How to Set Goals Before Product-Market Fit

Do you need to grow 7% week over week? Dalton and Michael discuss how founders should set goals for pre-product market fit startups. The classic YC advice of growing 7% every week works but you can't cheat the data. Discussion includes: making products people actually love, why founders shouldn't worry about graphs in the very early days, why founders tend to cheat the data, how Paul Buchheit built Gmail very slowly over months to focus on the users loving it, why you should focus on love not like, not focusing on investors, how Stripe built slowly, why focusing on investors is habit forming, doing research on your favorite companies, building enduring companies, why customers loved Whatnot DoorDash & Twitch, and why you need to get to 100 customers that love you before you graduate to graphs. Dalton + Michael is brought to you by @Standard_Cap. Standard Capital is the AI-native Series A fund. Learn more at standardcap.com – About Dalton: Dalton Caldwell is Co-Founder and Partner of Standard Capital. He spent 12 years at Y Combinator, where he served as Managing Partner, worked across 25 YC batches, and advised more than 1,000 startups. His investments include Whatnot, Brex, GitLab, PostHog, Stock Space, Rappi, Razorpay, and Oklo. Before becoming an investor, Dalton founded imeem and App.net. About Michael: Michael Seibel is a Partner Emeritus at Y Combinator, where he served as Managing Partner of the early stage accelerator from 2014 - 2024. Michael also serves on the board of three companies: Reddit, Dropbox, and Kalshi. He moved to the bay area in 2006, and was a co-founder and CEO of two Y Combinator startups Justin.tv/Twitch (2007 - 2011) and Socialcam (2011 - 2012). In 2012 Socialcam sold to Autodesk Inc. for $60m and in 2014, under the leadership of Emmett Shear (CEO) and Kevin Lin (COO) Twitch sold to Amazon for $970m. – Are you an AI builder? Check out StandardDB. Discover offers, credits, tools, and partner programs from the StandardDB ecosystem.

Dalton CaldwellhostMichael Seibelhost

Sep 28, 202615mWatch on YouTube ↗

EVERY SPOKEN WORD

15 min read · 3,216 words

0:00 – 0:41
Why 7% weekly growth is the wrong starting point pre-PMF
1. DCDalton Caldwell
  So do you think that your product is helping them in any way in their business, in their, like in their goals? And that- they give, they give me that look. They're like-
2. MSMichael Seibel
  [laughs]
3. DCDalton Caldwell
  Is that what products are supposed to, like-
4. MSMichael Seibel
  Yeah
5. DCDalton Caldwell
  ... I thought we were supposed to be growing 7% week over-
6. MSMichael Seibel
  Yeah. [laughs]
7. DCDalton Caldwell
  Like... This is Dalton + Michael, and today we're gonna talk about setting goals as a pre-product market fit company. So Dalton, we just launched, and I know that, um, if we're not growing 7% week over week, we're failing, right?
8. MSMichael Seibel
  I mean, let me, let me-
9. DCDalton Caldwell
  I'll just start you with a banger. [laughs]
10. MSMichael Seibel
  Uh, let me attack, let me attack this from a different direction.
11. DCDalton Caldwell
  [laughs]
0:41 – 2:04
Big-company analytics instincts break when you only have a few users
1. MSMichael Seibel
  Let's say you used to work at Google or Meta-
2. DCDalton Caldwell
  Mm-hmm
3. MSMichael Seibel
  ... or whatever, and you know a lot about product and analytics.
4. DCDalton Caldwell
  Yes.
5. MSMichael Seibel
  And you're used to, like, working in those environments, and you get into YC.
6. DCDalton Caldwell
  Yeah.
7. MSMichael Seibel
  And you launch, and you come in, and you wanna show, show me some graphs, and you wanna tell me about your AB test and all this stuff. And I'm like, "How many users do you have?" And they're like, "Three."
8. DCDalton Caldwell
  Yeah.
9. MSMichael Seibel
  And I'm like, "Okay, never show me these graphs again." Like, graphs are a waste of time.
10. DCDalton Caldwell
  [laughs]
11. MSMichael Seibel
  Measuring weekly growth when you have zero users-
12. DCDalton Caldwell
  Like, they're like, "We have three users."
13. MSMichael Seibel
  ... are projecting.
14. DCDalton Caldwell
  "And they all churned."
