🎧🍌 Gokul Rajaram | Lessons from Zuck, Jack Dorsey, Sergey Brin
Defining your ICP, how to size a market bottoms-up, the importance of a fast response time and how to do it, when big acquisitions can go well, and why startup titles are overrated
Gokul Rajaram is an early stage technology investor. As a product leader, operator and board member, he’s helped build seven generational technology companies, including Alphabet, Block, Coinbase, DoorDash, Meta, Pinterest, and The Trade Desk.
We talk about lessons learned from Zuck, Sergey Brin, and Jack Dorsey, when big acquisitions can go well, how to define your ICP, why you should always size markets bottoms-up, the importance of a fast response time + how he does it, the evolution of seed investing since 2007, and Gokul’s hot takes on titles at a startup.
Support this Episode’s Sponsor
WorkOS: Building an enterprise-ready SaaS app? WorkOS has got you covered with easy-to-integrate APIs for SAML, SCIM, and more. Start now.
To inquire about sponsoring future episodes, click here.
If you want to help me make the podcast better, reply to my posts on Twitter and LinkedIn letting me know what you thought of the episode!
I’m also a bit behind on emailing these out - there’s 5-6 episodes from the past few months that I haven’t gotten a transcript out for yet.
👀 Watch out for these bonus emails over the next few weeks.
Timestamps to jump in:
2:14 Common thread of success between the founders of Google, DoorDash, Facebook, and Square
5:50 Gokul’s first job in Silicon Valley
7:46 How Serendipity led to PMing Adsense, one of Google’s biggest products
12:20 Lesson from Sergey Brin on reducing friction before a products magic moment
18:50 How Zuck used founder mode to beat Google Plus in 2011
22:51 When big acquisitions can go well
24:47 How Gokul switches from startup helper to public company board member
28:09 The evolution of Seed investing since 2007
33:27 How to have a fast response time
37:40 Lessons from Jack Dorsey always selling
39:54 How to define your ICP
42:40 Using bottoms-up to size a market
44:05 Why Director and VP titles are bad for startups
Referenced:
Follow Gokul on Twitter / X and LinkedIn.
👉 Find on Apple, Spotify, and YouTube
Transcript - Read on Rev.com
Find transcripts of all prior episodes here.
Turner Novak:
Gokul, how's it going? Welcome to the show.
Gokul Rajaram:
Turner, thanks for having me. It's great to be here.
Turner Novak:
Yeah, I'm excited because we've talked about this for a while. We finally got it on the books. You have helped scale, it says on your website, seven generational companies. That's a lot. Are there any common threads you've seen just between all those businesses, the founders, the products, one big common thing or maybe multiple, that just leads to a generational company?
Gokul Rajaram:
It's a very good question. I've been lucky to observe these companies, and they were fairly young, especially Google, four years, five years after being started. I joined Facebook five or six years after being started, and the others also. I think one of the common themes I have seen across all of them is the fact that the culture of the company and the things the company focuses on were the things that were unique to them and were the things that were the right thing for the market they played in or the product they built. For example, at Google, Google was a technologically-focused company. In other words, what mattered at Google was building the best technology products. That mattered because ultimately, the products that Google competed with, the markets it could compete in, technology was what it took to win.
Unfortunately, they were not able to win when technology was not the driver. For example, Google Plus, it wasn't about technology, it was about something else, they were not able to win. Similarly, DoorDash is a very operational-focused company. They have a lot of product innovation, but it is infused with operational and really deep understanding of details, and that's what it takes to do well in the food delivery space, for example, or delivery space. Facebook was a very growth-oriented company and metrics-oriented company, and that was what it really took because you can't just throw technology and expect to do well because it's a true consumer experience. You've got to be fixated on subtle things that move the numbers, that move retention numbers, that move MAU numbers, etcetera.
Square was a very design-centric company because a big part of the value prop was how do you make the product frictionless so that SMBs can sign up with zero CAC and the experience is better than for a consumer product. Each company, each founder emphasized things that were different than the other. I always say that if you put any of those founders in charge of the other companies, they wouldn't be the same and they wouldn't probably win. The founders and the culture and the values all tied together almost, what they call it, in a steel thread versus a steel-man-like thing where there was a steel thread passing through all of them that was the right thing for that company and that market.
Turner Novak:
You could maybe call it not founder-market fit, but also founder culture, founder competitive advantage fit. You're just leveraging what your strengths are and really leaning into what you are uniquely good at.
Gokul Rajaram:
Well put.
Turner Novak:
Interesting. I was going to ask you if there's anything that made them unique, anything in particular of certain founders or companies?
Gokul Rajaram:
There are many paths to greatness. That's the thing you realize, that the thing that was common to all of them was the uniqueness and the fit. Like you said, the founder market fit. Again, I think the founder was the right fit for that market. They were just the right fit because they cared about the problem, they were deep in the problem, and their specific set of skills, like Liam Neeson said in Taken, "Their particular set of talents were a perfect fit for that market."
