Pod Insights a16z Podcast Distribution, Channel & Partnerships About this episode The route by which you distribute your product to customers -- i.e., the channel -- is key, as is the decision whether to partner with others here.
A large and diverse set of channels can help with a lot of classic “b-school 101” problems, because you’re typically creating more routes to market for your product while also displacing competitors and building up a barrier around customer acquisition. So getting channel distribution and partnerships right really matters for the “little guys” — i.e., startups. But when does a channel partnership make the most sense for you? How do you determine who the appropriate (and most beneficial) channel partners are? How does channel pricing work, and when can you offer different “channel economics”?
Before joining a16z, General Partner Alex Rampell ran a company called TrialPay (which was acquired by Visa). Using the case study/example of TrialPay in this deck -- first presented at one of our company building events in early 2017 -- Rampell shares how you can best use channel partnerships to grow your business … including understanding what they can do and what they can’t.
Transcript 0:00 Speaker 0 Hi, my name is Alex Rampel. I'm a general partner at Andrews and Horowitz, and I'm going to talk about distribution, channel, and partnerships. So before I joined the firm, I ran a company called Trial Pay, which was acquired by Visa in 2015. And I wanted to give a very, very quick background about what Trial Pay did and how we use channel partnerships to grow because I think it will help illustrate what channel partnerships can do and what they can. So just some very, very basic background. You have currency as a customer. So as an example, you're worth $1,000 to Comcast because you pay Comcast a lot of money every month. Therefore, they could afford to pay Google $1,000 every time they get a new customer. You're worth $2,000 or more to Geico. If you're a relatively good driver, you're worth $100 to Netflix, you're worth hundreds of dollars to the cell phone carriers. And these companies will pay some percentage of this to whoever refers you to them.
0:48 Speaker 0 So before Trial Pay came along, uh, the main, the main uh exit valve for a lot of advertising was Google search. So take washing machine. You do a search for a washing machine. That's a relatively transactional search. You want to buy a washing machine, all these different companies would advertise there, lots and lots of them. And it was great for the, you know, for the May tags and the best buys of the world because you're looking for a washing machine, they sell washing machines, Google makes that match happen and they charge a cost per click. What Trial Pay did was we said, you know what, let's try not to occupy when people are searching, but before they're searching or after they're searching, really when they're buying. So here's an example uh when when DVD players were big in 2006 when I started the company, Blockbuster wanted to advertise to people that were buying DVD players, and there's no better time.
1:34 Speaker 0 Like, sure, they could buy Google ads, and they did, and that that worked well for them. But what worked even better was showing ads at the point of transaction, and that's what Trial Pay did. The key thing was how do we get lots and lots of transactions? We need to partner with lots and lots of merchants. Some of them are very big, but a lot of them, actually, the vast majority of them are not like Amazon. They're a little tiny guys that might make five or 10 transactions a day. Another example of what Trial Pay did was we worked with almost every large social gaming company. So social games became big on the Facebook platform, you know, starting in 2008, 2009. Most people didn't want to pay for social games, so we would show advertisements there. And here's a game that nobody had really heard of called Crowdstar Happy Aquarium. We helped them monetize by saying, buy flowers for your mom, get happy aquarium for free.
2:20 Speaker 0 And likewise, company called restaurant.com, where they would show offers from Ghiradelli or ShuBai after you bought your restaurant gift certificate. Last example that I'll give you here is kind of makes sense. If you're buying a movie ticket from Fandango after you're done with your movie, you might want to go see, uh, you might want to go shopping at the mall because the mall is near the movie theater, get 40% off of the gap. When we did this promotion, uh generated about a million dollars in sales for gap in 24 hours. We had lots and lots of partners. We we got up to tens of thousands of clients, both on what we call the publisher side, the people that showed our offers and also on the advertising side. It took us a long time to get there. But the key thing that caused us to cross the 10,000 merchant barrier was not having individual salespeople or actually not even buying ads. Like, how do you how do you advertise for something that people haven't heard of?