15. MSMichael Seibel
  [laughs] Yeah.
16. DCDalton Caldwell
  So what you're saying is you have zero users.
17. MSMichael Seibel
  Yes.
18. DCDalton Caldwell
  Let's not talk about weekly growth.
19. MSMichael Seibel
  Yes.
20. DCDalton Caldwell
  And, and so where I'm going with this, and I know we're on the same page, is that the tactics and skills that you have working at a post-PMF company have no relation-
21. MSMichael Seibel
  No
22. DCDalton Caldwell
  ... to tracking progress and goals in an early-stage company. If anything, I noticed that college kids that knew nothing about analytics were sometimes ahead of the game 'cause I- they didn't have to unlearn all this crap-
23. MSMichael Seibel
  No
24. DCDalton Caldwell
  ... of, like-
25. MSMichael Seibel
  No
26. DCDalton Caldwell
  ... how they were trained at Meta to establish if a product is working or not, right?
27. MSMichael Seibel
  Yes. Well, you can run an AB test on 1% of Utah and, like-
28. DCDalton Caldwell
  Right. Like at Meta-
29. MSMichael Seibel
  [laughs]
30. DCDalton Caldwell
  ... they have all these great... You know, it's great, and if you work at Meta, you know, good for you that you have-
2:04 – 2:35
What 7% growth advice is actually trying to enforce (retention + value)
1. MSMichael Seibel
  No, no. [laughs] But I think you're, I think you, you, you dodged the, the gotcha question.
2. DCDalton Caldwell
  Go ahead.
3. MSMichael Seibel
  I think the gotcha question is, I think growing 7% week over week is great or horrible depending on how you do it. I think the genius of that advice is that when you're small, growing 7% week over week is not actually-
4. DCDalton Caldwell
  Hard
5. MSMichael Seibel
  ... adding that many users.
6. DCDalton Caldwell
  Yeah.
7. MSMichael Seibel
  Right? I think the other genius of that advice is that it's a lot harder to grow 7% week over week if you're churning all your users every week.
2:35 – 3:06
Metric cheating and "good-hearting": how founders fool themselves
1. DCDalton Caldwell
  Sure. You and I both know, and this is where we're going, what people do is they choose metrics that cheat.
2. MSMichael Seibel
  Yes.
3. DCDalton Caldwell
  And they're like, "Oh, we're not gonna do, uh, DAU. We're gonna use cumulative users that ever signed up, and we're not gonna worry about if they're using the pr-" Like, they're basically founders.
4. MSMichael Seibel
  Yes.
5. DCDalton Caldwell
  I think the term is good-hearting. What they'll do is just, like, find a way to doctor the metric so that they hit it.
6. MSMichael Seibel
  Because-
7. DCDalton Caldwell
  And all you're doing is screwing yourself.
8. MSMichael Seibel
  Well, because the, the i- because it's like, I think that growing 7% week over week is a result of doing a lot of things right.
9. DCDalton Caldwell
  Yeah.
10. MSMichael Seibel
  Whereas they think it is the cause of the company-
11. DCDalton Caldwell
  Amen
3:06 – 4:38
The "gotcha" user-value questions that reveal you have zero users
1. MSMichael Seibel
  ... working, and it's just like, th- ah. Like, when you mix those things up, you really screw up. The number of times I talk to a founder where I'll be like, "These users do you have, how many of them spent more than five minutes in your pro-" Actually, no. The first question, because it's a gotcha, it's like, "To get value out of your product in a week, let's say, how much time should your users be spending in the product?" And they'll always be like, "Oh, you know, like an hour a day. Like, this is a core-
2. DCDalton Caldwell
  Yeah
3. MSMichael Seibel
  ... workflow. Like, this is really important to them. Da, da." And then I'll be like, "Great, so how much time are they spending in the product?" And there's always, like, two answers I love. One, "We're not measuring."
4. DCDalton Caldwell
  And you're like, "Oh, we'll have to get back to you." [laughs]
5. MSMichael Seibel
  Yeah. It's like, you only have, like, six u- like, you don't know? Y- and then two, "Oh, it's like an hour a week." And you're like, "So you said that to get value out of the product, they have to be using it this much. Are they getting value?" "I don't know." Then, you know, I like to pile on. It's like, "So do you think that your product is helping them in any way in their business, in their, like in their goals?" And that- they give, they give me that look. They're like...