Turner Novak:
They're like the Liam Neeson of social networking or SMB payments.
Gokul Rajaram:
Delivery out of payments, exactly. That gives them a competitive advantage. Gives them a competitive advantage from day one.
Turner Novak:
I know this is going back a long time. I've heard the story a little bit. You've never actually told me, but I've heard on other podcasts, getting your very first job in tech in Silicon Valley, how did that come about?
Gokul Rajaram:
It was a company called Onetta, which was an optical networking company funded by Sequoia and Matrix, if I remember correctly. It came about because I was a business school student on the East Coast, and it was in 2001, where it was in the aftermath of the NASDAQ implosion. Turns out that when all the dotcom companies died, there was one sector that was growing, which was telco. If you remember... You might not, you were too young then.
Turner Novak:
Yeah, I was in fifth grade.
Gokul Rajaram:
Yeah, exactly. There was a mini telco boom in the fifth grade because there was a bunch of regulations that loosened, and these things called CLECs, competitive carriers basically, came up to challenge the monopoly of back then, the large carriers. Because of these new customers and new technologies like optical networking, in fact, we know Coastline had some of its biggest wins during that time, [Inaudible 00:04:51] and some others which was sold for billions of dollars. But there was an optical networking boom. That was the one space there were jobs for product managers, and I really wanted to be a product manager. So instead of applying to the consulting or banking firms, I somehow got to an alum of the business school whose husband had started this optical networking company and got to them and convinced them to offer me a job. That was the only job I got. Took the plane out to the valley and have stayed ever since. So whatever it takes. I knew I needed to get out of East Coast and be where technology was created, I felt.
Turner Novak:
I know one of the companies you're at was Google. It's one of those stories where I've heard you say it on podcasts before, but you built one of the initial ad products. How did that come about? I know there's an interesting story on how that evolved and how you got involved with it.
Gokul Rajaram:
The biggest lesson for me from that story is the power of serendipity and being open to opportunities every day. I started at Google, I was working on a product called Syndication of Search. So basically search is all syndicated to other websites like AOL, Yahoo. I think AOL was our biggest partner with smaller partners like Ask.com, etcetera. We provide them with search and also provide them with search ads.
Turner Novak:
That was a pretty big part of Google's business at that point in time, right?
Gokul Rajaram:
Exactly. Web search syndication, ad syndication, search ad syndication. Every day at around 6 PM people would go home, but there'd be still enough people working, so I'd just wander around the office and I just came across in a room a few engineers who were working on stuff. I was like, "What are you working on?" Turns out they were working on a product, a project almost at that point, of which Sergey Brin the co-founder had asked them to, which is just like Google, the search engine, when you give it a query, it finds the best pages that match that query, they were taking the opposite. They were taking the search index and flipping it, which is given a page, how do we find the best topics or keywords for that page? And then using, we already had ads from AdWords, how do we use AdWords ads to show ads there?
So they were running a bunch of pilots. They were just starting trying to get a few web publishers and seeing if the contextually focused ads, because those ads are contextual based on the content of the page, would perform better. Very early on they were trying to build this engine, run ads, and they wanted more help. They wanted product help because they were all engineers. I offered to help them on the product side and they said, sure, they wanted help. They said, "Aren't you working on this other thing?" I said, "That's okay, I can also do this thing in addition." That's basically what I did. I did two jobs for a while, for a few months. I was in presentations doing the presos with them, so the exec team-
Turner Novak:
They probably thought you were on the team.
Gokul Rajaram:
I was on team X, but there was enough things going on at Google. That's a great thing about being at a growing company, that there's always opportunities and it's up to you to be proactive. After a few months, my boss saw that this was taking more time. AdWords was also growing fast and so put me on this full time. I was very happy to just work on that full time. That became a product of Google AdSense, which was a large part of Google at that point.
Turner Novak:
Was that their external website, like if I wanted to monetize my product?
Gokul Rajaram:
Exactly. Back then blogs were huge. So if you have a blog, you want to make some money, you basically just sign up for AdSense, put a piece of JavaScript on your website, and as soon as that page loads and someone clicks on the ad, you start making money.
I think we had, actually back then, fairly large, millions of monthly active blogs using it. I think it got to a hundred million in gross revenue in a couple of months, a billion in gross revenue in a year, year and a half. So it was exceptionally fast going. It was global from day one because it was just a pretty remarkable product. I still get where people who used to run blogs, I still periodically will get inbound saying, you helped me pay for my kids' education or helped me pay for my house, et cetera, with AdSense. So it's a great feeling. We helped people make a living doing that.
Turner Novak:
I know you have another fun story around when you were at Google working on, it might've been this ad product, it might've been a different ad product, but you're working on the signup flow, getting advertisers on, and there's an interesting thing. I think it was Sergey Brin who was like, "No, let's do this instead." What was the story there?
