President of Pearmill, ex-Head of Product at Taplytics, ex-Head of Mobile at Frank & Oak. YC fellow.
ex-Sequoia Parter, ex-Uber Growth, ex-Facebook
CGO at Portag3, ex-CMO Policy Genius
As companies grow, the challenges aren't just limited to finding your customers and acquiring them.
The company needs to learn how to build the right processes to continue its trajectory, and the growth teams need to learn to tap into company resources to continue scaling.In this episode, we'll be having Amy (early Uber growth) and Jonathan (ex-CMO of Policy Genius) on to talk about the challenges of managing growth at fast-growth companies.
Both of our guests have been both operators and recently been on the venture capital side of startups, giving them a multi-dimensional perspective.How do you manage up effectively? How do you work with product and engineering? How do you think about channel composition as you scale? What do the best companies look like from the outside?
These are questions that are seldom discussed in our community. Join us as we dig a little deeper than usual!
Hi, I'm Nima Gardideh, I'm the co-founder and CTO of Pearmill and this is "The Hypergrowth Experience”. As companies grow the challenges aren't just limited to finding your customers and acquiring them. The company needs to learn how to build the right processes to continuous trajectory and the growth teams need to learn how to tap into company resources to continue scaling.
In our second episode I chat with Amy Son, early Uber growth member ex-partner at Sequoia and Jonathan Metrick ex-CEO at PolicyGenius and Chief Growth Officer at Portag Ventures about the challenges of managing growth at fast-growth companies. Both have been operators and recently been on the venture capital side of startups, giving them a multidimensional perspective.
We cover important topics like how to manage up effectively, how to work with product and engineering and how to think about channel and composition as you scale and what do the best companies look like from the outside in. We selldom discuss these topics in our community. So join in for a deep dive with Amy and Jonathan.
Amy Sun: Hi everyone. I'm Amy Sun. I am currently the founder of a new startup called Daylight. We're still sort of in steth mode I guess. I'm prior to, starting my own company, I spent three years as a investor on the growth stage side at Sequoia Capital.
And prior to that I was in growth marketing at Uber in the early days, transitioned to product management and also spent some time in product management at Facebook And one factoid about myself is that for the last 10 or so months, I've been a vagabond uh, and just driving across the country, like living in various places from Boston to North Carolina, to New Orleans and I'm currently in Austin.
Jonathan Metrick: Very cool. That's gonna be a tough one to follow on Amy. My name is Jonathan Metrick. Uh, I'm the Chief Growth Officer, at Portag Ventures, which is a FinTech focused VC. Some of the companies in our portfolio is Well Simple, Albert based out of LA, Clark out of Germany's and Insurtech out of Germany
Prior to this role. So I'm kind of a growth advisor. I advise on marketing and growth with our portfolio companies. Prior to this, I was the CFO of PolicyGenius for three years, helping scale up 10X over the past three years in New York kind of have a deep experience in the marketing side, across startups, Proctor and Gamble.
I used to work at Live Nation Entertainment for awhile in the entertainment sector but all things related to marketing. And most recently in the startup space. An interesting factoid about me prior to the pandemic, we could actually do interesting experiences and travel a bit.
I would highly recommend if anyone ever goes to Tokyo was an experience you can do where you can dress up and get it a go-kart and drive around the streets of Tokyo as Super Mario. And I would highly recommend it. It's very cool people stop on the streets to take photos of you. if you're ever in Tokyo out of the pandemic, you should check that out.
Nima Gardideh: That's hilarious. It's funny as I actually, I don't know if this is a service, but in New York, I was on the pier playing volleyball a [00:05:00] couple of months ago when it was. And there was a race going on around our field with a bunch of people in Mario Kart outfits. And it was hilarious. We spend like 30 minutes just watching them and making bets on who's going to win.
It's amazing. Cool. Well, thanks for doing this guys. I think what was really interesting out of the previous interview when we started chatting, I think both of you have seen a couple of brands go from like very early to massive scale. And one of the things that always makes me curious is there's some things that you do in the early days that will work for awhile and then they stop working at some point.
I think we've talked about how like, even on this show that the good really strong brands essentially have one channel in the first while, like first few years, and then they just doubled down on that channel to become massive company. But I wanted to sort of put it out there. It's like, what are the different stages in your mind?
Like if there are there frameworks of like, all right, maybe pre-product market fit. This is what I do post [00:06:00]product market fit. This is what I do. Or like, what are the different stages where you think it started very shift your mindset when you're trying to grow, grow a product.
Jonathan Metrick: I mean, I can start, I think in the earliest stages and I do it more so kind of like by funding structure. So like in the pre seed stage or seed stage, a lot about trying to find product market fit. Right? And so a lot of the companies I work with in our portfolio it's been scrappy, right?
Generalists trying a bunch of things trying not to invest too much in a certain channel for a certain experience until you know, that that works well, that you can get in front of customers, that they're actually going to buy your product. You know, that they're, you know, valuable customers and they actually might tell their friends and refer you. There's a lot of kind of scrappy experimentation and the very, very early stages that are critical to establish product market. And really hone in who are you, for who, and what value you bringing for that. And then I think as you start to scale up, you get a bit more of an idea of that.
You're getting into series A and you're getting into series B, [00:07:00] that's where you get a little bit more into the specialties, right. And you carve out, okay. Actually one of our best channels for acquisition is going to be Facebook or it's going to be podcasts. And, you know, you start to bring in expertise in that area and start to establish best class in that area.
Procedures, processes, new building, a bit of a moat for yourself. But I think those are kind of, some of the differences I see in the really early days is investment keeping nimble. And then as you get bigger, doubling down on what works and bringing in specialists and expertise to, to expand.
Amy Sun: Yeah, that that really resonates actual with my own experience at Uber. So just when I joined Uber, when UberX was just launching. And that was when the growth team was formed because it was like, oh, we need actual some growth expertise to help with this like mass scale behavior change problem of. You all of a sudden, you need to try to convince millions of people across the world to quit their jobs and like drive strangers around full time.
And that's like a completely legitimate thing to do. [00:08:00] And, and prior to the growth team being established. So when I first joined everything was like that scrappy hustler mentality. Like people were just like calling up limo companies. Literally something that I did in the early days is we would go to the airport and like knock on the windows of like black car drivers and be like, Hey, like here's a card work at Uber.
You should sign up for this thing. And we were just like hustle drivers. But the thing was interesting because. Real user feedback, in addition to like onboarding these drivers, literally one by one. And so you got a sense of like their needs, what they do, where they hangout, like what would convince them to like, take that leap and start driving on your platform.
And so that was like really valuable, but it was not scalable, right? You can't onboard a million drivers through this process. And so that was to try to automate things and like, how can you make this more efficient? How can you reach people without having to deploy a bunch of like interns into malls and try to like scale people onto the platform.[00:09:00]
And like a lot of referral loops were built. I think referrals in the early days can be super valuable. And, and Uber actually had zero budget dedicated to acquisition in the, in the beginning. And so. things like trying to get to our reserve for each other, trying to get writers, to refer drivers, doing some promo code stuff where you gave away free rides and or like credits on the driver's side or super valuable.