6. DCDalton Caldwell
  [laughs]
7. MSMichael Seibel
  Is that what products are supposed to, like-
8. DCDalton Caldwell
  Yeah
9. MSMichael Seibel
  ... I thought we were supposed to be growing 7% week over week. [laughs]
10. DCDalton Caldwell
  Yeah.
11. MSMichael Seibel
  Can we get back to the point?
12. DCDalton Caldwell
  And, and again, where we're going with this is you do this line of thinking, and you know where we end up, which is we have zero users.
13. MSMichael Seibel
  Yeah. [laughs]
14. DCDalton Caldwell
  Like, like basically you, you go through this line of questioning-
15. MSMichael Seibel
  Yeah
16. DCDalton Caldwell
  ... and, and-
17. MSMichael Seibel
  We have a graph like this, and we-
18. DCDalton Caldwell
  And it's like-
19. MSMichael Seibel
  We're like, "Uh." [laughs]
20. DCDalton Caldwell
  We actually, we have zero users.
21. MSMichael Seibel
  We got it. We got it. [laughs]
22. DCDalton Caldwell
  And you're like, "Okay, well, I'm glad we had that little chat." And, and so where this goes is-
23. MSMichael Seibel
  [laughs]
4:38 – 5:31
Pre-PMF goal setting: get a few people to genuinely love the product
1. DCDalton Caldwell
  It's t- yeah, it's so true. Um, where this goes is you don't worry about the graph when you have zero users.
2. MSMichael Seibel
  Yes.
3. DCDalton Caldwell
  And you're not helping anyone.
4. MSMichael Seibel
  No.
5. DCDalton Caldwell
  And you're just, you just wanna make something that a few people think is awesome.
6. MSMichael Seibel
  Yes.
7. DCDalton Caldwell
  And you don't need a graph to know that now. You're like-
8. MSMichael Seibel
  'Cause you're talking to them
9. DCDalton Caldwell
  ... do you have three people-
10. MSMichael Seibel
  Yeah
11. DCDalton Caldwell
  ... that love your thing?
12. MSMichael Seibel
  Yeah.
13. DCDalton Caldwell
  Do you have five people that love your thing? And that, and I always think you should start with that.
14. MSMichael Seibel
  Yes.
15. DCDalton Caldwell
  Don't even worry about graphs and charts-
16. MSMichael Seibel
  No
17. DCDalton Caldwell
  ... and weekly growth. Just get something that someone loves, right?
18. MSMichael Seibel
  It's almost impossible, in my experience, to have something that 100 people love that's, like, somehow not growing.
19. DCDalton Caldwell
  Yep.
20. MSMichael Seibel
  You're either lying about the love, or you're not looking at the growing. Like, it's really, really hard. And I would say that the sad truth of the 7% is that, like, sometimes it takes time to get someone to love it. Sometimes it takes time for you to figure out what the hell to build that someone would love.
5:31 – 6:40
Gmail’s early playbook: slow user adds, relentless iteration for love
1. DCDalton Caldwell
  You know, we, we speak about him a lot, but our colleague, uh, Paul Buchheit, created Gmail.
2. MSMichael Seibel
  Yeah.
3. DCDalton Caldwell
  And this is how he created Gmail, just in case you haven't heard this story, friends. Um, he built it for himself, and he was the first user.
4. MSMichael Seibel
  Yes.
5. DCDalton Caldwell
  And then he gave it to other people internally at Google. He would only add one or two people a week, and he would ask them what he needs to add to, to Gmail.
6. MSMichael Seibel
  Yeah.
7. DCDalton Caldwell
  And so it went through this state for months and months and months where it had less than 100 users, but he wanted to build something that people loved.
8. MSMichael Seibel
  What I love about that story is by the time Gmail was beating- Right? It wasn't even open distribution. It had hundreds of people who loved it.
9. DCDalton Caldwell
  They lo- And they, again, love is the word.
10. MSMichael Seibel
  Yes.
11. DCDalton Caldwell
  They weren't like, "Oh, I guess I could take it or leave it."
12. MSMichael Seibel
  It was their email client [laughs] .
13. DCDalton Caldwell
  They ran their life off of it.
14. MSMichael Seibel
  Yes.
