CP3: Enhancing Efficiency of Network Operations with Qwilt – Podcast with Dan Sahar
Welcome to the third episode of our podcast!
You can download the podcast to your computer or listen to it here on the blog. Click here to subscribe in iTunes.
Martin: Hi, today we are having a very interesting interview with Dan from Qwilt, which is mainly about video and network operators. Hi Dan. Who are you, and what do you do?
Dan: Hi Martin, My name is Dan Sahar. I’m the co-founder and VP of Product Marketing for Qwilt. So Qwilt is a startup selling infrastructure products for network operators. We essentially help them make their network adapt better to online video. One thing that pretty much everywhere around the world you can see today is that consumption of television is fastly moving into online video formats. So companies like Netflix, YouTube, Amazon, Hulu are definitely leading the way and the consumption pattern is headed that way. What we’re enabling is, essentially, a better experience, making that video experience better. We’re helping network operators optimize their networks in order to allow their consumers, their subscribers, watch as much videos as then can at the highest quality. That’s what we do.
Martin: Cool stuff, Dan. So, what is your background and how did you come up with those kinds of ideas? I guess maybe you have some kind of background with network operating, or video, or something like that.
Dan: Yes. Both myself and Alon, our CEO and Co-founder, we are enginners in background. So we went to university about the same time and one of the first companies we started to work with, we ended up working in the same company. This was known back then as Deep Bridge, later on it became, it was actually a company that was known as Siemens. Later on it became Nokia Siemens.
So, we were in companies that were in network infrastructure for pretty much most of our career. So, our roads parted on afterwards I wnet to work for a start-up company that was later acquired by Juniper. Juniper is one of the largest equipment manufacturers for network operators and Alon went to work for a different company, also a start-up company called P-Cube that also got acquired by a giant in the networking space called Cisco, that most people know of, and he spent most of his career in the networking space as well.
So we have both development backgrounds, creating products that are tailored to that segment, and also somewhat sales experience into that market as well. I think I started my career as an engineer and went up the ranks in engineering. But, I think, after several years I moved on to doing more product and product marketing activities and that’s where I am right now, but at core I’m an engineer.
Martin: And Dan, when did you decide to start Qwilt, and what was it like in the first one or two months or so? Did you leave your job before and then start a company?
Dan: So Alon and I, we were squash partners for many years. And what we were doing, we used to play like once a week, and after every game we talked about the industry, what we want to do as individuals. I think at some point we realized that the place that we were at, at the time, is hired employees- we could do more than that and we realize that we want to break out on our own.
We started looking at a few directions of products, markets that we can address. I think we had three solid ideas. I think that we ended up selecting this one because we felt that we had a very good insight into the customer segment and into the problem area. And at some point we dove into the water and just, like, decided to cut our ties with our previous employers and just go at it on our own. And I think that’s the basic advice to young entreprenuers. Like ultimately, there’s like a million ideas that are out there. But, the first thing that really dictates whether you do it or not is if you have the balls to, basically, break loose and just go at it on your own. That’s like, the first, I would say, cut off point between if you’re a real entrepreneur or just, like, a dreamer.
Martin: Yes. And when you started, how much money did you invest in the MVP? How long did it take you and how did you test this assumptions with potentially one of your contacts, or so?
Dan: So, the product that we’re selling is the system, it’s the infrastructure product. Typically these types of products are a lot more resource intensive then a web application and they require many more man years, a lot more investment, and a lot more time. So the MVP, the mechanisms are very similar to a web app, however the time that it takes you to get there is longer and the process that you undergo is a little bit different. It doesn’t matter if it’s for network operators, for enterprise customers, I think you have to do a lot of customer discovery in parallel. Building the MVP will take time.
So we had to make be like a certain functionality in the product that we had very quickly, probably in less than a month. But it wasn’t the real product. We could test some assumptions but, definitely, it didn’t produce the value that the product creates. So building the real MVP took over a year and a half, I think. So, I think products that are tailored for enterprise you typically build them, fund them in a slightly different way than you do for your application. So you typically have to raise larger funds initially. The investors have to kind of like take more of a gamble, basically, on inputs that you get from potential customers and partners. So that’s what we did a lot in the early days. We went out and touched the market as much as we can. We talked to probably over a hundred network operators around the world and got feedback from various people inside those organizations as to, like “Is this a problem that they think is relevant? If it is, what are guidelines that they would have for a product that addresses this problem?”
