A Little Bit About Our Guest For This Episode
While reaching people for the “Let’s Talk Growth Series”, Ivan Kirigin was one of the first guys to reply and accept our invitation. I can’t really describe the way I felt when his response hit my inbox.
I’ve been tracking Ivan for a long time now but in case you don’t know him, he served as a Product Manager at DropBox (growing it 12x in a 2 year period) and as an Engineering Lead with Facebook’s commerce team, creating the first games dashboard and FB Credits payment flows.
Now Ivan is the CEO of YesGraph, a platform that helps your app grow by recommending exactly who a user should invite, allowing you to boost the performance of referral and invite flows in your app or website.
Here’s What Ivan Had To Share With Us
Hi everyone, I am Evie Samouris, a Growth Hacker at GrowthRocks. I have the opportunity to speak with Ivan Kirigin, founder of YesGraph. Previous, Ivan led growth at Dropbox and helped build Facebook Credits. Hi Ivan and thank you for taking the time to speak to us.
Thank you very much for having me!
Tell us a bit more about who Ivan Kirigin is? We know that you have served in several roles e.g. CTO, Product Manager, Software Engineer and you’re currently CEO at YesGraph, but which role is the one that you feel that fulfils your personality?
Well, I think I like running things. You know, if you have a large company versus a small startup, like running a part of a small company is a lot like having your own company.
I think there’s a lot of inflation of founders of title cause anyone can say “Hey, I’m a startup founder, I started my company 5 seconds ago’’ but I really like to get into the details and in terms of product, which I think depending on the role, it can be different things. I like running my startup for that reason, but I have worked at different larger companies and that feels pretty good too.
I like running my startup for that reason, but I have worked at different larger companies and that feels pretty good too.
I think one big difference for me was, going from a software engineering role to more product focus. A lot of people are really much better software engineers than I am. So, that transition was a pretty big deal for me. I went to grad school in robotics, I am really technical, but stopping coding was a big decision.
And that, I guess, has led you to be the CEO of YesGraph. Can you tell me more about what YesGraph does,and the referral software that you have build and the challenges that you have faced and what steps did you take to overcome them.
Sure, so what YesGraph does, is we help your app grow and we do that by a few different ways. One of them is by recommending who your user should invite. So we actually use a lot of machine learning and a lot of social graph analysis to be able to look at the user’s friends to say “Here are the five that you should invite’’.
That is useful for a friend-in flow, invite flow or referral program. So there is a lot of different ways you can integrate it to your product. We also have a sort of gun up in the stack, so we actually power the actual invite flow now, whether that is on mobile or the web. And so we are starting to cover more.
But really what YesGraph is all about, is about making apps more social. So in that sense, it is more like facebook connect, where we want to help apps understand their users and make it really easy to make their products great in social. And we soon had a lot of challenges.
I mean for one this is a pivot, so the company is a bit older and we switched to working on this just because we felt this is more compelling than what we started building which was more social referral recruiting. So away from recruiting and more towards this tool directed for this kind of social graph analysis.
Another issue is that it is really hard to hire. So you know various things like having remote developers and switching over to a local team, which you know it is pretty interesting to try to manage different structures of teams, and we are sort of in a startup boom right now, so finding good people is very challenging.
And you do that, and the good news is that you solve that with diligence, you really work hard on sort of presenting your vision for the future of the company, and making a great place to work, and you work through the process of recruiting. I will be happy to dive into the detail about any of these.
Please do. Could you give me an example of one of the challenges, and what steps you took to overcome them?
Sure, let’s talk about the pivot. Basically, we were working on a product that didn’t have enough traction, because I think we were building the wrong thing. So what we did was we had this idea “What if we took part of the product we had and sold that directly?”, and this is interesting because it’s like the first step for almost any company is to try to talk to potential users and see if you understand their problem and see if you can help them out.
I think a lot of people in growth focus on tactics, that are all about acquisition or you know, just getting more people to hear about you, in the traditional marketing sense. But really the kernel of a new company should be our new product should be about the users, should be understanding them and you do that by talking with them.
And the way we justify the pivot was by, having a lot of people interested in what we were doing, and saying “yes! i want that” and getting pre-orders essentially, or what we were planning on holding. So before we wrote a line of code we were talking to thousand of potential customers and getting a few to commit to what we wanted to build.