15. DCDalton Caldwell
  And so when you think about that this is the most popular consumer product in the world, I use it every day, you use it every day, how did that product come to life?
16. MSMichael Seibel
  Yes.
17. DCDalton Caldwell
  It was someone that wasn't trying to get a hockey stick graph-
18. MSMichael Seibel
  Yes
19. DCDalton Caldwell
  ... within two seconds of building it. It was someone that took the time and the care to make something that a small number of people loved-
20. MSMichael Seibel
  Yes
21. DCDalton Caldwell
  ... and, and w- were willing to run their life on. And again, I can't emphasize this enough, this is how you build great products.
6:40 – 8:12
The fundraising trap: building for investors instead of users
1. MSMichael Seibel
  But Dalton, I have to fundraise in N months, and you know that investors love graphs. It's gonna take too much time. It's gonna take too much time to build something that people like. You're not solving my immediate problem, which is how do I build something that investors will give money to [laughs] ?
2. DCDalton Caldwell
  Yeah. I think that's just the classic self-defeating mentality, where you, what you're saying, founder, is that you're building something for investors and not for users.
3. MSMichael Seibel
  Build something investors love.
4. DCDalton Caldwell
  Yeah, like-
5. MSMichael Seibel
  That's our model [laughs]
6. DCDalton Caldwell
  ... basically you just ex- you just made the pitch-
7. MSMichael Seibel
  Yeah
8. DCDalton Caldwell
  ... that your actual customer is an investor.
9. MSMichael Seibel
  Yeah.
10. DCDalton Caldwell
  And nothing good will come of that. Again, like-
11. MSMichael Seibel
  Yeah
12. DCDalton Caldwell
  ... this is, these are not the droids you're looking for.
13. MSMichael Seibel
  No.
14. DCDalton Caldwell
  If you start thinking this way, that your entire startup is a product made to convince investors, you will not win.
15. MSMichael Seibel
  You will not win. Well, okay, no, no, no, no, no, no, no. You will win in the short term.
16. DCDalton Caldwell
  Yeah.
17. MSMichael Seibel
  I bet you, I, I, I... If you're smart and somewhat capable, you could probably win in your seed round. But what I love about that seed round, what's the winning percentage of ultimately winning when you raise your seed round?
18. DCDalton Caldwell
  It's not, it's, I mean, it's better than zero, but, you know, it's just-
19. MSMichael Seibel
  [laughs]
20. DCDalton Caldwell
  [laughs] Yeah.
21. MSMichael Seibel
  2% chance?
22. DCDalton Caldwell
  Yeah.
23. MSMichael Seibel
  Okay, so you just took your odds down to 98% chance of failure. What if it's, you spent that same amount of time actually interacting with your customers-
24. DCDalton Caldwell
  Right
25. MSMichael Seibel
  ... learning what they needed, learning whether any of your hypotheses are right, learning how to make their businesses better, their lives better.
8:12 – 8:52
Why user obsession beats pitch optimization (and YC’s counterexamples)
1. DCDalton Caldwell
  And how many stories do we have of founders that did a poor job of fundraising a demo day, but they had the half million dollars from YC, and they just powered through, and then a year later, they followed this advice we're giving-
2. MSMichael Seibel
  Yeah
3. DCDalton Caldwell
  ... and they're doing phenomenally well.
4. MSMichael Seibel
  Yeah.
5. DCDalton Caldwell
  Way better than their batch mates that were actually... They weren't making something that people want, they were making something for investors, making something investors want. And-
6. MSMichael Seibel
  Competing for the highest price.
7. DCDalton Caldwell
  Right?
8. MSMichael Seibel
  Yeah.
9. DCDalton Caldwell
  There's nothing good comes of that, friends.
10. MSMichael Seibel
  Well, and to your point about, like, that's habit-forming in a negative way. Like, you don't want your culture-
11. DCDalton Caldwell
  No
12. MSMichael Seibel
  ... "Oh, we're just gonna worry about investors now, but after we raise money, then we'll worry about the-"
13. DCDalton Caldwell
  You know, yada, yada, yada.
14. MSMichael Seibel
  [laughs]
15. DCDalton Caldwell
  We'll talk to users. Yeah, whatever, guys.
16. MSMichael Seibel
  Yeah [laughs] .