Gokul Rajaram:
Exactly. I think it was actually right around the AdSense product launch, which was web publisher could sign up, they would get a piece of JavaScript code and they would place it on their site and then they would show ads. However, when we initially were building the product, we were worried about all kinds of publishers signing up and getting access to this code and then showing Google Ads on a not safe for work site or a site that promoted violence, all kinds of stuff. We didn't know what we didn't know. We said we should put the publisher on approval queue once they apply, and only when they get approved will they get this JavaScript put on their website. So we built this approval queue, we were like, "Who would object?"
Turner Novak:
It seems like a good idea, yeah.
Gokul Rajaram:
It seems a great idea, of course. Google was a growing brand then. Who wouldn't want Google brand to be associated with the publisher we didn't want to. So we built all this. In fact, a huge amount of work went into it because Ops was involved, a small Ops team. They were getting ready. They were hiring people for that. It was like two, three engineers.
Turner Novak:
So they were staffing up to approve all the things?
Gokul Rajaram:
They were staffing up, I mean classic fashion. They were staffing up. So one day Sergey, he would drop into the launch meetings and we were drum beat towards launch in I think, six or eight weeks. And he said, "What is this approval queue?" Because we had already check marked that and moved on. He looked at the screen, I was like, "Well, the approval queue is something publishers have to be put in a queue. We approve them." "Why?" "Well, what do you mean why?" We didn't even know how to answer this. "Why? Because publishers could force a reputational risk to Google." He was like, "Why?" We were like, "Well, if a publisher with bad content runs." He's like, "I don't really care. I think they should be allowed run." We were like, "Hang on, do you think they should be allowed to run any publisher?"
He was like, "Yeah." He anyway, pointed out this interesting thing that we had overlooked. He's like, "You're not understanding that publishers can just change. They can apply with a different website." One of the biggest challenges we faced was validating that someone owned the website because you or I could apply with nytimes.com and there's no way to know that you owned it. Even today I don't know how to actually prove, there's some complex ways that you can prove it, but back then it was super hard to prove that you owned a certain domain or something like that. So you could apply with that. And then people also had subdomains because people would host it on some subdomain of Blogger.com, so that was super hard to prove. So that's what he said. The reality is whatever you're doing has lots of holes in it anyway.
He also pointed out that Google and AdWords, you could actually run an ad against any keyword you wanted with any website you wanted for a hundred or a thousand impressions before they started charging you, before they started doing approval. So he suggested the same thing that, "Hey, why don't you just let the ad run for a thousand impressions, and at that point, trigger an approval or some kind of check. It can be automated in many cases because you can see what pages they run on instead of basically focusing on the site they apply with. So you know really that they own the site because they got the JavaScript onto the page, and second, you know the exact pages they run on." That's the other thing, there were many sites that were mostly good, but ad pages that were bad because maybe it was a news article.
There was a crazy one around the news article about a serial killer. I think it was a top news publication, but one page had that content. We didn't want to show ads against it and we were able to turn it off because of that. So we were too coarse-grained and we were not thoughtful enough. So I think overall, we took it off. I think the key lesson though for me was the magic moment lesson, where most products have a magic moment and the magic moment for AdSense was publishers seeing that they made money. That was his primary objection. It was not the holes. It was the fact that you are impeding publishers from signing up for AdSense because they want to make money, and you're saying now wait for a week before you make money. You cannot have hurdles or barriers in making money.
Turner Novak:
Yeah, because they might churn and they go try different product.
Gokul Rajaram:
Exactly. There's very different product. They won't ever respond. You lose them. You have them then. So I think that removing friction from the magic moment is a lesson that has stuck with me, so I love products that you can go end to end and sign up and instantly get to the magic moment of the product, whatever it is. I think sometimes you can have the wrong sense of magic.
Caviar, which is a product I ran, a premium food delivery service. We thought a magic moment was for people seeing beautiful photos of food, which is what we had. And so we would basically not gate people from seeing the photos. In other words, you could just download the Caviar app and start browsing the photos and you don't have to have an account at the end. But some enterprising engineer saw this, or some marketing person, I forgot who it was, they put a login block or sign up block before seeing the photos. So basically, I was super upset. I was like, "Hang on, why are you doing that?" But turns out they were right because our magic word is not seeing the photos, it was getting an error-free delivery, and that didn't depend on photos.
So that login before seeing the photos didn't do anything. It was really when someone didn't get a delivery, that was what really lowered attention. That was a magic moment. The first delivery error-free from a great restaurant.
Turner Novak:
Interesting, okay. I could see that. We have Gopuff in Ann Arbor. It's probably the original 15-minute delivery company, the dark store model, the one that got super popular throughout COVID. So we've had this in Ann Arbor for a while. The first time I used it, I don't remember what we got, we got a bottle of wine or some chocolate for date night, and it came in 12 minutes. I was like, "Oh my God, this is insane." Because I'd used DoorDash and Uber Eats, it takes maybe half an hour or the restaurant you want is an hour away or whatever, but when it shows up that quickly, I was like, "This is game changing." You can order anything you want and get it instantly. Right there I was sold. We order from Gopuff all the time.