Well, that’s really great and that’s why it’s become one of the greatest referral software programs. Could you give me an example of a company, like a case study, that has used YesGraph and the results that have come out of it?
Sure, so we are across the web and on mobile, so there are lot different kinds of Apps that are on us. One of my favourite Apps, because of how little I have to do to make the product work, is called Paribus and the way Paribus works is that they read your e-mail and they see what products you bought, and then they check if the products have gone down in price, and automatically ask for that difference back. So you essentially author email, pair with Paribus, so connect your email and it works! That is all you need to do. Then you start saving money and they take a cut. They take a smaller cut if you invite your friends. So they have this referral program.
So you essentially author email, pair with Paribus, so connect your email and it works! That is all you need to do. Then you start saving money and they take a cut. They take a smaller cut if you invite your friends. So they have this referral program.
So they added YesGraph, and because they have their email contacts by the product, every user has other contacts in there. What YesGraph can do is recommend who they should invite. So they added YesGraph to the referral program, and I would say the biggest example of how we could add value, is that they shifted focus.
Originally they focused on social sharing, and had prompts to go to facebook, and to twitter and then they switched this to be front center our recommendations, with back to top context that we recommend that auto selects, all you need to do is press “go” and then they are invited. And they saw a 10x increase in the performance of the referral program, which is a pretty big deal. Getting more invites sent is important, but this is where people go on the wrong track, as far as growth hacking, because they well say “Oh! We need more invites”, let’s just make it push easy as to invite everyone, like thousands, and there are problems with that.
Getting more invites sent is important, but this is where people go on the wrong track, as far as growth hacking, because they well say “Oh! We need more invites”, let’s just make it push easy as to invite everyone, like thousands, and there are problems with that.
The problem is that the signal to noise goes way down, and it results getting marked as spam, and your invites stop being delivered because they are now in a spam folder. And so these big problems of that kind of certain growth hacking of “invite all” is kind of a dark pattern.
And so YesGraph helps you have incredibly high-quality user experience which does not get you marked as spam, but also has a great deal of performance improvement, when it comes to the recommendations that we come up with. And so it is an interesting example, because the app you know sort of is a really well-structured referral program, but even with that having quality recommendations on top of it, makes it a much smoother user experience and a much higher quality performance.
Exactly and the right audience as well.Thank you so much for that example, actually, I would like to know, a little bit about you and your background story. Your first job, milestones and basically what sparked your interest on growth.
I got started out of grad school, like i mentioned at the robotics. My first job, my title is a scientist actually, so basically i was a software engineer working at a research project. I focused on computer vision, sort of the perception side of robotics and that was lot of fun, did a lot of software engineering. But then i wanted to do my first Startup and we got accepted into the YC Winter08. So like eight years ago, the batch a long time ago, so that startup was called Tipjoy and the way it worked was that, you can support things you love with micropayments and they were social.
So like eight years ago, the batch a long time ago, so that startup was called Tipjoy and the way it worked was that, you can support things you love with micropayments and they were social. So you could do things like tweet out of payment, the sort of like you could say “d-username send message” and send somebody a direct message.
You could do p at 2 dollars, and make a payment through twitter.The mentioned traction was ok, but the problem really was that we never really figured out the marketplace we were building, as far as merchants and buyers, it was kind of more complicated than just an app somebody signed up for. That’s a big deal and we really didn’t figure it out. We didn’t really know how to really drive traction. We understand our product, about making experiments and all that, but it wasn’t really organised in a good way.
We didn’t really know how to really drive traction. We understand our product, about making experiments and all that, but it wasn’t really organised in a good way.
So my next job after that was at facebook. And the cool thing about facebook is that they really have figured it out. They really understand how growth works. On the internal of the facebook, i was working on the credits team and folks from the growth team were also working on growing the ads, like the number of merchants that are posting ads, brands posting ads to facebook.
And the most fascinating aspect of facebook was that they had the same tactics applied to varied different products. So if you think facebook the social network vs the facebook the ad platform, those users are really different. One of them is a business, they spend money, they have some performance to care about, they need to see a dashboard and configure things. The other is a product where you share your cat photos with family right?
It is not the same thing at all. The same tactics surround looking at retention, and driving repeat engagement, looking metrics in that driving, making experiments. That is where I learned how growth is important.