8:52 – 10:42
Stripe’s slow launch and the "Collison install" as pre-PMF superpower
1. DCDalton Caldwell
  Um, let me give you another example. Stripe is a company that we talk about a lot at YC. Phenomenal company. It's w- 100 billion dollar company, you know, just-
2. MSMichael Seibel
  Yeah.
3. DCDalton Caldwell
  What a lot of people don't understand is, like, they didn't launch for something like two years. They weren't in the batch growing 7% week over week. And like, no, I mean, that's not how they built Stripe, man. It was invite-only. It was beta-only. They built it for themselves for a side product, and they only added customers super slow. And I, again, it was two years plus before they publicly launched, right?
4. MSMichael Seibel
  Yep. And I remember back then, um, 'cause they pitched us at, at Justin TV to use it, and, and it, it wasn't good enough.
5. DCDalton Caldwell
  Yeah.
6. MSMichael Seibel
  Like, we didn't use it. And it turns out to make a payments product that, like, we had a global audience. Like, to make a payment product that could accept payments around the world, that you don't just MVP the-
7. DCDalton Caldwell
  You can't do that in two weeks.
8. MSMichael Seibel
  Yeah, no [laughs] . Doesn't, no. Doesn't.
9. DCDalton Caldwell
  And, and they did all this amazing stuff, where we talk about the Collison install-
10. MSMichael Seibel
  Yeah
11. DCDalton Caldwell
  ... where they would come and sit down with you at your computer and help you implement Stripe into the payments flow.
12. MSMichael Seibel
  Yeah.
13. DCDalton Caldwell
  They were that customer-obsessed. I was actually an early Stripe customer at my startup.
14. MSMichael Seibel
  Okay.
15. DCDalton Caldwell
  We were probably the first couple hundred customers.
16. MSMichael Seibel
  Yeah.
17. DCDalton Caldwell
  And I remember Patrick would send me messages in Google Talk all the time.
18. MSMichael Seibel
  Mm.
19. DCDalton Caldwell
  And so I had my new friend, Patrick Collison-
20. MSMichael Seibel
  Yes
21. DCDalton Caldwell
  ... um, because I was an early customer, I'd probably talk to him, like, every day.
22. MSMichael Seibel
  Yeah.
23. DCDalton Caldwell
  And he'd be like, "How's it going?" Like-
24. MSMichael Seibel
  Yeah
25. DCDalton Caldwell
  ... like we were an important customer to them.
26. MSMichael Seibel
  Yeah.
27. DCDalton Caldwell
  And how cool was that for me, as a consequence of being an early customer, to feel like my payments provider cared about me that much.
28. MSMichael Seibel
  Yes.
29. DCDalton Caldwell
  That I had a relationship with the CEO.
30. MSMichael Seibel
  Yes.
10:42 – 13:07
Use AI to learn the real timelines of great companies—and reset expectations
1. MSMichael Seibel
  Yeah. No, I think that that is a superpower that founders never wanna tell. And here's what I'll say. Th- this will be my closing point. Pre-AI, it was hard to do research on how companies won, right? Like, you have to go through articles, and you have to listen to a bunch of podcasts. Like, if you wanna hear the early, early stories of Stripe, it might take a couple hours, and maybe that's too hard. And certainly, if you wanna hear the early stories for 50 companies, that, that's, that's a lot of research. It's a lot easier to accept, like, the pop knowledge. And I'll tell you, like, with, with AI, research is easy now. You can hear about the companies that you like. You can learn about what they were like when they were small. And I just, like, I challenge you to take whatever assumptions you have about early-stage startups and test them against your top-
2. DCDalton Caldwell
  Yeah
3. MSMichael Seibel
  ... 10 favorite companies.
4. DCDalton Caldwell
  How long did it take for them to launch from when they created the company? How long did it take them to get 100 users? How long did it take to get them to a million in revenue?
5. MSMichael Seibel
  Yeah.
6. DCDalton Caldwell
  And you're gonna be shocked-
7. MSMichael Seibel
  Shocked
8. DCDalton Caldwell
  ... that the real stories are not what you think they are.
9. MSMichael Seibel
  And, like, that doesn't mean you have to do it the same way, but it's just like, hey, just get that stuff in your head. 'Cause I, I... You see that graph where it's, like, the companies that are fastest to 100 million in revenue. It's like, is that the race that we're... That's not the race, guys.