3:09 Speaker 0 Nobody was looking for transactional advertising. We had to find them, but we had to find them in a scalable manner. So just some generic advice that I'll start with, because I think that this is uh tried and truism at this point. But whether your Facebook or a or a brand new company, it always helps to focus on a tight vertical, even if you don't have a network effect business, just because you can kind of go around, walk around, find a couple friends, get them to try your thing before you cast too broad of a net. So in the classic example here is that Facebook started off at Harvard, they won that vertical, then step two, they expanded. And I remember a very uh meaningful uh day for me was I sat next to Reed Hastings at dinner. This was probably 10 years ago, and I asked him at the time, why aren't you expanding Netflix to Canada or to Europe or to Japan? And he said, you know, I can still grow 10x in the US.
3:55 Speaker 0 And as soon as I don't think I can grow 10x in the US anymore, you're going to see me expand. And lo and behold, as soon as Netflix hit 30 million paying subscribers in the US, they started expanding everywhere because the population of the US is 300 million, makes a lot of sense. Now, how do you conquer the vertical? The key thing that I tried to do with trial pay, trial pay was a personal project because I used to write software called shareware, tried before you buy software. I uploaded it on the internet. I wanted people to pay me. Most people didn't pay me. So I said, you know what? If you don't want to pay me, I'll give it to you for free. If you switch to Geico or get a Discover card or sign up for Netflix, this doubled my little shareware business revenue. So it seemed like a pretty good idea. So the personal experience for me was all right, I'm going to start with consumer software, then I'm going to kind of go to the friends or friends and family, which were incredibly helpful towards winning my first customer.
4:46 Speaker 0 My first customer at Trial Pay that was not me was a company called Winzip, which at the time had hundreds of millions of downloads. Nobody would pay for it. But I knew the guy that ran Winzip. So he just said, Oh, yeah, I know I know you. That kind of makes sense. And he started working with me. Meanwhile, I had a number of competitors to Trial Pay that had emerged, but they just didn't, they were inorganically constructed. They didn't have any basis in this particular software industry. So it was hard for them to really gain a foothold. But the key thing around software, what actually made what made it hard for all of my competitors was that there were so many little guys out there, and it was consequently not cost effective to do classic sales to them. So, you know, if there are 100,000 clients out there that each write software products or think of it today as iPhone apps. There are there are millions of developers that develop iPhone apps, Android apps.
5:34 Speaker 0 If you could make $500 per year against $100,000 small clients, that's $50 million a year in revenue. But you can't run a business if you're making only $500 per year and you have to pay a salesperson $100,000 a year. And the salesperson can only sign up 10 clients per year. You just have a mathematical disadvantage that you're never going to be able to surpass. And that's really where channel partnership is kind of the name of the game, along with marketing and sometimes inside sales. So let me rewind a little bit and talk about what is a channel. And I'll use an example that hopefully people are familiar with, which is a hotel. So Hilton, big hotel chain, they have different distribution routes of getting their product, namely a hotel room into the hands of their customers or the people that stay in their hotels. So the first one, they're called the OTAs, the online travel agencies. So Expedia, Priceline Orbits, a bunch of these have merged, but they're the aggregators.
6:23 Speaker 0 There's another big one called Booking.com, which is now priceline.com. There's a second group, which are the agencies. So if you work at a large corporation, you probably use somebody like Carlson Wagonlit for your travel, as opposed to going to Hilton.com directly and then booking your room. Or airlines. So if you are booking a flight from here to Minneapolis, at the end of your flight, you're probably about to go stay in a hotel. Hilton really wants to be at the back end of every Delta or United or Jet Blue flight. So that's another distribution route for them. And then there's the last one. And there are many others that kind of are some amalgamation of these. But uh the last one, which is the most important one for Hilton is direct. That actually might not be the most important for them in terms of what share of volume, but that's where they want you to go. They want you to go to Hilton.com.
