I finally made the decision to call it quits. Now I think it would be a good time to reflect on my experiences and see if I could learn something from them.
As I talked about in the previous post, initially, I thought that the market size was hundreds of thousands of users, and that I could make something like $100/user. After talking to people in the industry, I now believe that the market is more like 5-10k users, and not all of them are willing to pay $100.
This is incredibly important! Going after a $20M market is very different from going after a $200k one. If I knew it was the latter in the beginning, I wouldn't have pursued this as a business. Spending 2+ years for the chance at making maybe $200k just isn't worth it, given the inherent uncertainty with startups, and the alternatives of pursuing a startup with a higher upside, or getting a job that pays approximately that much but does so with 100% certainty.
So what happened? Where did I go wrong? Let's see. This was (roughly) my initial logic:
100k subscribers to the poker subreddit. Educational YouTube videos gets 100k+ views. Popular posts on TwoPlusTwo (big poker forum) have 1M+ views over the years.
The kind of person to subscribe to the poker subreddit, watch an educational YouTube video, or spend their time on TwoPlusTwo is probably somewhat serious about poker. They're trying to get better at the game. All of this is indicative of a market size of hundreds of thousands of users. Possibly more. And poker is expanding. So there seems to be a big market here.
And poker players are probably pretty willing to spend money. Investing in software is +ROI. Poker players love ROI! And they tend to be on the wealthy side.
Maybe I really underestimated the divide between passive + free and active + paid. Watching a YouTube video is something that is passive. You sit there and consume information. Using poker software is something that is active. You have to sit and think and mess around with numbers. Watching a YouTube video is free. Poker software costs money.
But then what about the existence of poker books? There around 500 poker books on the market. The top ones get up to 100k sales, and many others get in the tens of thousands, I think. Maybe it's that with books, someone is telling you what the right answers are, but with software, you have to figure out the right answers yourself.
Anyway, I think the bigger point is that I should have found people in the industry and asked them about the market size. I started doing that towards the end of my journey, but I should have done so from the beginning. People in the poker world all seem to have a pretty solid idea of what the market is really like. Why screw around trying to figure it out myself with these questionable proxies when I could just ask the people who actually know?
I really can't emphasize enough how huge this is. It would have only taken a few hours, and it would have saved me so much time.
So why didn't I go out and talk to people in the industry? I'm not quite sure. I think a part of it was that I didn't actually feel like I had to. It seemed pretty clear to me that the market was big, so I was more concerned with making the product awesome.
Another part of it is that I didn't see it as an option to talk to people in the industry. Because why would they want to talk to little ol' me? They're basically B-list celebrities, in some sense. They've written books. Tons and tons of people know them. Don't these people have hundreds of fan emails every day that they never respond to? That was my thinking in the beginning. Now, I've come to realize that they're just people, and that they're often happy to chat and provide advice.
It also would have been good if I had access to the actual financial data of my competitors. But none of them are public companies. Does that mean this isn't an option, or are there still ways?
I came across one cool approach last night. If you don't have access to their financial data, you could look at how many employees they have, and multiply by something like $125k or $200k. Something in that ballpark should give you an idea of their revenue. My competitors, from what I can tell, are all working solo. So that is a sign that they aren't making millions and millions of dollars. Not definitive, but definitely points in that direction.
Another interesting option is to actually call your competitors and talk to them! Eg. you could pretend to be a prospective employee, and in that conversation start asking about revenue and stuff. I'm not sure how I feel about that sort of deception though.
Here's a closing thought for this section: the world isn't that big. For the majority of my journey, my thoughts on the market size for poker has been, "I'm not sure exactly how big it is, but it's big enough." The world is just such a big place, and so many people play poker. The market just has to be huge. Now I realize that this isn't true.
"Just another month or two"
As I explain in the first post, for a very long time, I kept thinking to myself:
I'm not sure if I should really pursue this as a business or a long-term thing, but I do know that I want to finish up X, Y and Z. It'll only take a month or two, and I think there's a good chance that it finally gets me over the hump.
This just kept happening over, and over, and over again. It's crazy.
And each time it happened, it felt like this was the time it really would only be another month or two. I was wrong last time, and the time before, and the time before, and the time before... but this time... this time I'll be right.
Wow. Articulating it like that really helps put things in perspective. I need to diagnose myself with a chronic case of the planning fallacy. I need to do a better job of adjusting in the opposite direction. I have a tendency to be overconfident, so I need to adjust in the other direction, and be less confident. That's what you have to do with known biases: try to adjust in the other direction.
And with the planning fallacy in particular, there's a known cure: the outside view. Don't try to reason from the ground up. Look at how long similar things have taken in the past. Maybe use the reference class of "times I've thought it'd only be another month or two".
Man, that makes me laugh. "Times I've only thought it'd be another month or two." Ha! That reference class is full of miscalculations, so it's pretty clear that I need to adjust in the other direction pretty hard.
I say all of this stuff, but I still worry that I'm going to make the same mistakes again.
It always takes longer than you expect, even when you take into account Hofstadter's Law.
Check out this excerpt from the first post in this series:
And I have this little voice in my head saying:
Hey Adam... it's been over a year... you don't have any users. This like totally goes against the whole lean startup thing, y'know.
And then I say in response:
I know, I know! But I really question whether the lean startup applies to this particular scenario. I know you want to avoid wasting time building things that people don't actually want. That's like the biggest startup sin there is. I know. But like, what am I supposed to do?