And then when you're like, oh wow, we really we're really constrained by supply. We need to like, turn this up 10X or a 100X. That's when started doing some performance marketing channels, which is when I first joined. Yeah. And then uh, the first channel that I actually worked on, which I actually talked to Nima on within the in-person panel, years ago, like four years ago was on Craigslist.
It's a super good channel for Uber. And it was also like. inexpensive, right. There was like $25 a post. And I literally spent my first, like weekend Uber, like sitting in a closet, waking up at five o'clock in the morning and posting to Craigslist for like, like five or six hours [00:10:00] straight, like manually.
Right. And then you're like, oh wow, this is actually working. We're tracking it. We have some systems. And then turn up performance or more dollars in, or scale the similar channels. And I think as the company grew larger Performance was a huge channel for drivers, but there's a certain point where it's no longer, really efficient to keep acquiring that like next incremental new customer.
And you have to think about like engagement and CRM and like you already have this huge group of people that are signed up or I've driven at some point, like, how come. It's much more effective to re-engage that existing base. And then later on in the company's life span, when it's like a household name, then like brand becomes super, super important.
Like how do you come that first choice? But like switching from that performance mentality out of into like a more engagement brand mentality was super, super hard because the whole organization was like, and the whole data system was like set up [00:11:00] to think about what is the cost of this next incremental fiber or rider on the platform or trip.
And it's like, the whole team has to like shift.
Nima Gardideh: Super interesting. Right? Cause like, basically if you become a household name, you no longer have a performance problem, a for performance problem. People who've never heard of us, please. This is, this is what we're doing and then come and use our product.
Then you shift upon where you're like, everybody knows about Uber. There was no human doesn't know about it. So now it's around like, okay, can we create integrity around the brand and then be on top of mine? So then they use us over Lyft or whatever. And so my sense of the later stages. Yeah.
Jonathan Metrick: Yeah. And I think the interesting piece and Amy, you tee this up is who are the folks that are actually best suited for that work?
Right? They're often very different, right? The folks who are scrappy and hackers want to break the system one to find a hack, they want to try something new and if it doesn't work, that's okay. But as you become a household brand you know, having worked at Proctor and my, the early days of my career, you actually had more to lose.
By taking a risk with a brand like Tide that [00:12:00] everyone knows that's buying the has high loyalty than to win by some hack or some growth hack and has a very different use case and how you deploy marketing and the folks that want to work on those sorts of brands are very different. Right? And I think as you transition out from the stages sometimes those folks may, may grow with you and many times they might not.
And I think depends on what what's needed in the business at that time. But I think one of the major inflection points and Amy, you, you mentioned that a little bit too, is as you start to point incremental amounts of capital, right? So it's kind of like if your budget's low and you're super scrappy, the risk is low of a mistake or trying to different pieces.
That's okay. But if you're deploying hundreds of thousands, millions of dollars into a channel. kind of not really knowing what you're doing, you just kind of being scrappy. I wonder, sorry, go on. I mean...
Amy Sun: OH, I was just going to completely agree. And I was just going to mention, like when I joined Uber, I had a year and a half of experience in something that was like completely non-related that was in product marketing at Microsoft working on Windows. [00:13:00] And, and so like and that, like the kind of person was like on the team with someone that was like super young, super hungry, shallow knowledge and a bunch of different things.
So you can go and like run a quick test, understand if something works. And then at some point like I'm like, I am currently like creating SCM campaigns. I know nothing about like ad-words non-branded search keywords. once, you know what works, you can hire someone that's like best in class search for what you need best in class for like video advertising or something.
And then like, bring those folks in. Um, but in the scrappy, both can like stay on and keep testing stuff. But, but that is an interesting transition as well. Like from a career perspective, I was like, well, I'm like, I don't have that deep knowledge. What is the next thing?
Nima Gardideh: Yeah, it's interesting because I think basically the way I think about it is there's like the exploratory stage.
And then there's the exploit stage where you're like, okay, you figured out one of the channels. And then now you want to master that channel. It was a totally different persona and or team that is required to master one of these [00:14:00]channels. Like I joke around with folks that hire us a lot.
We're like, oh yeah, you want to master page, look at the composition of our team. We have engineers, we have designers, videographers, cinematographers, like illustrators in order to like succeed on one channel to like scale it up. You have like a totally different set of skillsets that you need to tap into.
it's sometimes like actually scary when they realize, oh, if we want to go from like the $50K a month to a million a month on this channel, it takes a whole team to do it. and it's like, it's not an easy thing. the thing that ad came up around referrals, Amy, it was, I surprised that you mentioned that they didn't have a budget for it because that is a budget, right.
When you're, when you're giving away money, that is like, how did you guys think. CAC in that scenario, was that like baked in and then when you started doing performance, were you thinking about, oh yeah, we're getting these drivers and then they have like a K factor on top of that. And then we lose money this much when they refer new drivers and like, how would you do the [00:15:00] accounting there? That seems like pretty complex, right?
Amy Sun: The referral stuff was a bit complex and honestly, We were not super scientific about it. And, you know, we had some metric apples, but a lot,[laughing].
Jonathan Metrick: Behind the curtain [laughing].
Amy Sun: But like in terms of there was a lot of, oh well, if someone's coming in through a referral channel, like what's the likelihood that they stay on to take another trip, another five trips, another 10 trips and referrals were actually a super sticky channel. So like the drivers that signed up and the writers as well, who signed up via referral attended to have a much higher LTV than someone who just signed up director or sell through it through a different channel.
And I think it's because they have a network of friends around them who are also doing the same thing, right. There are people to like guide them through the process. it was like, give her a ride, get a ride. And then on the driver's side, it was a varying bonus amount for the drivers, like where the driver who referred would get some, and [00:16:00] then the new driver that signed up also got them.
And in terms of like, thinking about the like LTV, CAC math, as long as they were like retentive to a certain degree and they didn't, they didn't just drop off, like immediately, there was a point where it like paid back. Right. So that was how it was thought about. And yeah, and I think the budget wise, that was actually how we initially pitch, please let us do a little bit of paid marketing because we are already spending this much provider and per driver on this like referral program.
Like, why don't we just, we can keep that as like the max budget that we would allocate to something new tech, we're just going to go and see if we can get the same job elsewhere. So it was a kind of crazy. So this is like Uber in 20. Early 2014. So it was like already unicorn company kind of darling, a Silicon Valley never had run branded SEM campaigns. So like when you searched Uber, someone else's ad would [00:17:00] like come up.
Nima Gardideh: Oh I remember we were running ads towards Uber actually. Cause I was making money. [laughing]
Jonathan Metrick: He was driving up your CAC!
Amy Sun: We would query these accounts? And like, there'll be like, oh my God. Some of these writer accounts have like thousands of dollars of credits. [laughing]
Nima Gardideh: Yeah. So I was basically being a couple of hundred bucks at a time. It did it with Airbnb too. These startups were blowing up. People are searching their brand name. All you have to do is throw in your referral. I mean, I think everyone has like systems against this. Now all you have to do is throw in your referral link as the landing page and boom, you make money or like you can ride around for free afterwards.