Then when i started talking with folks at dropbox, it was not like they had some job posted about someone wanting to learn growth. It was more like, I loved the program, i knew the folks there, and i wanted to see if there was an opportunity.
So when we started talking, it was more about maybe this kind of role focused on growth, it would be great. So, then we started to install that same process at facebook, but on dropbox. You know, they had a really high-quality product, everyone loved it and a lot of different channels that they could get traction with, a lot of complex user onboarding issues.
And so working on things like email, a/b testing scepticisms, and different kind of analytics questions getting answered, and doing a bunch of experiments on their social flows. That is where I spent my time. So they had a good referral program, we made it better.
We had some sharing channels, we added more. And we all agreed with the company as far as my understanding of how to run growth and working with some really smart people. It was a really good experience.
And so coming to YesGraph my second startup after Dropbox really wanted to apply all of these lessons. And the good news is we could apply them directly in the product. The product itself is essentially the best practises in growth that you should use when it comes to high quality invite flow and how to think about metrics and how to drive things.
So if you look at facebook or Linkedin, they have these people you may know (recommendations), the thing about that is that it drives a lot of performance, this is where they recommend to send invites.
They drive a lot of performance but there is an incredible amount of work to do. So the cool idea with YesGraph is that you can sort of let companies jump ahead in line, where they could get this high-quality performance, normally where huge growth teams are required to build it. But we do that for a service, so anyone can do it even if they are two co-founders who have just launched and app, they can do it.
While working for Dropbox, you helped the company grow 12x in 2 years. What in your opinion is the best way to model growth, and what do you consider as a prerequisite before doing so?
I guess a lot of people, like I was saying before, focus on top of luck and I think that is a mistake. I think the very, very first you should do is make sure that people like your product.
So you know what our sales calls are trying to get people to understand with YesGraph and I don’t just talk about virality, my first question is “How are things going?”, “Do people like your product?”, “Do you see retention” “do people come and then come back?”, because if you try to grow, if you try to get top of acquisition, but your retention is poor, then it is almost impossible because all the users have been gone versus a retained base;
However, a small percentage there was, that you could do a proper analysis, some people are sticking around. Growing that retain base is tenable. It could actually work. Then you grow from there, and all the different tactics about how to do it. And so I think that is the first step, make sure it is a really really high-quality product, in order to get people to stick around.
It is a bit more complicated than that, depending on the application. For example, some apps are social. Some involve your friends or your co-workers, then your experience on the app is going to be based upon the people who have signed up for the app already. So it is interesting because top of acquisition might be seen as just a way to grow but actually, it is a matter of driving deep engagement with your users, so that they have other friends there.
And an example of that is Google+, where basically we have the basic features of facebook but no one is there. It is terrible product experience. User experience is impacted by the density of users. Other examples include Uber, where you have the density of drivers and motor drivers that are out there, are the best experience you are gonna have, because there is this deeper relationship between growth and user experience. So there are different ways you could focus depending on your application.
To speak to a consumer that is interested in your program, how do you spark their interest, from the beginning through your sales calls? Do you talk to them about the product and then make it a more of a friendly conversation? How, from a consumer point of view, would sit down and analyse your product, and talk about your product without receiving anything in return. I feel that is more of an obstacle right?
We actually try, this is not perfectly known but a lot of B2B applications, we don’t think that talking someone over the phone that is the first experience they have.
Actually, the first experience is probably reading one of our blog posts, because we have some pretty active content marketing. So if we have started to build trust in people, to say we know what we are talking about as far as growth, then once we get on the phone, and we hear this from a lot of people, we do get a lot of sales calls, that they heard us through a blog or through a friend.
That kind of social proof is really important. So for example, we write a couple of posts like “The top three metrics of any viral flow” and the actual top three metrics are that you have some application users actively participating. They sent a certain number of invites. Those invites convert at a certain rate. So you have three simple metrics to look at, three simple things to understand.
Let’s say you could apply to Instagram, to say “oh, Instagram everybody posts your photos on facebook or twitter, so almost every active user is actively sending out a link where someone could come in and sign up for Instagram. Those links are really engaging. So they have a high conversion rate because there is really high-quality content because of the photos that are there.