It's just the local quick, fast... It is the fast version of Amazon. It's like Amazon, but you get it in 15 minutes. It's incredible. So one other thing, I know you were at Google, but then you worked at Meta, formerly Facebook, you talked to Zucker a lot, worked with Zucker a lot. There's the moment when Google Plus launched, I remember when I was, I think I was a freshman in college or it was between my senior of high school and freshman year of college, it launched. All my friends were like, "Google Plus, this is going to beat Facebook. This is so cool. We used it." I don't think it ended up having any killer features. That's, I think, why we stopped using it. It was just a clone of Facebook, basically. I think it had video calling. But I'm curious, what was it like inside of Facebook? I know that was like a pivotal, completely pivot the company, let's make sure we survive this. What happened inside Facebook when Google Plus launched?
Gokul Rajaram:
I think the great thing about Zuck is he knows how to switch between founder mode and manager mode very well. Even then you could see that. He would delegate things very well, but when a crisis hit, he would go into founder mode, which is he would be extremely hands-on until he felt the crisis was passed. So as soon as Google Plus hit, or maybe even before it hit, as the rumors started spreading, he called what is called lockdown. So that is his version of founder mode. He did that, I think, a few times in the company's first several years. Whenever there was a crisis facing the company, an existential crisis or what he perceived as an existential crisis, he would put the company into lockdown. Back then there was a sign outside his office that said lockdown, it would be off. But turned on-
Turner Novak:
That's awesome.
Gokul Rajaram:
Lockdown was on. So you knew that lockdown was on, and that means that he called in all hands of every single engineer at the company, and he said that you've got to drop what you're doing. He would basically, personally run the roadmap for the company at that time, which was maybe a thousand-plus engineers for sure. So he would basically say, "I want teams built and I want these eight or nine things [inaudible 00:17:27]." I forget what they were, but they were very specific to counter the perception that Google Plus had. And so he said, and then of course the impact, he was very smart because it was lots of impact, but it has to be used by X number of people. So he said once we get to those things and ship them and get certain amount of usage, that's when you go to lockdown.
He also established exit criteria for the lockdown. It was not just a two-week thing, it was not a four-week, it was once all these things are done and there's usage, and basically, we've staved off. That was a forcing function and it was really energizing. I think we were on the ads team, so our job was not as.. Because ads still needed to keep running for the company to make money. So there was some part, but many of the ads front-end engineers, we didn't build any new features. That was more the ads infrastructure that has kept running. Many of the ads engineers also were pulled into the consumer side to build basically consumer features.
I ended up PMing one of the consumer features, I forgot which one. So even the PMs were pulled into other stuff. But I think when the company is in an existential crisis, I think that's for sure when the founder needs to just step on, you can't delegate at the point. You've got to be very, very close to the work and almost dictate, micromanage what happens.
Turner Novak:
It sounds like you don't remember any of the features. Do you remember? Was there anything that countered Google Plus? I'm trying to remember.
Gokul Rajaram:
I think Google Plus ultimately fell due to its own weight versus anything we did, to be honest. I think it was because they just tried to copy Facebook within the context of Google.
I think it was the pay limitation. I think when you try to copy something versus build something that goes to your strengths, I think you end up not doing well. I think Google was not something people associate with social first, of all. Even the brand name was probably just like Facebook and Instagram, right? The smartest thing Facebook did was buy Instagram. If I were Google, I would've just bought Instagram or something like that instead of trying to build Google Plus
Turner Novak:
Was Google Plus before Instagram?
Gokul Rajaram:
Yeah.
Turner Novak:
Really? I didn't know that.
Gokul Rajaram:
Of course.
Turner Novak:
Wow, man. At least they got YouTube though. That was a good acquisition.
Gokul Rajaram:
Yep, exactly. I think that is a great example, right? Google tried to do video, didn't work, and there was this fast-growing private company called YouTube at that point. Kudos to the leadership for shutting down video, even though there's a lot of people on it, it sunk cost didn't matter quite YouTube did all the warts. If you remember back in the day, YouTube had lawsuits just burning a lot of money and still we had to pay 1.69 billion, which is a big part of Google's market cap back then.
Did doing all that and we were all thinking, what is Larry thinking? I hope he knows what he's doing. Obviously, he had made a lot of great bets, all successful, and this was again. That's founder mode.
I think buying a company like that, going in directly, it was a version of what Zuck did a few years later, going and buying Instagram and then going into WhatsApp. I think in many cases an acquisition, when it's a culture that goes to actually what I was saying. Google is a technologically focused company. Google does really well when technology matter was YouTube was not a technology feedback. Yes there is technology, but ultimately it's an environment, it's a culture, it's a vibe. It's a different kind of company. I don't think you can build that within a company. In fact, acquisition in many cases, is the best way to do it when the fundamental what it takes to win is different than what your core DNA is as a company.