7:08 Speaker 0 They want you to pick up the phone and call Hilton, because then they don't have to share the revenue with anybody. And moreover, if you go to Expedia and you see here are 400 different hotels that you can stay at, you know, only one of those is a Hilton hotel. They want to capture all of your mind share, but they also recognize that is not realistic, which is why they rely on the channel. And if you are able to build a business with a large channel, it actually helps with a lot of kind of classic kind of business school 101 problems because you're able to construct more defensive motes and barriers to entry. Because you know, imagine that you make serial, and it would be great if everybody goes to your website and buys your serial directly. But if you get prime shelf space at 40 of the largest supermarket chains in the country, well, you're actually this is a zero sum game. They don't have infinite shelf space.
7:55 Speaker 0 They're giving it to you at the expense of somebody else. So if you're able to build a large and diverse set of channels, you typically are displacing other people and you're able to build up a little bit more of a barrier around a key thing, which is customer acquisition. For trial pay's business, this is how there's some arrows here, and I'll try to illustrate what this meant. But we had a sales and marketing organization, which normally means that the marketing team is meant to drum up demand. Like what is trial pay? What does it do? Let's sponsor events, let's get people at the very, very top of the funnel. When people raise their hand and say, okay, I'm kind of curious, what is this thing? That's when sales would come in. And then once the client was signed, we'd have account management work with them. But we also had a self-service channel. And sometimes marketing would go straight to self-service because if you click on an ad from Google saying, hey, are you a little game developer that makes 100 a year, please sign up for trial pay, that would have to go through self-service in order for it to be cost effective.
8:51 Speaker 0 And then the channel would typically try to deposit people into self-service land. And that's not always true. Sometimes if you have a big system integrator that is a channel partner of yours and you're you're making a firewall security appliance, that's not going to go into a self-service channel. But for our business, channel was purely meant to deposit more people that were too small for us to go after directly. Because if we go after them directly, then we could afford to pay a salesperson. They were too small to go after directly. We couldn't find them via marketing because they hid in weird places, but we wanted to partner with people that they worked with. And the way that we thought about this is looking at it as a technological and chronological stack. So we said, okay, who could use trial pay? Remember, we want to work with lots and lots of merchants that are selling things to put ads or offers before they sell or after they sell.
9:38 Speaker 0 So how do you become a web merchant? Well, first you need to get a domain name. Then you need to get a web host. Then you probably need a commerce store like Shopify. Well, then you probably need to accept credit cards. So maybe you're working with strong. Or you're working with Vantv or you're working with Chase Payment Tech, then you need accounting software. I mean, maybe you don't, but most people probably would at a certain scale. So you have a QuickBooks account. And then you probably need to buy customers. So you're using AdWords or you're buying ads on Instagram or Facebook. And maybe you're also using email marketing software to go communicate with your customers. So every one of these is a potential channel route for us. Because if we could partner with every merchant bank out there, wow, like we'd have a phenomenal business. Or if we could partner with QuickBooks, we might have a phenomenal business.
10:24 Speaker 0 That doesn't always mean it's a likely scenario because take domain names. If we were to go partner with GoDaddy, GoDaddy, probably 90% of their traffic is not people that are selling things. So it would not be a very, very relevant offer for GoDaddy to go enroll people in trial pay. But this is the rough way of thinking about how do you build channel partnerships. Think about the whole chronological stack leading up to the discovery of your product or service. And then once you have discovered this kind of short list of who might you want to work with, there are numerous ways that you can actually build out a channel partnership. Like what does that actually mean? What does a channel partnership mean? It could just mean like they send out a newsletter and they say, hey, uh trial pay is great, go use trial pay. That's kind of a referral relationship. So there's a web hosting company called web.com.