Jonathan Metrick: That's so funny about that. Yeah. Yeah. It's interesting. I think at that point Amy, we were discussing a little earlier around kind of like that referral metric. Right. And I think that is one of these big barometers and the earlier stages of how well do you have product market fit, right.
In the sense of, are you able to actually are folks using your product? They like it so much that they [00:18:00] then continue to use it. That's the first basis of it, but then are they willing to refer a friend and to what degree? Right. And I think with referrals you know, on the VC side of things, we evaluate that a lot.
Like, What is your referral efficient? Right. And folks who are referred are probably higher conversion, probably lower churn and, you know, cause they're doing the selling for you. Right. And I think that's a really interesting metric Not enough businesses, I think, really try to capitalize and take advantage of their referrals.
Right. And I think it's the folks who love your brand are there's no better person to sell it and promote it than them. I think Uber is you know, best-in-class example of that, a few others, of course, but that's a really untapped channel, I think with a lot of, a lot of folks. Yeah.
Amy Sun: I guess the end, the thing that I think was I've, it's probably common knowledge now, but was very enlightening at the time is that the give part is super, super important for referrals, but like you're actually giving your friend something and it, Uber tested a bunch of different variations and even the, like give a friend a free ride and you get nothing.
Like people are still willing to share because they're actually [00:19:00] more excited because the product market fit was so good. It was like, Hey, like you need to try out this new service here. Let me give you this free ride. And like you can try it out, which is really interesting. Yeah. That's exactly the product market fit is so good.
I'm willing to market your product for you for free in order for my friends to win something cool. Like you're like, look, I'm going to get off of black cars. Can we be real?
Jonathan Metrick: [laughing] Social capital.
Nima Gardideh: Yeah, I remember that I was like 18 or 19 or whatever. We were launched in Toronto, Canada where I lived at the time. I think Jonathan you're there right now.
Right? And the black cars were just cool. What 18 year old is riding black cars. Nobody. Right. But for sure, I felt like on top of the world be at 18, like living the black car life. Meanwhile, in my late twenties, there's no way I would have been in a black car in Uber.
Amy Sun: Actually, for me personally, I loved gossip girl. [00:20:00] And like in gossip girl, they wrote these
I've been to that like human psychology of it with your product, there's something interesting there, whether that's like looking cool with your friends or like, something like that on the, probably the referrals though.
Jonathan Metrick: I think it's really interesting as well, like around like value and perceived value. Right. And I think some of the most viral referral pieces I've seen are things you know, like Robinhood, giving away a free stock what is the value of that? They're not giving like a Tesla stock, you might be getting a stock at Joe's coffee company that costs five bucks. But the sound of we're going to give you a free stock or a free ride.
It sounds way better than, oh, that rides probably $15 or that stock is $8 you know, it's kind of perceived value. And I think you know, where, brands have really done well is through leveraging mechanisms like that versus, Hey, here's $20 Amazon cash. Right? Cause people have a lot more of a sense of what, what that's worth
Nima Gardideh: Totally. I think that that's, what's an interesting thing too, in general in referrals is it doesn't work just because [00:21:00] you put it in there, right? Like it only works when you have level of product market fit to begin with, and then I'm going to like gain something by doing it. And that game thing could be so many things. One, it could, could be just social value which seems like Uber had it at some point where I'm cool by having access to Uber. So I'm going to give you access to Uber and that, that is intrinsic value for me to do it.
So like, I'm going to go do it. I think Clubhouse did a great job at this right. For a while. And I don't know, maybe if you're like not into deep tech in all the Clubhouses, but it's just like a new social network. And having a Clubhouse invite for a while was like the cool thing to have in the valley.
And so like they had that created this sort of like social dynamic or giving it up mates. It made a lot of sense. Yeah, I think this is like an interesting leeway, cause you both have been in growth, but they also been both in venture. we talked about this in the previous interview is I think one thing that VCs do later on in the stage of a company, let's say you've gone through the seed stage, you're done series A, you found like one of these channels that work [00:22:00] super well for you.
And maybe in like the context of Uber would have been referrals in the context of like Dropbox, it would have been referrals too, right? Maybe booking.com would have been search advertising, whatever it is. And I hear this a lot because what happens is founders come to us and say, Hey, I had a board meeting and my VCs are like, you gotta diversify your channels.
Right now 80% of your traffic is coming from one or two of these channels. So you got to do more That's usually when I, when I figured out I get pretty upset, cause it's just the wrong advice in my opinion, but I want it to first I mean, I already biased you guys, but yeah.
What do you guys think about that? Because I think that's not the right call, but I'd love to hear what you guys think about that. Cause you've obviously been on the other side where you've been in board meetings and stuff, and maybe you're not giving advice and maybe it's the same advice who knows. I'd love to hear what you guys think.
Jonathan Metrick: Yeah. I mean, I think from an investor perspective, right? Putting yourself in their shoes for a second if they're bringing on a new investor for a new round, they've given you a hundred million dollars, $50 million, whatever that is they're looking at their return and they're looking to diversify the risk of that return in theory.
Right. So [00:23:00] if you've got a super defensible moat that you own. From a channel perspective. That's, that's potentially interesting. Right? So if SEO as an example, right. Kinda like the nerd wallet's in the credit Karma's of the world, building a huge motor out SEO and organic search. That's something that I think is really interesting and defensible, but if you've built it around, say Facebook or Google and the platform continues to allow other folks to just come in and bid over, you continue to splash money and not folks want to potentially then diversify you know, the risk of, okay, with 80% of your lead volumes coming in from one channel and other people can come and outbid you.
And you're eventually gonna go up against an incumbent that probably has more money than you how do we think about that? And so I think that's where a lot of these questions of, okay, well, what else, what other channels can we bring into the mix? And I think it's natural for folks to want to diversify the risk, because when you think about it VCs have a portfolio of companies.
They don't just invest in one, they invest in a bunch because one will do well, but a few others won't. And so they're probably employing a similar. Kind of logic. But it is a challenge, I've been [00:24:00] in those meetings on the board and the kind of operating company side, when the board members are like, Hey there's two big channels you've got, but what about all these other ones?
And you're like those don't necessarily work for us. And we've got really strong efficiencies here. So it is a balancing act. And I think on my side, it's you probably have to have one or two lead channels that you can disproportionately win in that are your champion channels, maybe one or two other testing, but you don't want to be spread too thin.
But at the same token I've been in scenarios where a ton of Facebook brands that were built off of Facebook, 80% of the lead volume was there. The Facebook algo changed. CPMs went up and their economics went upside down. So you don't want to be a non scenario either.
Amy Sun: I would agree with that as well. And it's like a, I think there's most investors have had one or two examples that have led a really like a deep impact on them. Like, Oh, I think there were a lot of companies built off of SEO when Google changed the ranking. It just like same thing. the business unit economics and longer really worked.