10. DCDalton Caldwell
  It's long-term value, right? Like, it's-
11. MSMichael Seibel
  Yeah
12. DCDalton Caldwell
  ... that's the ultimate thing we're trying to do here, is to create enduring value to help customers and to build enduring businesses and not- ... flash in the pan type stuff, right? Or at least that's why I'm here. Um-
13. MSMichael Seibel
  Well, and not only that, even if you're o- like, let's just say you're agnostic.
14. DCDalton Caldwell
  Yeah, you're like, "Either way." [laughs]
15. MSMichael Seibel
  Either way, right? Like, this is the way that is a higher likelihood of success. Like, so like, yeah, you can like it because, like, yeah, you like helping people, or you can like helping people 'cause, like, that's gonna get you richer. Like, either way- [laughs]
16. DCDalton Caldwell
  Yeah
17. MSMichael Seibel
  ... like, you get to the same answer.
18. DCDalton Caldwell
  And the inventory you should do, person watching this video, is to be really honest with yourself. Do I have any users that love my product? And I would use that word. How many love my product? If they don't, what can I do to cause them to love my product? And just break it down and think about it that way, instead of just staring at graphs all day, 'cause that's, that's dehumanizing to your users. Again, if you're, if you're Meta, I get it, they stare at graphs all day.
19. MSMichael Seibel
  Yeah.
20. DCDalton Caldwell
  But if you're dehumanizing your user base and staring at graphs, and you're not thinking about the people that you're serving and whether they love you or not, y- you're hurting yourself.
13:07 – 15:02
Actionable pre-PMF goal framework: 1 → 10 → 50 → 100 users who love you
1. MSMichael Seibel
  So I have a controversial question to ask you at the end here. We started this with how we should set goals as a pre-product market fit company. Have we answered that question?
2. DCDalton Caldwell
  I think it's do you have more than zero users? So again, this is what I would do at YC, is do you have one real user with no asterisks. You know, like-
3. MSMichael Seibel
  Yes.
4. DCDalton Caldwell
  [laughs] Okay, then I would say, "Okay, do you have 10?" And then I would say, "Do you have 50?" And then I would say, "Do you have 100?"
5. MSMichael Seibel
  Yes.
6. DCDalton Caldwell
  And if you could get 100 users or customers that love you, that is an awesome goal, and that's when I would start to graduate to graphs. Like, that's-
7. MSMichael Seibel
  Yes
8. DCDalton Caldwell
  ... that's kind of the point where you're like, "Okay, now let's start talking about graphs."
9. MSMichael Seibel
  Yes. And I think the only thing I would add is, like, I love talking to startups about how do we define love, especially early.
10. DCDalton Caldwell
  Yeah.
11. MSMichael Seibel
  Because, like, so we don't cheat it. And the second thing that I like to think about is that is there any way you can measure the benefit that you're providing the customer, right? Like, you worked with Whatnot. In a marketplace, it's really easy, right? The GMV is the benefit. [laughs]
12. DCDalton Caldwell
  Yeah. Well, their sellers were making so much money.
13. MSMichael Seibel
  Yes.
14. DCDalton Caldwell
  It was life-changing.
15. MSMichael Seibel
  Yes.
16. DCDalton Caldwell
  So yeah, with Whatnot, someone would start selling on Whatnot-
17. MSMichael Seibel
  Yes
18. DCDalton Caldwell
  ... and you could see how much money they were making, and it was like, "Wow!"
19. MSMichael Seibel
  Yeah.
20. DCDalton Caldwell
  These people are, like, earning a crazy living on this thing.
21. MSMichael Seibel
  Yes.
22. DCDalton Caldwell
  It was obvious they were providing value.
23. MSMichael Seibel
  I mean, DoorDash, similar example. Like, you wake up and DoorDash is a third of your restaurant's revenue. Okay. Like, you might have opinions, but for us at Twitch, like streamers, they could quit their jobs. That's the thing that I wish I said more in years past, was that, like, can some part of love, can you quantify the actual benefit of the customer, as opposed to just time with you or that kind of stuff? Like, if we could put a dollar amount on the benefit of the customer, now we're cooking with gas. All right. Good chat. Good luck with your goals.
24. DCDalton Caldwell
  Sounds good. Thanks, Michael. [upbeat music]