11:09 Speaker 0 We started working with them and they just put a little banner on their page saying, hey, trial pay is great, go partner with trial pay. That's that was basically it. Worked great for us. We got a lot of junk because it turned out there wasn't a high amount of relevance. And then there was the kind of, ooh, super duper relevant. And maybe we integrate with them, or maybe they'll say, wow, this is so relevant that we want to white label you. So we don't want to send our valuable asset, our customers to you. We want to just say it's not trial pay, it's uh e junkie, or it's PayPal because they were so relevant for us. And one thing that I learned through this process is that the people that were the least likely to be a channel partner were people that were growing quickly, which kind of makes sense. Because if you're growing 100% year over year, why would you bother kind of like making a little bit of incremental revenue here and there?
11:56 Speaker 0 The key guys to go after ended up being people that were looking very, very seriously for other revenue opportunities because their growth engine had sputtered. They had lots of customers and they wanted to figure out what else do they do with them. And this is actually I I had a guy that worked for me that worked for BCG and he liked to do two by two matrices. So this is actually how we uh we constructed this, which was it needs to be relevant for us because if it's a client that only has one new merchant per day that signs up with them for their like little accounting software, probably not gonna be relevant for us, and it's not gonna be relevant for them. But you know, take somebody like PayPal. PayPal was by far number one on our list because wow, PayPal, they had at the time like 7 billion transaction pages where they could put our offers. We make 50 cents per offer that we show. That's three and a half billion dollars.
12:43 Speaker 0 Incredibly relevant for us. And also, you know, pretty relevant for them, but it just wasn't very feasible because this wasn't their top priority. They didn't really care about it. There are a hundred reasons why it wasn't quite the right fit for them. We really zeroed in on ones that we thought were high feasibility, largely because partner growth had tapered. So PayPal is still growing 40, 50% year over year. Do they have to change the formula, try something new? Not really. Whereas download.com, ooh, they they really did, or there was a um a company called Kaggy, which was an online payment processor, kind of full shopping cart payment processor, same deal. They needed the growth, and they also had a high amount of potential benefit. So we zeroed in on a company called download.com. So download.com had 64,000 products, or rather, the way to think about it is 64,000 different small merchants or software developers.
13:29 Speaker 0 Now, in today's era, iPhone app developers or Android app developers, but this was kind of back in the era of Windows and Mac software, it offered us defensibility by sending traffic to our existing partners. And I'll talk about that in a little bit. So the key thing about this channel partnership was it didn't just get us 64,000 clients. We didn't get all of them, but we got a big chunk of them. It also helped our positioning with all of our existing clients because it made it harder for them to move to a competitor. The key thing though, um, and this is a phrase that I used in my company a lot, which was it was bi directional relevance. It was relevant for us because holy moly, we have 500 clients. If we get 20,000, that's incredibly relevant. But also from download.com's perspective, like if they could make $5 million a year from real estate that they weren't doing anything with on their site, that's incredibly relevant for them against kind of a shrinking PL.
14:18 Speaker 0 This is actually, I took this slide from one of the original decks. I kind of masked out a couple numbers. So it says XM per year as opposed to the actual number of millions of dollars per year. But we came up with a whole presentation to download.com of, hey, here's how this is going to work. Here's how much money we can pay you. And it actually worked. They were like, yeah, we got to do this because it's very, very relevant for us. And it was just so uniquely tuned in to what they were doing as a company. But this didn't happen by accident. Like we really scoured the market. And sure, we would have rather done a deal with PayPal, but it just wasn't PayPal was growing too quickly. It wasn't as central to PayPal. It wasn't addressable to all of PayPal's existing customers in the same way that it was to download.com. And this is what I was talking about aiding defensibility. So the way that our business worked, if you remember some of the examples at the beginning was we would partner with merchants, customer traffic would go to merchants, then merchants would say, Hey, you don't want to go pay for this little app, go to trial pay.