So you can model growth based on that. But the point is that when we talk to people, after they’ve read this blog post, we’ve already incurred on how to add value with YesGraph, to say “ok, we care a lot about the high-quality invite flow, that somebody wants to engage with, sent more invites, higher quality conversion rate, so I understand how your invite flow and your contact recommendations are going to actually help me.
I think this is why companies are so focused on content marketing because the kind of users that are coming through that are prime in order to be in a good position.
Another example of something we have done is we’ve actually just launched a new class, which is a class focused on growth. So we have taken some of our low content and added a bunch of video lectures and some exercises, and it is like learn at yesgraph.com. and it is actually a paid class, it is not free.
We want people to come in and take the class and learn how to do a good job. We are running growth with a lot of companies, and there are two sides to this one. One is that not all of our knowledge about growth can be productized, into the YesGraph product services, and so having a place of premium content, where you can come in and buy a class is really good to help spread our knowledge.
The other side to it is that if you are paying for a growth course, you are probably in a very good position to use our product, which is you know a sales product, it is like a way of qualifying the leads that come in. It’s all experiments on our end too, so we are thinking about adding different channels, and paid content marketing channel is a recent thing that we just announced.
So there is a theory behind it, now as far as qualifying leads, there is also a lot of really good metrics that can come out of it, like every step getting nurtured. It’s a class, it can be purchased right away, structurally it is really different than a SaaS app, normally for a lot of companies they spend a great deal of money upfront to get recurring revenue.
They will still get as in a sales team, what are the tactics they try. They spend a lot of money upfront for a subscriber that will pay off over time. And a nice thing about a class, that is paid, is that it inverts that so that one of the first experiences someone has, kind of pays for the marketing to get that person to have that first experience and then it’s much more efficient to acquire new users.
Structurally this is a really good way of doing content marketing. What we are working on sales as an experiment, but I think the structure is really good since you are thinking in ways a growth hacker should be doing. Should be understanding how they can get high leverage in terms of the kinds of work that they could do.
So aiming for this class we are going to spend almost the revenue on ads, to get a larger audience in the class, and then essentially will be our lead funnel. And the scale of the ads will be much much higher, than what we were able to do.
Plus on top of that, we are being very transparent, where we are recording how we are running ads and making that part of the class. Because the class is about growth, so paid acquisition campaigns examples are great to learn about. So I am really excited about that.
That is really great to hear because Growthrocks is actually launching their own Growth Hacking University too, with the same model that you just explained right now. So it is great to hear that, even from your point of you as an expert in this field, we are following something that could in the end, yield some results for us, so that’s actually very, very happy for us to hear. Thank you for sharing that.
That is so exciting in this world of content marketing because people share what they know and everyone learns and there are a lot of great things that come from that.
So I think this is a really wonderful time to be entrepreneur where there are interviews like this or read blog posts, or take a class or whatever, and then get better. Because the first time I was doing this there wasn’t nearly;
First time I was a founder late in ‘07, I had far less of an understanding of what was going on. So I have been there as a novice, I know exactly what it’s like to not know what you are doing at all. So it is really important to kinda commensurate into that, helps you understand people to help them.
Exactly, exactly the same point of view. About hypothetically now, I want to get your point of you on something; so if you are starting your company and want to validate your idea, what is the first thing strategically that you would do to start and gain traction?
That depends on the company of course. But you kind of glanced over of what I would do, so you already validated the idea, what does that mean?
So you have talked to people, you have a product in place, it works and you are starting the face where you talk to people, you want to see what they are talking about your own product and from there how do you achieve growth for your product?
You want to skip the acquisition, the main funnel. What are the main steps that you would take to go to the next level after you get the feedback from the customer?
Sure, so I really want like to stress that this is the part where most people fail. Most people fail to actually understand their customer, talking to them enough to find out what to build.
So most apps that fail on growth, fail because nobody wants what they are building, so this validation is really important stuff to understand you must have a product experience that actually solves the problem, can help people out but let’s skip over that because you asked me about what I would do next.
I think the next thing would be -if you do understand your user, where they are, how to reach them. Depending on who they are, there might be different channels. So for someone who wants to reach the dentists, it might be like some kind of conferences or some kind of magazines, or some kind of publication.