Turner Novak:
Yeah, I think you can maybe in a recent parallel with that was the reason that TikTok was successful and got as big as this because it was its own unique kind of culture of the product. It was different than YouTube, it was different from Instagram, obviously copied the features. It probably kneecapped the growth a little bit, but it didn't die because it was kind of its own product, its own DNA.
Gokul Rajaram:
TikTok had nothing to do with identity as much. I mean there was no identity. You can post any video, sometimes with yourself, sometimes with other stuff. It's just pure video. You're right, there was some copying and some kind of... But yeah, it's still growing and doing well.
Turner Novak:
I have a listener question from Hari Raghavan. He said you are on the board of a couple of public companies, I think Coinbase and Pinterest, maybe it's changed, maybe those are the two, maybe there's more.
Gokul Rajaram:
Trade Desk is the third one.
Turner Novak:
And then you're also helping founders starting their companies. What is the difference? Maybe it's founder mode versus manager mode. It's like public board member mode versus startup helper mode. How do you switch between those two and why do you think you can do it successfully?
Gokul Rajaram:
To be honest, I feel both of those modes compliment each other for me because the things you learn from a skilled company, some of those things can be brought to the startup and then some of the things that you learn from the startup around nimbleness, agility, new ways of doing things, can be brought to a skilled company. They're quite similar. The key difference I think, is that at the earliest stage of a startup, the founder has nobody. They don't have an executive team. It's much closer to founder mode when you're helping because you're trying to be more prescriptive in those cases. Why? Because a founder doesn't have a team. You can't say, "Okay, here's six things, go and discuss with the team and come back." There is no team in many cases, the founder, a few engineers, literally you are brainstorming.
By the end of the meeting you have a course, a plan of action that you agree on, and the founder goes and execute. I'm sure you see this yourself. At a public company as a board member it's not your job to prescribe, far from it. Because this is a leadership team, they're pretty senior executives, they've skilled this company. Your job is to be almost Socratic.
Turner Novak:
The philosopher.
Gokul Rajaram:
Exactly, and ask questions. Ask the right questions and make sure they're thinking about those questions. Have we considered this earlier? I joined my first board in 2008, 16 years ago. Back then I would conflate the two modes when I would very much try to prescribe things. I realized that over time that it was not my job as a board member to prescribe, but my job was to make sure the company was looking around the corners and asking the right questions to make sure you're looking around the corners.
Now of course, if you ask the same question three times and they don't have good answers, then you start becoming more prescriptive. "Hey, can we do this?" Again, it's not like I want you to do this. Can we please at the next board meeting discuss pricing or discuss competition, whatever the case may be. But first you start with, "Hey, have we looked at this? Have we considered this?" And you see what they have to say. That's how you do it versus, "Hey, let's discuss pricing. In the next half an hour, let's discuss pricing. Let's discuss to hire this person." You've got to just execute. I think there's no time to discuss.
Also, the timeframes around the startup, the iteration speed is just so much faster that you don't have time to wait till the next board meeting, for example, to think about that. But at public company scale, obviously, you've got to operate fast, but if you operate too fast, you have a multi-thousand person company, you can't yo-yo them up and down. So you've got to be very thoughtful and intentional about what you do.
Turner Novak:
Interesting. I definitely had it before with founders where we're talking about something, I'll be hanging around their office and then two days later they send the investor update and they're like, "Launched the thing we talked about." I was like, "Whoa, nice." That's like a key KPI for a startup is just like how fast you move, how fast you get things out there. So you not only worked at a bunch of companies, you're a board member, you also invest in companies sometimes. How did you first get started with Angel Investing and making the jump to making that transition?
Gokul Rajaram:
It's a great question. I think even today it's the same thing. It's supporting people. In this case actually, it started by investing in Google colleagues back then in 2007 who were leaving Google. I think 2006, 2007 was when the first wave of Google people started leaving to start companies. People like David Friedberg who started All-In. He left to start the Climate company, which became actually the Climate Corp. It was called WeatherBill back then. So a bunch of people left to start companies and was lucky enough to invest in a few of them, but it was just a support. I realized that that's the best way. You invest to learn and you invest to support as an angel. As a professional VC, obviously now you have obligations. Turns out if you do it well, you invest in good people to support them and you invest in spaces that you're not too much of an expert in, but you want to know about, you think it's interesting, you do well financially, but it's not the primary goal. Obviously, as a venture capitalist-
Turner Novak:
You kind of have to make money or you don't last very long. Actually, you could maybe last kind of long if you don't make money, but there's a point eventually where-
Gokul Rajaram:
The fund three is what [inaudible 00:26:09] fund three or fund four is when the LPs truly want to start seeing.