15:14 Speaker 0 And again, we had five competitors. So we always worried about the merchants saying, don't send, don't, you know, take out the trial pay button. There was one of our competitors called Lyft Media. There was another one called MyOfferPal. What if they put my offerPal there? What if they put uh Lyft Media there? We'd lose that customer. We didn't have a very, very sticky relationship with them. The nice thing is that the second example here on this slide where trial pay is in the middle, that's really download.com. So a lot of people would go download winzip from Winzip.com. And that's that first example. But a lot of people would go to download.com and say, I'm gonna go download Winzip. And then there and right there, they would see trial pay. They would see an offer. We didn't brand ourselves as trial pay, but they'd see the offer for Winzip that was powered by trial pay. It wasn't powered by Lyft Media, it wasn't powered by OfferPal.
16:01 Speaker 0 For them to replicate this, they would have to get Winzip as a client and they would have to get download.com as a client. So it really, really helped us. I mean, in fact, almost all of our competitors went away after we did this deal because it turned from being a very, very vulnerable position where merchants could kind of switch us on or off or like rotate us and five other competitors to one that was a lot more defensible because we had this channel and nobody else didn't. And this is what it looked like. So when you went to go download, uh, this is a antivirus product called AVG. When you went to go download the AVG antivirus tool, you'd see, and this is a little bit I'm I'm embarrassed by the graphic now, uh, but you'd see, hey, get this for free if you go shop at Petco or if you sign up for Blockbuster, if you buy flowers from FTD. And we actually, this is part of how we got AVG as a client.
16:47 Speaker 0 And ABG was a massive client. We actually tried selling into them, but they wanted to have, they wanted to participate in this download.com program because download.com was a big channel for them as well. Now, what about pricing? This is actually very, very important because you often hear this concept of channel conflict. What does that mean? What it means is that from the perspective of the manufacturer, channels should not compete with each other on price. Because imagine that Hilton example where you want to go find a Hilton hotel room. Hilton says we're going to sell it directly for 199. And then all of their alleged partners go undercut them and say it's 150, it's 140, it's 120. Who's going to go buy the hotel at Hilton.com? Nobody's going to get a hotel room in Hilton.com if they're known for having the highest price for their own product when ironically they don't have to pay anybody else.
17:32 Speaker 0 So the way that this works is channels are often not forbidden legally, but forbidden by by contract from competing with each other on price. But channel partners can be offered different economics behind the scene. So Hilton might give up 30% of the economics to American Express. So they might give up 30% of the economics to Booking.com, but that's not transparent to the customer. And this was how we structured all of our channel partners as well. Because if we end up getting Winzip or a tiny software developer through download.com, we don't want that developer to get worse pricing because they weren't going direct to us. We didn't want to disincent download.com from sending us customers. We had to keep the transparency of pricing sacrosynced. And the way that we did it was a little bit tricky, which was we convinced the end partner, the end client, in this case Winzip, hey, the person came from download.com, go pay download.com their half, because normally what we would do is we'd take about 20% of the economics.
18:23 Speaker 0 So if the person went to Winzip directly and then monetized via trial pay by signing up for Netflix or getting flowers, we would get $6 and Winzip would get 24. If the person went to download.com, well, we had to pay download.com. How did we end up paying them? We split the bounty that went to Winzip 50-50. And this was our kind of clever way of doing this. But the key thing is we didn't want Winzip to be kind of clandestinely worse off if they went through download.com. Because then we or our salespeople would have an incentive to say, hey, potential customer, don't sign up through the download.com link, sign up through the trial pay link because that way you're going to get a better deal when you start really generating revenue from our service. So once we had proven this distribution model, now we knew we could expand. And actually it was funny. We tried working with a dozen companies that were like download.com.