Maybe a list you could buy to do up on sales. There might be different ways of reaching dentists versus if you have, let’s say you want to reach local businesses like breaking water businesses, maybe like if I was blurring my growth ad, you could say “yelp, can we crawl in there and yelp and get everyone’s number?’’ starting with where you know you can find your users to reach them.
And it could just be as simple as an ad campaign. Right? So it might be that you can target people by some interest. So if you look for users and let’s say you are an online vet, and you want to target people that are pet owners, maybe that is the interest you could target on facebook and run ads to acquire that kind of user.
And you could of course, be more refined. One reason people do content marketing is because of the ads they can run afterwards. So if you’ve already grown an audience based upon that content, then retargeting those users in ads is actually incredibly efficient compared to trying to find who those users are out there.
So using some ad like dashboard or AdWords to start trying to find a new user, that’s harder than making compelling content that people share. Just retargeting people to that content, ad retargeting is a big deal. If you are on mobile things are different as well, because there are also sort of different things on mobile, but like depending on who your audience is, how you reach them etc.
There is a higher bar on quality, so you typically do much more work on having a higher quality experience, or to even be to be featured, to be able to talk to Apple, you have this relationship to support. So there are all sort of things you could do on mobile, but overall it’s like a general challenge, there is a big problem because basically when you had the web you could just click a link and use an app.
So it’s much easier to get people into your product. On mobile, the barrier of acquisition is so high because you have to go to the app store and install the app. So that barrier alone changes how you think about growth. So depending on the app you might do different things, to try to get validation.
After you have validation, to try and get traction. So it might be an ad campaign, some kind of social campaign, it might be some kind of getting featured in the app store. Really depends on what you are trying to build. A lot of apps are even more complicated than that. Because often these days they might have some kind of ground game, if you have some kind of logistics at the app, if you have a delivery service, some kind of restaurant deal.
You have to worry about hiring drivers and you are competing with Uber on hiring drivers. It is really hard, or like Uber it might be a marketplace, we have different sides, so it is not only one side to grow. There are actually multiple sides where you are trying to get both the supply and demand side and there is more.
But trying to get those two together and then sync. So if you look at Instacard let’s say, one of their big growth challenges was, matching up the supply of people that get your groceries and deliver them too with the demand of the people that want those groceries. So it is not hard to acquire new users I think, these ad campaigns are easy enough, the hard part is matching things up when the marketplace grows without getting out of balance.
An out of balance marketplace is problematic, so it’s a big deal. So if you are an Uber driver and don’t have enough fairs, like you lose money, drop out of being an Uber driver. If you are a consumer looking for a ride and there is no one to give you a ride, then you are going to stop using the app. I see it like a double sided growth problem.
Anyway, depending on the context there are a lot of different things you could do, but I think to the meta growth, to understand how to think about growth, it is to break down the problem. Like you have a goal, and you break it down to understand its component parts and then run experiments, come up with ideas and experiments to run and 3 hours then by cost and benefit to try to work on improving accrual. And that process is getting down to like running these kind of science experiment, is really like how you run growth, whether it is ads at a strategic marketplace or social app or whatever it is that is the way you break it down.
So it is ultimately a test. Lots and lots of test and experiments to understand what the right channel is for the hypothetical companies growth.
Definitely a lot of experimentation, a lot of numbers, a lot of understanding. A lot of strategy too. So for example, if you have a financial product, like if you are Weathfront, and you run ads and you are competing with people who are selling incredibly lucrative products.
So when Merilyn tries to manage your retirement account, your accounts is worth a lot of our to them. So as a result, the structural issues around financial instruments is that they are really expensive, as far as ads, because other people are competing for the ad space.
If I search for how many K people are spending, you are not going to be as good as them, and so you could strategically say: “Probably running a top level Google Ad campaign is not gonna work”.
So, Wealthfront has a referral program, Wealthfront does content marketing, Wealthfront does a bunch of other things; They ran ads, but they are retargeting ads; and I’m sure they’re doing other things as well, but I don’t want to speak about them because I don’t know their growth too well.
Great thank you so much Ivan for all these insightful information. I am sure every product manager, growth marketeer, anyone who owns a business, or wants to own a business in the future can really take this insightful information and model it and actually make it grow and they can also set the right key objectives for growth. Thank you again and have a nice day, have a nice time.
Thank you very much, thank you for having me!