And so the way I think about it the way [00:25:00] as a growth marketer, as well as an investor is just like, how has the performance on that channel looked over time? Have you have your CPAs over the last few years been really steady? Are you actually diversifying within that channel? Right. Cause like social is so big.
Right. And even like Facebook itself is so big. Oh, you're able to find new AD like your targeting groups and new tactics that work really well. And also what's the age of the channel? Is it your first mover on something like snap and as a result, there's not that many advertisers in the platform yet.
So you're able to get really high returns or like, is it like a channel that's already at maturity? I would say. Cause like, if you're an early mover and then snap ads open to everyone, it just, the CPA's are going to go up across the board. So like how do you think about it? And also I would differentiate the way that investors think about.
Paid acquisition channels, which is like a lot of Facebook and Twitter and Google and things like that. Versus the organic channels. One thing that really like annoyed me as an [00:26:00] investor is people would always ask oh, what percentage of your traffic is organic? When it's like, what do you mean by that?
Do you mean attributed? do you mean like PR like, cause you know, you should be able to attribute most things and like, it depends if you're using an attribution model is gonna look wildly different. So you're not really comparing apples to apples, but that's something that I hear investors ask a lot.
And those, those sort of question behind the question is are people organically telling their friends about the company because they love it.
Nima Gardideh: And that seems to be the much more important thing is can you maintain this level of spend, let's say I'm on one of these channels that maybe you're doing search or.
And then rely on the fact that you have product market fit. So then people will talk about your product because the problem actually is not with paid acquisition, being sustainable the way these algorithms work. It's pretty sustainable. If you want to continue to spend the same amount you're spending for years to come, it's problematic when you have to triple and double your spend that for your [00:27:00] that's when it's problematic.
So what I give this advice to a lot is basically, if you do have one of these channels, it's working, can you master it? And I like that the way you put it, Amy is are you discovering new ways to get days in the channel constantly every couple of months where you're like, actually you're not ballooning over time.
Like if your graph, your CAC is looking like this, and it's going off over time in the last, like 12 or 18 months on a channel that may be bad news, but if you're maintaining a CAC or even maybe getting efficiencies maybe it's, it's a really good one to stick with. And I dunno, I, and this is the reason I think it's bad advice is because if you study some of these like bigger companies, like Dropbox, Uber,, booking.com, TripAdvisor NerdWallet, all these like sort of, unicorn esc companies, what they had done is they had matched their business model to one of these channels, such that they had basically unbeatable systems for that.
Right? So like Dropbox did a great job with their referral system. In fact, it was the main way that grew for the first [00:28:00] five or six years. And I think they did this like Samsung deal where the Dropbox app was on every android installation after awhile. And that's when they got they're sort of second, big growth growth thing, but that's like a partnership thing.
If you look @ like booking.com, they're just incredible at search. There was like 50 people inside of Google.Who's in charge of the paid growth that booking.com does. That's how insane they are with the rates in there. Right.
Jonathan Metrick: I would, I would totally agree with that. And I think that gets back to our point earlier, though, of kind of like being scrappy in the early days to identify what is that channel for you, that there is arbitrage opportunity in, and then as you identify that you're bringing in specialists to not only become scrappy, but best in class in that, and building your moat.
If you're building It just seems to me, just, just looking at these companies that they have one bond being incredible at one of the, one of these channels.
I would, I would totally agree with that. And I think that gets back to our point earlier, though, of kind of like, you know, being scrappy in the early days to identify what [00:29:00] is that channel for you, that there is arbitrage opportunity in, and then as you identify that you're bringing in specialists to not only become scrappy, but best in class in that, and building your moat.
If you're building, you know, booking.com and it's, you know, paid search uh, if it's SEO for NerdWallet, if it's kind of podcasts for ZipRecruiter, whatever channel you want to choose and, you know, eventually you become the absolute expert in that. And I think that curve is an interesting area.
Like, are you continuing to get new wins in that channel? So that may be your lead channel. If not you know, you still have to be looking scrappily around cause maybe you need to find a new channel to continue to get gains. If you're one channel you're in.
Nima Gardideh: Yeah. Do you guys have any frameworks on the exploration phase?
Cause you know, I've, I've used like an 80 20 approach, but I'm not quite sure if that's like the best approach where it's like, oh, the 80% of your time should be spent on channels are currently working and then 20% maybe on exploration, continuous it's either time or money you know, it depends on how you're thinking about it, but I [00:30:00] wonder how you guys have thought about this problem before.
It's It's like, okay, you want to diversify a little bit, but how much effort should you spend on the, on the exploration after you figured out one of these channels?
Amy Sun: I think it depends on the size of your team and the stage of the maturity of the company.
Uh, so in the early days that at Uber, everyone was trying to find new channels, right. It, it was just, we only had two and so we're constantly seeing like what else was out there. And and once we had found oh, okay These are two that we definitely know work. And we started to hire in experts in each of these fields.
Like I was actually, I mentioned I was one of the more generalists early employees. I actually started a new channel team that just only did new channel. And it was only like me and one other person. So there are two people. Um, but it was, uh, we just constantly were testing new channels because a lot of these new channels, like one example, it was like affiliates.
You, it requires a heavy lift to get off the ground. You don't want someone spending [00:31:00] 20% of their time, like trying to run an affiliate marketing campaign because then you could actually run into a lot of issues. And so we actually had a dedicated, static, two dedicated people. And then we actually had a framework of this is going to be our budget for new channel tests.
This is our metric for success. this is where we track the, like our goals and, and at what, like one point after X number. Weeks. We just make a go no-go decision. And then we were able to like cycle through them. And a lot of them didn't work, but one or two hit and those like became large channels and we actually hired an, like a person for life oversee a full-time.
Jonathan Metrick: Yeah. I think that that framework is interesting. And I, I, I love your point, Amy, around it. Some of these channels do take a heavy lift, right? So for example, like how long is that testing window? Like a policy genius. podcasts is a great channel for us, but if, if we had done three podcasts tests and taken the read on the CPA is after three of them, we would be like, oh, this channel doesn't work that well, but it does take a lot of iteration around that.
And other ones influencers, right? Like there was one, one of my friends is a CMO at a, at a direct to consumer [00:32:00]company and they'd three people staffed against influencers. Right. And so I I'd asked them some like, oh, we tried influencers and it didn't work. He said, well, how long did you try it? And I was like, oh, One or two months tests. He's like, no, we did it for six months with a really good sense of you know, who are the influencers. And then we had the influencers refer other influencers, and those are great audiences for us to find. But it wasn't just through just one little task.
You've got to really take a run at it and ensure that the first three, four or five iterations. Okay, great. Maybe you've uncovered something there and there's gotta be early indicators of that. Of course you can't just waste money. But I think I've seen a lot of startups take a run at a channel, do it lightly, and then be like, oh, this channel doesn't work.
So there is that balancing point between really validating. Did you take a solid run at it? And it's okay if it didn't work, but you want to make sure you've got a solid read on the test, not, oh, maybe it could have worked, but we didn't execute it properly. We didn't run it along.
