How did you... What was that evolution of saying... Did you just come to your engineers and say, "You're doing support now"? How did you get buy-in, how did you get them to care about it? Or did they care about it because of your culture?
Nicolas: I think the latter. So what happened is, we didn't come one day to engineers and said, "Oh, now you are doing support." We started this way. Actually we started... We were going much further when we started. When we started every single person in the company was expected to support. And then we actually reduced that scope because quickly as we were growing communication was becoming more difficult. If a sales person for example was having to answer a technical question, quickly didn't make sense anymore. So we started to reduce the scope of the number of the people who would answer simple questions to engineers, especially for tech questions. Pricing, business questions are within directly to sales. And so it's always been the case, it's always been transparent. So candidates know that when they join. They are going to meet engineers of the team and they're going to discuss that. And so I think that's never been a surprise, never had to adapt to new thing. Part of the expectations... It's like English. Going back to that, that's super funny because in the very beginning that was difficult but then we started to apply that so consistently that French candidates, they all do every single interview in English including with French employees, so when they join it's expected for them, they're expected to speak only English. So it's never a question. Same thing for support.
SA: There's a question here in the middle.
Audience Q2: Hey. Seems that you guys have three main use cases with Algolia that is SaaS, media, and e-commerce, right?
Audience Q2: Wonder if you can describe how was your process of discovering these particular use cases because in the beginning you may have... [23:58] ____ set because of being so horizontal right?
Nicolas: Yes, but the product is always [24:05] ____ work... I wanna say not everywhere, but like in any consumer grade kind of search, when you have a user interacting with the service. In this, like segments, like these industries came progressively, when we started we were everywhere. E-commerce became obvious very early on because if you speak about search on the web that kind of like the default example you are going to have in mind. It was also a good start for us because it's very easy to prove the value, going to do a [24:37] ____ test, and if you win the [24:39] ____ test it's very easy to prove the value of the product. The next one was media and kind of discovered from our customer base, we started to see more and more media using us. It's between the video space like we had TiVo using us and then Periscope using us, and then we started to realize it was a very good fit for that category, and so we started to invest more in some features but also like messaging around media.
Nicolas: And SaaS is the last one we actually came, and what's interesting with SaaS it's a very different kind of use case. E-commerce, media is the big data so you don't have any privacy concerns or anything like that. SaaS, like the companies, they have to trust us with the data their customers trust them, so it's kind of a much bigger bet for them.
Nicolas: And so that's why in the beginning we didn't see that but possibly with those startups using us for SaaS products. And we were scaling with them because if they were successful, their need of search we are going to scale, too, which was awesome, and rapidly we realized that our fastest growing set of customers were SaaS players. It's not the biggest of all but in terms of upsell, upgrade that was the fastest growing. And so we started to concentrate there, starting to look at what were their needs like compliance things, like SOC 2, all the certifications we're going to... All have to go. I guess during the...
SA: We had a talk on SOC 2 earlier this morning, yeah?
Nicolas: Yeah. [chuckle] You know about that. But also as we speak, for example right now we're developing a new... Let's call that a feature but something in the product dedicated to this segment, like their needs, like the biggest pain in SaaS is basically scaling. If you are scaling fast at some point you are not going to be able to use anyone's server, then you are going to have like tens of servers and hundreds of servers and user base is going to grow, you are going to have customers everywhere around the world, then you also have the privacy issue, you need to host the data differently, all that stuff.
Nicolas: And so that's exactly the next challenge we are going to address with our product for these players. So progressively it came like that, well, what is the next one? Kind of like what I'm seeking now, looking at our data, who our customers today, okay, what is our next segment where we can maybe invest a bit more to look at their need. Yeah, a couple of ideas there but a bit early to say.
SA: Are you ready for GDPR?
Nicolas: We are going to be ready for GDPR.
SA: We all are, right? We will be there by May 2018.
Nicolas: That's the biggest like if you're... I mean any company, not only SaaS, GDPR is kind of a pain but we have to do it, we'll be ready... It's actually a bit more complex than SOC 2 [27:49] ____. SOC 2 is stuff that is just a process. GDPR involves engineering time just to make sure everything is proper.
SA: Do you think you have an advantage there just because you have that French base?
Nicolas: Yeah, that's a big advantage. GDPR of course but also like all the psychological aspect. So it's not French or not, it's probably like data centers. Where are we going to host the data? To the web, I don't know, like maybe even a hundred servers in 50 data centers around the world. So one of the biggest advantage here is that if we have a customer in Germany, it means there are like, that's kind of the market where they are super conscious about where the data is hosted, yes, we have servers in Germany. No problem, you can host your data there. And even if it's not part of any specific regulation, just for the psychology impact that's very useful.
Nicolas: And of course in terms of product and availability and multi-data center setup, multi-region setup, all of that of course is part of the product we offer.
SA: Cool. Well, I have one last question I'd love to ask you about is, as you scale the... Is it harder to scale the product across the hundred data centers or is it harder to scale the culture across all of your offices?
Nicolas: The culture.
SA: Why? 'Cause a hundred data centers, that's nuts, right?
Nicolas: It's like a bit more than 50 data centers.
SA: 50 data centers? Okay.
Nicolas: Yeah. But more than 1100 or 1200 servers. We have actually a very small infra team, it's not big, we only have like four people on duty, on call. The nice thing here is that because there's only four if something happens, you are going to be woken up. Well, that's a very, very good incentive to make sure... I mean you don't want to be woken up. And so you basically, like from day one, we've invested so much into building automization everywhere. Also, we use our, if one today we're currently from, if one server dies, nobody's woken up. There is enough redundancy that that could happen any day. That once too often to have someone to have to wake up and correct that, so you deal with that the next day. And that's because that team was able to build all that automization and are building more and more and more to deploy new servers, like near instant, and so on. It was progressive. It was growing as the number of customers were growing. That was finally not... It was super challenging, don't get me wrong. But it's something we are confident about.
Nicolas: Culture, as you evolve, the culture has to evolve. We have these core values but you cannot have the same, especially the way you work. You cannot work the same way when you are 10, 50, 150. You are going to see new challenges. Maybe to give you a very quick example, there were several stages in the culture. There was one time when we were maybe 30-35, where the culture failed a little. The culture was already super important for everyone, so everyone would speak about the culture. But then you would ask people, "Oh, why don't we do that?" "Oh, no, we cannot do that." Why? Because of the culture. Why? You started to realized that everything was mixed up in their mind. We were 30, so we were still flat, completely flat, no middle management. Many people started to concern, oh that's our culture, to be flat.
Nicolas: We're not going to scale to 1,000 people by staying flat. That's when we actually did that work that really, worked on formalizing these five key values and separating the way we work from who we are, and that was successful. But, I'm sure that next month, next year, we are going to have new challenges, and that's not frightening, but that's a bigger, in my mind at least, in my position, that's bigger challenges than [32:16] ____.
SA: Are you excited?
Nicolas: I am, definitely.
SA: Awesome. Well, Nicolas, thank you so much for your time. We really appreciate it. Thanks.
Nicolas: Thank you so much.