Different people have different thougths. You have to do a lot of listening, I think, at this stage. And, basically, have very fine tuned ears for every piece of data that you gather and try to compose all those data pieces and compile them together. You could be wrong, but hopefully you’re able to really distil the main points and build that viable product. And we think that we did. We think that we managed to really find what the main pain points are, and solve them with a product.
I thnk that one thing we were fortunate in doing is we built an extremely solid MVP. It was an infrastructure product, one of the most robust products I’ve ever seen on their first release, which was very fortunate for us because ultimately the inftrastucture that we’re dealing with is very mission critical. Think of ICT Deutsche Telecom, if suddenly your customers do not have internet, it’s very bad and you would call them up and freak out, maybe, that you can’t access the internet. It’s that sensitive. We designed a product that has to be up all the time and cannot fail and out engineering team did a great job of building that product.
Martin: Great. Dan. How did you find and approach those 100-200 network operators?
Dan: So, I think one of the things that you have to have is some, I think that one of the VCs calls it, unfair advantage. When you’re going after a certain market, you have to know something that other people do not, or will have a hard time finding out. So one of the things that we had going for us, we had a fairly good network of customers that we knew from our past. And that’s how we started. So, you open up your address book and you reach out to everybody that’s out there and you talk to everybody that you can. And you’re friendly. If you approach this with a good vibe, we found out that people are very eager to help you. So even if they’re not the right ones to help you out, they’ll say: “Hey you know what? I know this guy and this other operator. I hear he’s looking for things like you’re describing. You should go talk to him.” And I think there is, espeicially in Silicon Valley, there’s a lot of that pay-it-forward mentality which is very, I think, welcoming for entreprenuers.
So we had a network, but within twelve months our network of connections grew substantially. And, to this day, it’s something that always expands. You start off with maybe, like, one or two data points and then all of a sudden within a few years you have a ton of data points and a ton of sensors spread throughout the world and the picture becomes clearer as time goes by.
BUSINESS MODEL OF QWILT
Martin: Let’s talk about the business model, Dan. So, when I look at Qwilt, what I see are three major things. So one thing is you’re helping network operators decrease their network costs. Second, you help them understand the video consumption. And lastly, you help them, potentially to earn them some money with video streaming by offering extra services. Is there anything with what you started, for example, maybe of those value propositions you said: “Okay, this is my entry model, so to speak?”
Dan: So, in our case, we didn’t really pivot. We had, I can show you probably presentations from five years ago, they look like, maybe they’re less refined then what they have today, but the idea is very similar to what they’re doing and those three things that you said are very aligned with what we had back then. So that was the order of things.
So ultimately, I think what we are providing is a better way of doing something that you already do if you’re a network operator. So ultimately, your consumers will want to watch video and it’s a matter of how well or how efficiently you’re able to provide that service to them. And I can give you examples in other industries. You take the storage industry, so there’s been a way of doing storage for probably over a decade of using hard drives for that purpose. And then all of a sudden a new technology came called Flash, and is it that much different? No, it’s a more efficient way of doing things that you are doing for many years and I think we found a way that is specific for video that makes the network a lot more efficient and ultimately brings the price of bringing that content down substantially. And, lastly, it improves the quality of experience to the consumers which is, some people find it hard to believe, but that is exactly what the network operators really want to do. They want to help you, the consumer, have the best experience that you can and it doesn’t matter where you are going to watch it. If it’s going to be over your regular television or on an iPhone or a tablet, they wanted you to have the best experience you can. We help them in doing that. That has not changed from day one.
Martin: So, normally when you are an entrepreneur, you are trying to identify a business opportunity and then exploit it. This is what you’ve done perfectly. What is your opinion on why network operators before were not trying to solve this issue themselves?