Turner Novak:
Yeah, they want to see some of those good markups. I don't think you need any big exits yet by fund three. It depends how long it's been, but you're probably four to six years in at fund three maybe if you're lucky. But on average I don't think you get those big exits quite yet.
Gokul Rajaram:
For seed stage, yeah, for sure. For the growth stage fund, maybe they'll see some exits.
Turner Novak:
Talking about seed stage, we were talking a little bit before we started recording how much seed has changed from when you first started to today. They're almost completely different. What was it like back when you did your first few angel investments and then over time how has it changed?
Gokul Rajaram:
2007 there were just a small number of firms. Basically, there were angels, so you always went through an angel round and then the firms would come in. So you had a time to do a friends and family round. Then there were a bunch of angels that became super angel. Jeff Clavier was an early super angel, a few others, and then the firms would come in. So everyone had their clear lanes. No one impeded in their lanes. It was all somewhat collegial. It was a collegial industry. 2014 I think, the first time, I think where you started seeing seed funds come up. Homebrew was one of the early ones. First Round was probably the earliest alongside Jeff Clavier. I think there were two or three, but there were not that many.
Again, they wrote early checks, but then seed explosion happened in 2013, 2014 I think. [Inaudible 00:27:34] was created around that time, your Former Shop, Homebrew, etc. Again, still the large firms and the seed firms generally, there was a clear demarcation where the seed firms would do the first check and they would then graduate from there and the large firms would do the next check. I think in 2018 to 2020, a seed change happened when the large firms started going down market. I think at this point, I don't think any large firm where they sold, there's a single one that has not done inception or seed deals. Many of them have their own seed programs. They call them different names, but they are the seed programs and they offer very, very good terms, sometimes better than what the pure seed firm can offer because these are parts of multi-stage firms. So it's not the only thing. This is almost in many cases a way to-
Turner Novak:
It's a deal flow to get the series A. Where they're going to put... They want to put 50 times more capital into the company.
Gokul Rajaram:
Exactly. Even to series B sometimes, exactly. So it's a way to dip your toe. So I think that makes it extremely challenging for both the entrepreneur to choose. If you get a five million on 50, or 10 on 50 as this first check from a large firm, or you're getting, which might come with its own pros and cons, you're getting a two million on a 20 from a seed firm, what are the pros and cons of that? Obviously, the check size is five X more, but it comes with a different set of things. So I think that it makes it hard for the seed firms, and it makes it hard for the large firms because the reality is large firms now have hundreds of companies, some of them. And some of them have retreated, but they're still doing it because they want to know about these great companies earlier and earlier, and the best way to know it is to be on the CAP table, even in a small way.
So I think that it's definitely a Cambrian explosion in some ways of choice for the entrepreneur, but it makes it confusing and hard for LPs on the one side and for the investors themselves. Everyone is a little bit confused, I think. Everyone's trying to figure out, especially with AI companies. I'm sure you've seen it, AI sales and marketing space in particular, and AI legal, there's a market. I can't list the number of AI sales and marketing companies that are out there. There's probably more than a hundred. A hundred good companies, good companies, good teams and doing well. These are all companies that are doing a few hundred K to a few billion in revenue.
Turner Novak:
And then it's like what is the form factor or the team that specifically breaks out or is there going to be a hundred of them that all break out and do hundreds of millions in revenue or billions in revenue? I certainly get a lot of emails from those AI email companies that are helping people. Not as excited about that part of it being on the receiving end.
Speaking of that, about busy inboxes, a lot going on, everyone I talked to when just asking what we should talk about in this episode, I think Kurt Lin at Pinwheel specifically, he was like, you are extremely responsive. You're on the boards of public companies. You're invested in and advising a bunch of founders all the time. You're super busy. How are you so responsive? I think even with me when we're DMing or emailing, you respond faster than I do, and I'm pretty sure you have more going on than I do. So what is your system for just being so responsive? And actually, maybe a way to lead into it is, why is responsiveness so important?
Gokul Rajaram:
I think it's because things change so fast. I've seen that if you don't act on something, A, I think by the time you act on it's too late and then the other side basically is like, "Okay, it doesn't matter anymore." So I've seen that being responsive, being fast at helping or even just replying to an email can make a difference as just things move too fast.
The second thing in responsiveness is to get it out of your own brain. I think if you don't respond, if something is kind of... It helps the recipient hopefully, but also helps you. For me, it's both because for me, I have almost like an OCD about cluttering my inbox or things that are in my mind. If I respond to something it's almost like I get the monkey off my back.
There's a famous article called Get the Monkey Off Your Back, HBR article. It's a 1960s or 1970s article. I recently posted about this a couple of months ago. Beautiful article. It uses a bunch of '60s languages, subordinates and so on, which I don't like, but I'm sure that it still holds true. It's really about within a company, but it's about how when someone gives you something and someone says, "Hey, can you help me with something?" You need to just help them with something and put it back on them to then move forward with it, versus saying, "Okay, I'm chewing on it." And then you ask them the next time, "How are you working on it?" They said, "I thought I had told you that I need help with it." So they put the monkey on your back. So the goal is I want to have as few monkeys as I need to carry.