Nima Gardideh: Yeah, there's so much execution risk with these. Like, [00:33:00] All the channels, like literally every channel there's execution risk because they're just hard to crack. One thing that comes comes up to this is basically that you may be wasting time doing all this exploration if you don't have product market fit, because no matter what you do, the channel is not going to open up for you.
this is a tough one because I think it gets, takes us to the other topic, which is growth in the early days has so much to do with product. Actually figuring out what part of the market will make sense for the specific product that you're building, or like maybe your messaging is completely different of the tech can be applied to a whole different slew of markets that you hadn't even thought of in the beginning.
Do you guys think that that's like the founder's job, the product manager's job in the early days? Or do you think like a head of growth can come in and just also own part of that? Like how you. How do you think about that? I'll have my own opinion on this, but I'm curious if you guys think.
Jonathan Metrick: I think it's, I think it's everybody's job, you should be watching your NPS as a marketer or growth growth person you want to be promoting products that people like. Right. And so there's [00:34:00] no sense in spending a ton of money on a product, if you're bringing in.
Users that other don't end up buying the product because they fall off in the conversion cycle or they buy in the mid turn quickly, or you've got a low referral rating or a low NPS score. Right. At that stage, you've gotta be like, all right, does it, you've probably got a leaky bucket. Should you just pour more gasoline into it?
Or should you try and actually improve the product, improve the product market shape, get it in front of consumers who actually like it. And I think there's a bunch of people in the company that can identify that. But as a growth individual your job is to drive revenue, right?
It's not just traffic, it's not just you know, awareness and to drive revenue, you've got to get people to buy the product and hopefully refer it and bring in some other people to buy it. I think the ownership is. How do you want to not mix, but definitely with the growth folks.
Amy Sun: And I would say like for the early in the, I was a growth person, but I, I would say that in the early stages, you might not want to hire someone dedicated to growth until you have some [00:35:00] sort of repeatability demonstrated like with the product. And the way I like to think about it is I do like this, like reality, like gut check thing of like, I think some people are like, okay, so like CAC is like a hundred dollars. How do we feel about that? How can we decrease it? I'm like what does that mean? If CAC is a hundred dollars, What's your like CPS, how many humans are seeing your ad and like not clicking on it and not converting. If you actually look at those numbers and you're like, Hmm, I think there might be an issue here where like, all of the hundred people that land on our site only one person actually signs up, like, why is that?
What happened to those 99 people? Do they, and, and really like getting to the bottom of that either by just calling people off, doing user research, spending time with their customers and like understanding all here are all the reasons that I didn't sign up. And it's like, oh, well, like some of those are product issues.
Like maybe we should go back to the drawing board and like with the product team and figure it out. And same with retention, right? It's You know, a hundred people didn't want my thought one thing on my platform. And then [00:36:00]they never came back again. why, why didn't they come back? And until you can get to the bottom of those like core questions around product market fit then you don't, you probably don't want to like pour a ton of gas on the fire, but.
Have something that you were like, okay, I think that people like this, and I think that they will continue to use it. Then I think it, it makes sense to bring in growth.
Jonathan Metrick: And back to the point of, of referrals, right? Like, Are you, you've tried this product, are you willing to recommend it to somebody else or not?
Right. And to doing some qual on, have you ever recommended this product? Why not? Right. If they're not willing to recommend it to a friend like that, that's a problem with the product you've got, right. Or you've got the wrong user and, or both. And I think, again, that gets back to that piece of referrals is such a valuable bit of information from a business trajectory perspective.
Because on Uber's point, if everyone's referring their friends they're doing a lot of that work for you, but it speaks to how satisfied they were with the product and their willingness to put their own social capital on the line to bring their friends in.[00:37:00]
Nima Gardideh: Yeah, what's interesting is I'm sorry.
I agree with you guys there, there is a question here. I think that would be a good wrap up on this topic. I think both of you have touched on the process that you go through of okay, we're going to attempt to unlock a channel and we're going to spend a reasonable amount of time on, on money, on, on the channel.
What I've actually realized that there's not a lot of written work about how to run this process. Do you have, do you have people that you follow on this where you just respect how they're talking about how to run a growth process? For me, it was discovered by being in enough startups where I'm like, oh yeah, there's like this like thing everyone's doing.
And there is like, basically some videos you can maybe watch on YouTube and see people talking about it, but there's no like, cohesive here is like the model for growth. Maybe, maybe Brian Balfour stuff from the old days, but I'd love to hear what you guys, where you guys learn this stuff. Or if you just have to solve on the job.
Jonathan Metrick: Yeah, my side, I think it's more than the job, right? Because the companies are really different, right? Like even, even with my [00:38:00] portfolio, I have, I've worked only with FinTech companies, but I couldn't say every single company should have a 20% allocated budget to testing and 80% should be business as usual tested channels.
Right. That might apply for a scale-up that's in series C or series D, but sure is hacked to Amy's point. Doesn't apply to a series A or a seed stage where it should be maybe 80% experimentation. And I think you know, there are nuances according to stage and then the type of company you've got. But you've got to be willing to test these pieces.
Like I always think it's critical to have at least some share especially in a scale-up or a tech mode that is allocated to experimentation. My heuristic typically similar to yours, I think Nima is 80% of what you know, in 20% testing, as you're scaling from series C to series D less so in the earlier stages. But that's my take.
Amy Sun: I have, uh, two frameworks, that I used to, when I, when I was leading the new channel team, and one of them was just like, talk to your customers, like really understand your customers and everything about them. Like[00:39:00]how do they spend their mornings? Like where do they hang out with their friends?
Like, what apps do they use? Like, how old are they? Like, how do they spend time with their family? All these kinds of things. And once you like, get that sort of understanding of your customers, then you're like, oh my customers are 14 to 24 and they love TikTok right. They love making videos.
They like, these are the surfaces that they spend a lot of time on. Like they hang out at school after what their friends and like they get rides from their parents. And so as, and then you can, I kind of like triangulate. Oh, okay. Like if TikTok, Snaps and stories might be good channels because these are places that they already are.
And this is like, Uber you heard from the drivers that they commonly peruse the gigs portion of Craigslist to find jobs. They con they like they spend a lot of time driving around in their car. Radio might be a good channel for, for that. And just really deeply understanding that and oh, what kinds of neighborhoods do they live in?
Maybe we can do some out of home in [00:40:00] those neighborhoods. Like all of the like trying to meet the customer where they are. And then the other thing is like trying to be. As early on new tech platforms as possible. And in that, in this scenario, it's like really hard to find writing on it. Cause like the writing will just be outdated really quickly, right?
Like what are the new emerging platforms you mentioned Clubhouse, the new emerging platform take talk obviously is huge now. Are people more people on mobile? What are these like massive like kind of trends and can you get into an alpha test to be the first advertiser on that platform?
That was a huge strategy at Uber. Like we like convinced Snap let us be an alpha tester on some of their ads. And they performed really well because there's not, it's not competitive. Right. You're one of and I think it's easier when you do have a brand associated with you and you're like, like an Uber and Airbnb or Dropbox or something, but even the smaller startups. Maybe you're friends with someone who works at take talk or something and we can get you in on the new app, install ads or something.