Dan: So, I think that during the time that the company has been running a lot of things have been changing in real time. So one of the things was video back in 2009 and 2010 was still in its infancy. It still is, by the way. What I mean by video is I mean online video. So, at the time you take the largest, long form video provider in the world today, which is Netflix. They, I think at the time had less than 10 million subscribers that were consuming their content online. And they were only in the U.S. at the time. And today Netflix, as one example, they operate on five continents. But the consumption was different. Completely different than it is today. People were still doing a lot of peer to peer downloads. So there was really no incentive for the operator to change their way because there was no demand. Now there is a lot of demand. So, that’s on the consumer side, what has changed.
The second thing is that sometimest there is technology innovation that simply wasn’t possible a few years ago and then all of a sudden it is possible. So, I gave the example of Flash and storage. In our case, one of the main differences was that software became huge. As market research says software is eating the world, it’s eating the world in networking as well. If you would go back ten years ago, there was only one way of doing things to build your own hardware and to have very fast A6 that were passing packets from side to side. But all of a sudden, you have commodities servers from HP or Dell or IBM and they use Intell chips and they’re very fast. And they achieve results that are almost as good as you can do with custom hardware, as long as you know how to build really smart software. Fortunately we have experts in software that can extract that capability from the commodities servers and basically find a new way of doing something that’s been done for many years.
So it’s, I would say, the combination of these two things. Both consumption changes on the consumer side, and technological advancement that enabled us to build a product that does things in a different way than the status quo.
Martin: Dan, when I’m looking at the revenue model. Can you tell us about how it is working? So is it just a SaaS model, or is it more that you’re trying to do some kind of cost reduction share? Or is it the network reduction? How does it work?
Dan: So we have two generic type of business models that we sell.
One is based on the capacity that we’re creating. So it’s kind of like a gigabyte capacity model. So, the more that we deliver, the more that we earn.
And then we also have a SaaS model, that’s the recurring revenue.
That really depends on the preference of the network operator, which model to go with, I would say the traditional model that network operators are used to procuring this type of infrastructure is the gigabyte per second model but more and more, you know, as cloud is slowly penetrating network operators as well we’re seeing a transititon to the SaaS model.
Martin: Isn’t there an incentive problem with having a revenue model of gigabyte per second? Because, if I understand your job correctly, you wanted to minimize the gigabyte per second for the video streaming, but on the other hand you are paid four gigabytes per second. On the one hand side you want decrease it, but on the other hand you would like, from a revenue perspective, to increase it.
Dan: So, basically the trade off that we’re creating in our product is that any byte that you’re delivering from our open caching solution is, because it’s in software, is going to be cheaper than having that same byte of data being transferred over the traditional routing infrastructure. So that is the incentive. So, the more that we serve, the more lucrative it is for the network operator. And because video is growing, exponentially, and it’s showing no signs of stopping or slowing down. It’s growing, in some countries, like 100 percent year over year, we are able to serve more and more content and basically improve the efficiencies that we’re getting inside a network.
Martin: Cool. When I’m looking at the sales cycle of the network operators, can you elaborate on how does it work, how long does it take, who are the budget owners, and what types of people are you approaching to make a deal?
Dan: So, this is, I would say, those of you that are familiar with enterprise sales, it’s very similar to enterprise sales cycles. It’s longer with the consumer, they can take over 12 months in some cases. And you have to interact with a lot of people in the network operator organization.
Budget owners are typically the guys that are running the network operations team. But a lot of times you have to have influencers on board to safe decisions. So, people like the CTO organizations, people that are in charge of the content side, in some cases, network architecture. It really depends on the operator, but the sales cycle and processes, you have to really understand the organization that you’re selling to and craft your approach accordigly. It’s very similar to how companies that sell to enterprise craft their approach.
ADVICE TO ENTREPRENEURS FROM DAN SAHAR
Martin: Cool. I guess you have learned so much over the last years, Dan. Can you share some of your major learnings and insights and tips and advice for first time entreprenuers?