I feel by responding, and always the response is, "What do you think?" Even the response can be, "Here's what I think. What do you think?" You're basically putting the action on them to think more. So I think it gets the monkey off my back. Anyone can only carry a small number of monkeys, and I feel every email in my inbox, if it don't get a response from me, is a monkey that I have to carry, and I need to lighten the load of the monkeys. The easiest thing I do, it's pretty simple. It's a principle called Getting Things Done. I did a seminar on it at Google 22 years ago, 21 years ago. It's always stuck with me. I think there's a lot of systems, but the thing there is about email. When you go to an email, it's a one-touch principle. What it means is when you open the email, you do one of three things.
If it takes you less than five minutes to process and respond to it, do it right then. Second, if it's not yours, if it's a response to delegate, you just forward it to someone and ask them to take care of it. Or third, if it takes more than 30 minutes, most things are bimodal, it takes less than five minutes or more than 30 minutes, you have a separate task list where you put this in the task list and then you basically come back on a regular. The task list is what you need to really work on an ongoing basis, create a presentation, do a memo, think about pricing, strategy, whatever the case is.
But then your email is clear and it's either responded to delegated or it's on your task list that you have to do. And then you use calendaring with yourself. Schedule time to do each of those tasks. So you estimate how long it takes, and you basically break it up into pieces and you use calendar with yourself.
Turner Novak:
Oh, so you literally make calendar events and you say, I'm working on this thing for 45 minutes. This is how long I think it's going to take and I'm going to get it done.
Gokul Rajaram:
Yeah, instantly, it frees up my mind because I no longer am thinking, "Oh my gosh, when do I find time?" Guess what? I found time with myself to do it. It again takes a monkey off a back to find the time to do it, and you find the time with calendar.
Turner Novak:
So I found the HBR article, Management Time, who's Got The Monkey, and then Getting Things Done by David Allen.
Gokul Rajaram:
Yes, GTD, exactly.
Turner Novak:
Okay, so I'll throw links to both of those in the show notes if people want to dig into it a little bit more. You gave a pretty good pitch for them. I'm going to read them after this. You also have another topic, I know you like talking about founders should always be selling. What does that mean?
Gokul Rajaram:
I think it means that almost always the first few customers have been acquired by the founder through hand-to-hand sales. Jack Dorsey still can recite to you the first few customers that he got for Square. Tony, Stanley and Andy can still tell you the first two customers they got for DoorDash, etc, especially in B2B business. Even for B2C, I bet Zuck can tell you who the first 10 users on Facebook were, and Larry and Sergey can tell you who the first 20 users on Google were.
But I think as the company grows, sometimes founders lose touch with that specific thing, which is focus on the customer. I think it's important to go into the field on a regular basis to engage with customers and convince customers to use the product. That's why I saw that whenever Jack was in the cafe, he would always... In fact, he would use excuses. When I was doing interviewing with Square, we would always meet in a Square cafe and he would always go up to the merchant and ask them, or any kind of cafe, "Why are you using Square? Why you're not using Square? Can we convince you?" He always would carry around some squares. He would try to get them to use it.
Turner Novak:
Okay.
Gokul Rajaram:
And it was amazing to me that at that scale, the founder was still doing that. Also, I think selling is important because you're always selling to recruits. So I think founders, many founders, still hundreds of people will try to close recruits as well as selling to investors, I think. So I think selling is a skill that across all your constituents, whether it's investors, employees or recruits or customers you need to keep doing, and you can't lose touch with it. You can't delegate it because you're going to [inaudible 00:36:07].
Turner Novak:
So you're either selling your product, you're selling a paycheck or equity to an employee, or you're selling equity to an investor. So you're always trying to sell.
Gokul Rajaram:
You're ultimately selling a vision to all three of them, a vision for how your business is going to be better using them, how your life is going to be better or how your returns are going to be better for investors. So you're selling, and selling is as you know, all about storytelling as much as anything else. Ultimately, storytelling is probably the single most important skill that you need. How to weave the story that compels the person on the other side to take an action based on the story or narrative you're telling.
Turner Novak:
Interesting. So when you talk about Jack Dorsey, he's in the weeds. He's always talking to customers. If I'm a founder, how should I think about drilling down on my ICP? I guess it can be different stages of the company.
Gokul Rajaram:
The most important thing about ICP is ideal customer profile. I think the biggest mistake that founders make is they make it too general, too broad. The ICP needs to be as narrow as possible.
Turner Novak:
But don't you want a bigger TAM?