Nima Gardideh: It's fine. That's so important. Like we, we, [00:41:00] cause we spent so much on Facebook, we get access to these like sort of new ad units and like they're the best part of all we do because we spend like millions and millions a month as a small startup, you don't get access to that unless you work with us. So then we can blow you up on that one ad unit.
it's an important one that I think Andrew Chen wrote about this. Actually he called it the laws of shitty click-throughs, the law of shitty click-throughs. If you, if you actually search that, you find his old article on this and it's basically the concept is as, as, as you mentioned, right, there is a new, new platform.
There's not a lot of bidders in there. So CPMs are cheap and they're more bidders getting their CPMs will Lou nap. So then your cat just shoots off. They no longer make sense. So it doesn't follow that you should be the first as, as early as possible in, in a bidding war for, for people's attention. And the latest one, I like we've tested a bunch recently is NextDoor actually, which has been a weirder one.
And it works for some companies and it definitely does not work for some companies, but there's super early [00:42:00]stage. Like the, their interface needs a lot of love. But it's an interesting one right now we're at a ton of our clients are testing for instance, other than TikTok and Clubhouse doesn't and have ads yet.
Jonathan Metrick: Yeah. It's an interesting theory around like we've found the same thing you know, working at PolicyGenius or even across our portfolio is, getting in front of that audience cheaper. Right. And where you find that lower CPMs is imperfect channels, right. That's where channel arbitrage exists. Right?
The second that I see like an IBM or a BMW advertising on a channel it's over, right. The fact that they could podcasts would be an example, right? So I PolicyGenius we'd be trying to get on a lot of podcasts, but one of the big ones was the daily. Right? And like, you take a look at who advertise on the daily Coca Cola, IBM, these huge car companies.
And if you're trying to outbid them from a CPM basis, not going to work right. Where you've got to be as areas that are hard to track, hard to scale, because the bigger brands we're working with agencies or these sorts of things, won't find that attractive. But like, oh, we don't know [00:43:00] how to track.
And we know how to validate it. That's where the channel arbitrage opportunity exists or in a beta where not everybody can access the inventory. So I think thinking about getting in front of your customers cheaper in a channel that has lower CPMs is usually a way for, for arbitrage.
Nima Gardideh: Yeah. And you know, what was interesting about this for a while?
And I think both of you were part of that way for a little bit and were paid was just not the thing because Facebook and all these different ops, still hot, real ways you can organically grow on there. And there was like a slew of startups, I think it Airbnb. And then we're part of this, this sort of stage even Dropbox in many ways where organic growth was a thing, I think growth hacking came around because that was possible where you can go on one of these platforms and then figure something out.
So I started telling people, we'll just share your stuff a lot. And there's all these startups, some of them failed. Some of them succeeded that came about. Do you think that that's going to come back at all? I feel, I feel like it's not, but I, I, I [00:44:00] wonder how you guys think about that now that the platforms have just fundamentally changed.
So then you have to pay for every question then using the network.
Jonathan Metrick: I think one of the things I've heard a lot is. The sophistication of these platforms through AI, it's just getting so much more robust. Right? And so there used to be a lot of hackability in paid search in Facebook, where if you were into the channels you know, and you were up at 5:00 AM posting on Craigslist.
Like There was a lot of these abilities to, to hack into growth. I think with AI, a lot of that's being removed. And I think that the pendulum is shifting back into the creative space and there's opportunity now to differentiate with kind of you know, dynamic creative, rapid creative, native, creative, right.
We're kind of, you got folks that look like they're, you know, it's probably an ad, but it's actually, the influencer was paid to speak about a product organically business insider you know, where it's like a mix of like, it's kind of news, but it's like kind of advertorials you know, these sorts of creative executions, I think are now a little bit more where [00:45:00] a lot of the opportunity sets because a lot of the hockey like get in front of the right audience on somebody who's established platforms, I think has been removed because of the sophistication of the planet.
Amy Sun: I definitely feel that that's probably the case on like Facebook and Google, which are the two dominant, like there's the way more dominant than when I was a growth marketer, today. But I think there is still opportunity to do creative things on like the newer platforms that are not as sophisticated with their ad models.
I think tick-tock is one of them. They just have so many impressions that like you can actually hack some of their organic loops and their challenges, if you know how to do it the right way. Like I've had a couple of portfolios find just a ton of portfolio companies. I find a ton of success on TikTok.
They don't exactly know why they're going viral, but like, if something does go viral, that's like millions of impressions, right. For very little cost. So I think there's like a lot of like interesting opportunities there, but you have to be willing to like really get into it and then try a lot of stuff that doesn't work.
And there was a [00:46:00] question on the chat earlier about like, What are some referral channels that for folks that don't have a ton of venture funding and I think that there's a lot of like creative, very creative ways that you can get people to talk about your brand, get something out of it without necessarily having to spend.
And Dropbox is one of my favorites when I was, when I was in college, Dropbox was like, just getting started. And they had this thing where if you install Dropbox, if you have your friend install Dropbox on another computer, you got like some free storage. And like my friends and I uh, one friend in particular, all of the like public school library, computers, she just installed Dropbox.
I'm like all of that kind of interesting because like you do, like now the next person that's gonna go on that computer is going to see Dropbox and potentially like use it. And, she like probably installed Dropbox on like, dozens of computers.
Jonathan Metrick: It's interesting. I think I, yeah, let's get back to the debt creative with the rewards.
Right. And so if I was an interesting [00:47:00] one, there was a period of time when you know, they used to give up these t-shirts and everyone wear them to the gym. Right. And it was kind of like a comfy shirt that you wear to the gym, but you're like, oh, like you're doing free branding for them. The t-shirt that cost $5 for them maybe shipping and a bit more, but it's not that expensive, but it's a way to get you were referring and these sorts of things, so you can get creative with the referrals and not have to break the bank on it.
The best use cases are giving your product away for free, right? Like space on Dropbox. Doesn't cost a ton of money. Fractional shares. Maybe they cost a couple bucks, but they sound pretty cool. Like, oh, I'm giving you a fractional share of Apple. Maybe it's $10, but it still sounds pretty cool.
Nima Gardideh: Yeah. I think that the referral thing in general seems like it's, it sounds like you're wasting money, but if you have it right, then you should be making money almost immediately. I mean, of course the context of Dropbox and Uber was that they had so much cash that they could do whatever the hell they want it.
Right. But I do think that there's a sustainable way to build that. So that program, right. There was another question in here and I actually think [00:48:00] that's a good one to think about a little bit, especially because you've been on the investor side is how, how to think about CAC. Do you guys think that the head count of the people that are working on marketing should be included for instance, in CAC?
Or do you think it's just the pure numbers you're looking at last click paid on Facebook. This is it like how, how sophisticated essentially to you, do you think people or companies should be when they're trying to calculate these numbers?
Amy Sun: That's a really good question, because I know I've seen companies do everything like you kind of really have to get into it, but I would say like one number that is just useful to think about from the investor's perspective. It's like what just the aggregate spent on all sales and marketing, including head count, including media, including everything over the aggregate new sign-ups.