19:09 Speaker 0 And before we had download.com, they're like, that's a dumb idea. That won't work. We don't like you. We never heard of you. After we had download.com, they're like, wow, that sounds great. Our sales pitch really gravitated from here's what we do, here's the management team, here's our venture funders, all this kind of stuff to we have download.com. That was the entire sales pitch, and it worked. And we kind of blanketed the whole industry. So all of the download sites on the internet started integrating trial pay. And most importantly, they only integrated trial pay. They didn't have room to do offerPal and Lyft Media and others, which really gave us more distribution and defensibility for that core vertical. And then we started saying, okay, we we've done well in this software. Where else can we expand to? And we really wanted to get PayPal. And actually, this is a key takeaway that I would also leave everybody with, which is you often think, forget about bilateral relevance where you have to do something that's relevant for the client and for yourself.
20:02 Speaker 0 If you go to the biggest guy first and you screw up, that's not good because then their institutional memory, this is like the best whale that you could ever hunt. And they say, oh, whoa, I remember that company. We tried them, it didn't work. And half the time it doesn't work, it's their fault, but you need to learn what type of tests work, which ones don't. So if we had worked with PayPal first, it would have been a disaster. I'm sure of that. Because either they would have screwed it up, we would have screwed it up, and then we wouldn't have had a second chance. The training wheels are often the smaller channel partners. So if you again go back to serial, if you have serial, you might want to try selling that somewhere else before you get safe way to do a 2000 store test. Because if it doesn't perform well on that test, you're not going to have a second shot. So we couldn't get PayPal. So then we went to eJunkie and they added us as a payment type.
20:50 Speaker 0 We figured out the economics, and we just kept doing this dozens and dozens of times, really as practice for the big fish like PayPal. So summarizing here, I mean, channel deals are key for virtually every company. I mean, there's a lot of stuff that channel will not do. Like you're better off owning the customer through and through. You're better off not going through a channel partner if you can avoid it. But you know what? Apple sells direct. You go to Apple.com, you go to the Apple store. They also sell at Best Buy. They also sell at Walmart. Channel deals are key for virtually everybody. The way to look at these from my perspective is really from the three lenses of defensibility. Like, does this help you build defensibility? Does it help you build distribution? And then how profitable is it? Because we often had things where it didn't actually help us with distribution that much. It helped a lot with defensibility.
21:36 Speaker 0 We had others where it didn't help with defensibility, but it gave us a ton of distribution. And we had others where it was just very, very profitable. So there was no reason not to do it. But if it's not hitting one of those three, it's probably not worthwhile. Third point I would make is uh complimentary, and this is complimentary with an E and not an I, the things that are complementary to your good or product or service or offering are the key poaching grounds for especially newer concepts. So if you are making something that's brand new, like it's different if I'm I'm a new pizza company. I know that there's a steady stream of people looking for pizza. I just go advertise on Google for pizza. If you're doing something that people have never heard of, how do you find those people? Because they're not looking for you. Well, what are the other things that they are currently doing? Those are the poaching grounds for that type of customer.
22:23 Speaker 0 And then lastly, as I mentioned, like even though this is all great in theory, it's much better for Apple if they sell an iPhone from Apple. It's much better for you if you can own the customer through and through. Because in many cases, if you're relying on channel partners to message what you do and how you do it, they'll screw it up. It's not their core business. Or if you're relying on their sales team to reach their own customers, how are those salespeople compensated? How do they know how to message things correctly? So it's really the way that I think about channel partnerships is it's not this opportunistic, oh, those guys already have all the clients. I'm gonna go work with them. I'm gonna let them sell it. That almost never works well for the company. It's more, well, I could sell to them directly, but I'll lose money on every sale because I'm paying my salesperson X dollars. They can close 10 sales per year.
23:09 Speaker 0 The economics aren't going to work. I have to take that through channel. So I hope this has been helpful. And thank you very much.
Previous Crypto and the Evolution of Open Source Next 3 Common Myths People Have About Crypto Page updated August 25, 2026 · transcript and summary are machine-generated