Gokul Rajaram:
You want a bigger TAM, but the reality is it will get bigger. I believe if you look at Facebook, the ICP was purely Harvard students, like whatever, 5,000 of them. That's it. If you can't succeed in that, then there's no reason for you to go outside of Harvard. So I think narrow ICP focuses you to just win that fully because if you think about ICP simply, a business, assume you're selling to business, it could be a business that is less than a $100,000 in revenue that has its physical storefront that is in the mission district of San Francisco, that's it.
Those are six, seven qualifiers right there. Then you win that segment, then you remove one of the qualifiers that are in the mission district and basically that are in San Francisco. You win that, then you remove that, then you go to the Bay Area, you remove that or whatever it is. So I think you slowly remove qualifiers and you slowly get bigger and bigger in concentric circles.
But I've seen time and again, the biggest mistake founders make is to have too broad an ICP thinking exactly the TAM, TAM, TAM, but this TAM thing is all useless. It's all theoretical. The actual thing is the ideal narrow, narrow, narrow customer, and I think the right framing is your ICP should be a segment where if you sell to a hundred customers of the ICP, your value proper offer it, 90% of them should say yes. If it's only 10%, or 90% of them say that's a problem they face, and then you obviously are selling problem not solution, and then you're like, "Okay, that's my solution." But [inaudible 00:38:35] should raise 90%. If it's only 10%, that's not an ICP. That's too broad. So I think you need a narrow enough thing that almost everyone you speak to in an ICP should resonate with that. So I keep pushing founders to narrow their ICP. Narrow it, it's okay, you can go broad later. Narrow, narrow, narrow.
Turner Novak:
I would not have expected 90%. I mean that's just super, super narrow.
Gokul Rajaram:
It forces you to then really narrow, and then you can go to 60 if you take one, but then let's go and check with this 90%, if it's true or not. As I say, it's very important to have, then you can dominate. I believe also the Peter Thiel axiom, I don't know if it's true. I think he said it in his book, it's much better to win a blue ocean niche market and nail it first and then you go from there where you're not competitive in some ways. The fact that 90% is facing the problem is no one is solving the problem for this 90%. Almost the whole segment is untapped.
Turner Novak:
So then if I'm thinking about how do I size my market, obviously I probably am not doing a top-down analysis, how would you advise?
Gokul Rajaram:
It has to be bottoms up. It has to be segmented out by these attributes. And then you try to understand how many customers in each of these segments, and then what's the willingness to pay in this segment, and then you multiply. It has to be bottoms up. In fact, one of the best TAM sections is Bling Capital, has a TAM, how to size a market because it turns out if you have any population more than 10,000, there's always going to be segments. You can't tell me they're uniform market or maybe a few thousand. So I want to see segments, I want to see how you're segmenting them, and I want to see how you characterize it. It could be by large, medium, small, like SMB, mid-market, could be something else.
But then you look at each segment, size it bottoms up, what's the willingness to pay, how intense is the problem based on your pricing model? And then you try to understand what it is. But it has to be bottoms up. I don't like top-down segment at all. I think they don't serve anyone because it's not actionable. Bottoms up tells you what your assumptions are. Okay, there's this many mid-market customers. Each of them has this many employees. Each of them is doing this, so therefore this is the size, etc.
Turner Novak:
Yeah, I think I found it. It's called How to Size a Market in 30 minutes.
Gokul Rajaram:
Correct, that's it. Yeah, it doesn't take too long.
Turner Novak:
Yeah, I guess you have 30 minutes. That's pretty quick. We'll throw that link in the show notes as well. You have a couple more. How to hire a leader for a function, Titles of a startup, which I think is a fun topic, the contrarian opinions.
Gokul Rajaram:
People have strong opinions. Ben Horowitz, I think even Mark and [Inaudible 00:41:05] they both believe that titles, and I had a part the other day with Sriram, they believe very strongly that you should actually give titles because they're an easy way to compensate people. But I think especially director and VP titles early on, I think can set the wrong tone for the rest of the company and can really make it painful for you to run the company. I've had enough people now reach out that they basically retreated back from titles after seeing how painful it was once they gave one of those title to one person because they felt they had give it to them to get them to join the company. Everyone at the company, they're like, "Okay, now directors are fair game. So what does it take for me to become a director?" And that's what they focus on, becoming a director, becoming a VP.
Turner Novak:
If I'm a director, you're quite older than me, and then you come in and you're a manager, you're going to be like, "Why is Turner a director? I should probably be higher than him, let alone at the same level."
Gokul Rajaram:
Exactly.
Turner Novak:
Well, this has been a lot of fun. Thanks for coming on. This is a fun conversation. We had to be really tight with time, but hopefully people enjoyed it.
Gokul Rajaram:
Turner, thank you, this was amazing. Your style is just so awesome, and thanks to the listeners. I'm excited too. Hope that it's helpful to folks.
Turner Novak:
Yeah, I think so. I actually took a lot away from this in such a short time, so this is fun.
Gokul Rajaram:
Thanks Turner.
Stream the full episode on Apple, Spotify, or YouTube.
Find transcripts of all other episodes here.