So like organic, everything, just all the new signups. And so like, so you kind of get that [00:49:00] like, oh, generally this is how much it costs to grow. And I think that, that, that kind of. Eliminate the head, like sort of headache that's associated with like attribution models and everything like that. And because from the investor's perspective there, we're just trying to understand like, oh, how much will it cost to grow?
How much did it cost to grow? Up until now? If we invest X dollars, like how many new...
Nima Gardideh: So you think basically like blended CAC, including salaries is an interesting one to look at. The problem with that is it just looks bad, very bad in the early stages of a company. Like the, the think like investors don't expect that to be as nice and clean.
Do you think like you're okay with it being, feeling a little bit, not performance in the early days? Cause you probably like or higher than some parts of the organization. You haven't fully figured out how to make your organization efficient on that front yet. Like you think that makes it harder [00:50:00] for the founders to raise money?
Jonathan Metrick: I was going to say, I think in the early days I don't think folks are looking at your CAC and they go, are you profitable? Like they know what stage you're at. They know that you probably need to over hire. You're investing in tech, you're investing in marketing in the early days. And you know, they're looking to the path to profitability, but they're not expecting a series a to be.
As efficient from a cost perspective as you would at a series C. Right. And I think um, you know, so they would know that if you're including head count and your business that scales, and you need to add head count to convert and drive revenue that may warrant you actually putting that in CAC versus if it's a completely SaaS company and it's more digital my tick as well, I think as you know, depends back on the business, right? So it's like if you're including head count and your business that scales, and you need to add head count to convert and drive revenue that may warrant you actually putting that in CAC versus if it's a completely SaaS company and it's more digital, there are different factors that would weigh in.
I saw it in the comments, there's a comment around you know, retail space being rolled into CAC and, and in [00:51:00]some of the retail startups, maybe that is a, if you got a heavy retail footprint, that may be a more pertinent factor, including your pack versus a completely e-commerce company that doesn't have.
Right. Or they just have their office space. So I think for me, I try to think about it, like what elements are variable costs that you're going to actually continue to incur as you scale that should be in CAC. And I think things that are just kind of more fixed or flat or investment, that, that I would want to not include because I want to get a sense of what is the incremental value?
What do you You spend a dollar incrementally to bring in new customers. What are you getting out of that? I'm very interested in looking at that trajectory.
Nima Gardideh: Yeah. That makes sense. Cause even in retail, we w when I was at Franklin, if we thought of it as, as marketing spend, when we, when we got real estate, because it was free impressions, right.
Or like, it would have been free impressions if we hadn't included it. And it also followed then whenever we were launching a store, that's the sort of click through rates would increase in that [00:52:00] market, on our ads. So it was just so obvious that this is totally like a marketing play. All we have to do is just launch more stores.
And that's kind of what Warby and Casper did for awhile. It's just like brought down their CAC a little bit. Right? It's, it's a hard one, I think for B2B, for me, at least when we talk to companies, I expect them to have the sales staff and the marketing staff baked in there, because it's such a big part of closing the customers literally at the part of your funnel.
So I know your funnel is variable, unfortunately, and that's an order to scale up your sales staff and your team to continue doing the work. But for B2C, maybe it's a lot more clear where, okay. Yeah. Dollar and dollar out. It's much more obvious and I get it back to like what, what is, what is required to generate the revenue on the incremental sale.
And then that can be to your point, like in B2B, if you're bringing in these leads, but you don't have the sales folks to actually convert it, that's a variable cost. Right. But if you're on a SAS platform, you're just bringing an incremental, everyone can convert directly online. You don't need that. [00:53:00] Yeah, I think it makes sense.
It makes a lot of sense. For us, like the way I like to think a lot about CAC is I'd love to hear, hear what you guys think about this. Cause we basically we'll run ads on like these different channels, right. And then we'll look at CAC on a per channel basis a lot, but then when you start scaling up, after a certain point, you can't really trust the networks anymore to tell you who they're getting, because there's going to be overlap in the audiences and stuff.
Do you believe in like pure last click attribution essentially? Or do you think there should be more complex models that people should build for their startups, especially in the later stage stages. I'd love to hear how you guys thought about that in the past.
Amy Sun: Yeah, I think there's like two that are, that come to mind. The first is if you're doing any sort of offline, like TV, radio, you can't just let last, just doesn't work. But like, those are actually can be very performance channels. And like once the digital stuff is like saturated, I think it can you know, lift all, if [00:54:00] you're doing like some sort of TV and radio and out of home, it can actually lift the performance of all of your channels.
So actually trying to figure out how to attribute all of that is complicated, but really important because otherwise you would never do it, but it clearly works. And, and then the other thing. I think it's super important early on to think about marginal CAC. And that's not easy to write, but if you can do like small attribution tests I put in extra incremental amounts of dollars, like how many additional people signed up.
And I think it's mind blowing for some people, like the first time you do it, you're like, I just put in $10,000. And I only got two additional signups from this channel. And I think Uber ones and we were like, oh my God like, we should just give this person a car. Maybe even like, it gets that expensive, but it, especially if you're doing a lot of volume It's sometimes hide within the aggregate blended numbers.
So yeah, it's like sometimes you want to share [00:55:00] the blended numbers with the investors, just be easy, like give them the holistic, like end-to-end view of the business. This is our gross margin, just our EBITDA margin. But like, when you're thinking about like the changes that you make we think about the marginal dollars you're putting in the system and it's like, if I have an extra $10,000, do I put into Facebook, do I put into this new channel?
Do I put it in part referrals engine? Do I put it into like text, text messages for existing users? Things like that.
Jonathan Metrick: Yeah. I mean, I think on the attribute now I was standing the attribution for like, that's, that's the age old question that we continue to try and get better at. Right. And you will never, by the time I left PolicyGenius, I was on you know, V 10 of our attribution model in three years.
Right. And the reason why, as you're continuing to take new information, testing, new channels, and some of them you can track quantitatively very easily, digitally others, you can't. Right. And so how do you think about. Incorporating the spend doing lift testing in certain geographic markets as a way to do that, or you know, how do you really track the impact of. Uh, Subway campaign, [00:56:00] right? Like this is an example. I got policy genius. We were among all the other startups in New York city who does these you know, brand takeover and these trends, these takeovers, and you see a huge lift in your awareness and knowledge of the business. As soon as you do that, it's almost like a pre and post of doing that in terms of awareness in the market.
But how do you quantitatively track that? you know, there's a big lift in the month that we did, that is all of that attributed to Subway? And you know, for life insurance, we couldn't put it on sale. So we couldn't put like 20% off life insurance this month. So there was no, there was no, there was no code that we could attach to that.
So you've got to get creative with how your attributes, that volume of, you know, do do you take a look at incremental search traffic? Do you look at the incremental lift in this market versus other markets, those sorts of pieces. And I think those are elements you need to look at, because that gets back to our comment earlier about channel arbitrage, the harder it is to track some of these things, there are opportunities.
Folks who are willing to lean in and try that versus, oh, it's too hard to track. We're not going to do it. So it's something to think through, I think, and revisit quite frequently in [00:57:00] terms of attribution modeling.