Dan: Okay. I think when you start off, one of the things that you have to bear in the back of your mind is that everything that you build, like there is that saying that the company’s DNA is formed in the first 90 days, and people, a lot of times, think that it’s about the cutture of the company, which is true in many cases, but it’s also about, you have to buld the foundations for the company at that time. It’s like building a building. If you have bad foundation, the building is not going to sustain for very long and foundation means that you try to opt for one of the best employees that you can get and try, as hard as you can, not to make any compromises. That’s a very common lesson. I think that it also means that you have to find the best invetors who ultimately become your partners on this journey. And if you have great investors I think it pays dividens throughout the lifetime of the company. It helps you in giving you the great strategic advice in every stage that you’re at.
There were points I think then we felt really good about ourselves, we met milestones and everything was going great, and then, you know our investors say: “Yes, that’s cool, but you should already be thinking two years out and plan on doing this initiative.” And we say: “Oh, we really didn’t think of this.” They always challenged us.
That is one thing that I think Qwilt as a company, we were extremely fortunate to have some of the best minds, I think, on the VC side still on our board and really helping us out on a daily basis. And I said, it starts with strategic advice, but it also, it helps you with recruiting the right people, it helps with giving the advice, making the right introductions to partners that you may need as time goes by. And I think the fundamentals to build the right foundation are crucial, because everything is going to be related to that as time goes on. It’s easier said than done. Obviously, everybody wants to have the best VCs but it’s not that easy to get them.
Martin: It’s true. Dan, how did you attract the first employees?
Dan: So definitely like the first, I would say, five to ten employees we knew from past lives. And I think that also helps when you build a certain company, that you see this often, that there’s this group of people that go together from company to company, and we had that core group. Then we built upon that core group with secondary circles of people, but most of them were from refferals of employees or people that we knew in secondary circles of ours, I think, probably over 80 percent of our first employees were people that we knew or one of our employees.
Martin: And can you still remember what have been the major three of four properties, or or skills, or attitudes whatsoever of the person, where you said: “Okay, this is the right skill set which we want to have for our company. And this is what we don’t want.”?
Dan: Well, probably the first people that we hired were engineers. And I think that’s fairly common. We were looking for people that were top 10 percent in terms of their skill set in software engineer. I think, also, we wanted people that were able to, we felt, that could really work within a start up environment, with all that – with tiny office, there’s going to be a little mess on the IT side, maybe, in the beginning, there’s a lot of uncertainty. That is, like, one thing that many people don’t realize, but I think engineers, at some point, realize that you can be like a 10x engineer and come into a product that is already, there is foundation. You come into Facebook or Google, there is foundation that other people already put in place. Even if you’re great, there’s only so much that you can do and you’re confined within those foundations that were already put in place in the past. So it makes your problem solving a lot easier. There’s less moving parts.
And when you’re a first engineer in a company, there’s no foundation. You have to, take for example, a logging system. Which log are you going to use when you write your code? When you come into a mature company, you know what that is. When you are first you have to make that decision on your own and there’s a million other decision that you also have to make. I think it’s super exciting for an engineer because you actually influence way more things, but at the same time it’s much more challenging than you would in an established company. Not every engineer is suited for that path of basically coming into an area where there’s nothing defined and you have to define everything from scratch.
Martin: Yes, but I guess this applies to every type of function, like marketing, sales?
Dan: Of course, yes.
Martin: Great. Dan, thank you so much for sharing your insights for Qwilt, and I wish you all the best and success for your company.
Dan: Thank you. Great talking to you, Martin.
THANKS FOR LISTENING!
Thanks so much for joining our third podcast episode!
Have some feedback you’d like to share? Leave a note in the comment section below! If you enjoyed this episode, please share it using the social media buttons you see at the bottom of the post.
Also, please leave an honest review for The Cleverism Podcast on iTunes or on SoundCloud. Ratings and reviews are extremely helpful and greatly appreciated! They do matter in the rankings of the show, and we read each and every one of them.
Special thanks to Dan for joining me this week. Until next time!
In San Francisco (CA), we meet Co-Founder & CEO of Quantcast, Konrad Feldman. Konrad talks about his …
In this article, we will look at 1) Angry Birds – an introduction, 2) how Angry Birds got so …