Nima Gardideh: It's funny what you mentioned on this. So two things came out of this from you, which one is a common thought where I think really growth people tend to be just really good capital allocators.
In general, that seems to be what the job becomes after a certain extent where you're just making budget decisions at that point. And then at that point then you're like, what, what number is am I looking at? And then how smart you are with looking at the numbers tends to be the reason you win.
But the other thing is, and it seems like you both have experience here on the idea of shared experience when it comes to advertising because the industry that we we've been a part of. Before Facebook and Google and these sort of very clear dollars in dollars out sort of channels existed.
Performance marketing was like a small niche. People couldn't really do what they're capable of doing now, but there was this concept in advertising where if you had like a Superbowl ad, everyone was talking about your brand afterwards, because everyone had watched the Superbowl. Or if you have an ad in the, I [00:58:00] don't know, in between friends, everyone was talking about the day after, cause everyone was watching friends, right?
So that doesn't exist anymore as much because media is so this joint and now it's not four channels where all families in your, in your city are watching the same thing at eight o'clock anymore. That's become less and less of a thing. But when you are able to get. Get to a point where you can create a shared experience moments, all your other channels become more performance, all of a sudden.
So the Subway app is a great example of that. You've done it quite a few times for our clients. We'll, they'll run a subway ad a click-through rates will go up that whole month because everyone's talking about the ads and it creates this sort of like old school way of like brand advertising. And then it ends up doing, creating like a lever for our paid and performance work, which is very interesting. Combining it together.
Jonathan Metrick: I think an interesting note on that um, two is, is, you know, um, and this is what gets so interesting around how do you value brand marketing, right? And performance marketing is one area and you're taking a look at that. You can [00:59:00] track it $1 and this is what you get out. But you know, as an example, Is it easier because you've li mean good and done some brand to recruit talent, right?
Like there was a, there was a moment where, you know, back at PolicyGenius, before we did our first subway campaign, I go to these growth meetings and I'm like, oh, I work at PolicyGenius. They're like, what is that? What are you talking about? Like Tell him, you, you spend a minute explaining what the business was before you even got into what's going on following that.
And I was like, oh yeah, I totally saw the ad. It was super cool. You did the poetry on the subway. And like, what is the value of that in terms of conversion for getting in front of people? And you know, we can do a ton of consumer research at PolicyGenius, obviously. And speaking of folks in New York you know, I would often ask folks when we're leading our discussion groups, like, how'd you hear about us?
Nine of the 10 people in New York would say are subway. Whether that was know, the only thing they'd heard of us, but they would mention that what value is that? Nine out of 10 in a qualitative kind of setting. And, you know, as you go out to these, these pieces, so there, there are intangibles that are attached [01:00:00] to really strong execution in a channel that may not be directly trackable, but what value does that bring to the business?
Nima Gardideh: Yeah, that's a hard one to figure out, because like, we think about this a lot where in the context of PolicyGenius, maybe it would have been like somewhat easy to like attribute and do like market differences where you're like maybe not advertising the same way in market day versus market B.
Some companies are so global that this is not possible or just pure software. So like the location won't matter. Like they might see an ad and they live in California and then sign up or buy something in New York from like a designer in New York. So there was like all these complexities and marketplaces that we work with.
And it's, it's a topic that is very hard to nail. I think there's just not a lot of written work around how to think about the different sort of aspects of attribution. And should you like do first to last or do partial? And it sounds like, and correct me if I'm wrong that you guys kind of like solve for whatever you're doing at the time.
Cause like it was unique to your business. So here's [01:01:00] like what I'm going to think about how to nail it and for my business. Does that sound right?
Amy Sun: I would say attribution is definitely like an 80, 20 rule kind of a thing where like 20% of the work like is like 80% of the value we went into like an attribution black hole so much time, so much, so many resources and like, not that much help at the end.
And it just like, it's just like so complicated and it's like, do you really want to like deploy data science resources to like really do this? Or like put data science resources on something like search pricing it's just the trade offs are get to be pretty steep with the amount of effort that you really have to put in.
One other thing though, that you mentioned the shared experience of like that Super Bowl ad. I want to like, bring that back to Jonathan's point earlier about con. I think like content and like morality now is that shared experience. like that example of the TikTok video of the cran, raspberry juice and dog [01:02:00] face, like everyone watched that, right.
It's like hundreds of millions of people probably have watched that and maybe they got cravings for cranberry juice afterwards, but I think there's some really cool things. And you can do with content. just thinking back to the like days at Uber and like when some of the portfolio companies that I've worked with at Sequoia, like when Uber put Uber kittens was like, just so viral of a moment.
Right. And it was like very cheap to do like. I think it was a woman named Jen and the Seattle office came up with the idea. It was a local like a SPCA, like cat adoptions and Julie, Hey, like, why don't we just take some of these kittens, put them in cars and hopefully you could get adopted, but the amount of vitality that, that generated and like the free ice cream thing, it was also just like super huge.
If you can create these moments of like surprise and delight, then they, and they can generate so many just free impressions, kind of a, to the something that everyone was talking about. Like a super bowl ad.
Nima Gardideh: Yeah. I mean, I feel like it's a very [01:03:00] hard, hard thing to do. I think it's, and this goes back to what Jonathan was mentioning earlier.
I think that our world, the world of marketing growth is turning more and more towards creativity in general, because there was a point where. It was about basically being a day trader on these platforms and then finding all the different, weird bugs. And that's how you want on Facebook and Google and all these things.
And then now, because of the AI, it's becoming essentially who has the better creative and the better product, and you're just going to win because of that. I do love that about the industry that this shift that's happening, because there was a point where I was feeling quite sad about where the world is going, where it's just like all these numbers.
And you just have to like when at, at bidding on the right auctions and that's how you're going to grow brands. And it has nothing to do with the merit and the beauty and the creativity and the products they built. It's definitely shifting again. But what you just described sounds way harder to nail of like creating an experience.
That's going to blow off in a way that millions of people will will, will watch it at the same time and have a shared experience. [01:04:00]
Amy Sun: Very hard to play book that. And to scale that from company to company.
Jonathan Metrick: Yeah. The, the number of agency briefings, I'm sure the agency is having like, we want to create a viral ad. We want it to go viral. Like that's our, that's our crew. That's our creative brief. It's like, okay, got it. Aspirations.
Nima Gardideh: Cool. I think we don't have to go all the way to the end. This seems like a really good ending to it. It's just like nice positive note of where the industry's going. thank you so much, Amy and Jonathan for doing this. This is really fun. And just talk about marketing with you guys.
Jonathan Metrick: Yeah, absolutely. Always a pleasure,
Amy Sun: Thanks for having us.
Nima Gardideh: Thanks for listening to our show. Get our episodes. As soon as they're released, just tap that follow or subscribe button, wherever you get your podcasts.
Plus if you want to join our live discussions where you can ask us questions, as we record, sign up at pearmill.com/hypergrowth-podcast, We'll see you on the next episode, on "The Hypergrowth Experience".