My hypothesis is that once my app gets to the point where it's in the ballpark of Flopzilla or better, that people will want to use it. It takes time to get to that point. There isn't really a quick two week MVP I could build to test that hypothesis. I'm already trying to avoid building non-essential features, and focus on getting to that point as quickly as possible. So what am I supposed to do?
If I released this and found that I had no users and no one wants this, what would that tell me? Just that people don't like this version of the app enough. Sure, ok. But the real question is whether they'll like the Ballpark Flopzilla version enough, or the Better Than Flopzillla version enough. My hypothesis is that they will, and releasing it now wouldn't invalidate that hypothesis. And I can't think of a way to test those hypotheses quickly. I think I just need to build it and see what happens.
I'm going to resist the temptation to respond to this right now. Right now I just want to tell the story. The story as it actually happened. But I do want to say that there were a lot of voices swimming around in my head questioning what I was doing.
I said in that post that I'm going to resist the temptation to respond to it. Here's where I do get to respond to it.
Here are my general thoughts about the whole lean startup thing:
1. I think the essence of it is to ruthlessly avoid spending time on things unnecessarily. Eg. you don't need to spend a week building a fancy navigation menu before you even know whether people want your product. I heartily, heartily support that message.
2. It can be tempting to think that there are no quick experiments you can run. This is usually wrong. Think more carefully, try to isolate the component assumptions, and get creative.
I definitely could have done a better job with (2).
3. But sometimes there truly aren't any quick experiments you can run. Eg. SpaceX. How is SpaceX supposed to build a quick MVP? (Well, there are some things they could do, but I don't want to get distracted from the central point that some hypotheses inherently just take a long time to validate.)
Now that I'm finished with Premium Poker Tools and am reflecting on what I've learned, it's time to add a fourth general thought.
4. If you are in a situation where (3) applies and are going to spend... I don't know... two years and three months testing a hypothesis... then the upside damn well better be worth it!
Pretty obvious. If the risk is big, the reward needs to also be big.
With Premium Poker Tools, the reward never big enough. It just wasn't. This is no SpaceX.
Even if my initial ideas about the market size were correct, and there was the potential to make $10-20M, that still isn't enough to justify spending so long testing a hypothesis.
However, I don't think it's that simple. This issue is very tangled up with the planning fallacy stuff I talked about in the "Just another month or two" section. I never actually decided to spend 2+ years on this.
Show me the money
https://www.youtube.com/watch?v=FFrag8ll85w (I had to)
I have another objection to the above section. I wasn't testing a hypothesis the whole time. No. About a year into it, I already had market validation. That's right. I already had validation.
I'm not sure when exactly I would say it occurred, but a big thing is when I posted to Reddit and someone asked where they can donate. Another big thing is seeing one of the most popular poker players in the world using my app. Another was having a bunch of people tell me that they would pay for it. Another is having random people email me thanking me for building the app and telling me how great it is.
Now, I know they say talk is cheap. I know how startup people always want to see traction. Users. Money. Growth. I didn't have any of that. But I did have all of the other stuff in the above paragraph! People said so many nice things to me. Sure, conventional wisdom might say that the above paragraph isn't enough, and that you need real traction...
But I'm a Bayesian! I'm better than that! Saying that you need actual traction is like how the scientific community waits too long before accepting something as true. Being a Bayesian, I can update in inches. I can be faster than science. I can be faster than conventional startup wisdom.
Well, that's how I used to think anyways. Now I run around my apartment with my fingers in my ears yelling SHOW ME THE MONEY!!!!!!!!!
I'm exaggerating of course. I don't actually do that. And I don't actually think you should ignore everything except "actual results". No, I'm still a Bayesian, and Bayesians don't throw away evidence. But given my experiences, I've come to believe that such evidence isn't nearly as strong as I had previously thought.
Deals fall through
This is similar to the above section. In the above section, I'm basically talking about how when people say "This app is awesome! I would pay for it!" it doesn't actually mean that much. This section is going to talk about how when potential business partners say "This is really interesting! Let's talk more!" or even "I'm in!", it doesn't actually mean that much.
You can read more about it in part one, but I've just had so many deals and stuff fall through. A lot of people were telling me that they were interested in working with me. Some even said that they would work with me. A verbal "yes". I thought to myself:
Ok, great! This is pretty strong bayesian evidence right here. I wouldn't have so many people saying this stuff if they didn't mean it. Sure, maybe a few fall through, but not everyone.
And also, I haven't even put myself out there too much. Just to throw out a number, maybe I've talked to 10% of the potential people who I could partner with. If I've gotten this much interest from the 10%, once I go after the 90%, I should end up with a good amount of partnerships.
Given my experiences, I've come to believe that verbal interest just isn't that telling. And this seems to mirror the experiences of others as well.
I don't want to say that verbal interest means nothing though. Honestly, I probably went too far in the above paragraph saying that it "just isn't that telling". I still have limited experience, and I'm not sure how strongly to weigh it as evidence.
But one thing does seem pretty clear: the lack of actual traction is stronger evidence than verbal interest. Eg. for me, I had a lot of people saying they're interested in partnering with me, but no one actually following through. I think the lack of follow through is stronger evidence than the verbal interest. Similarly, I had a lot of people saying they really like the app and stuff, but even when it was free I was only getting maybe 100 users/month, and they weren't spending that much time on the app. I should have paid more attention to that.
Build it and they'll come
I've always had a perspective that goes something like this:
My app is at least in the same ballpark as Flopzilla. I feel pretty confident that it's a little bit better, actually. So then, I would think I should at least get 10-20% of the market, if not more.
Yes, I know they were the first mover and have the brand recognition, but if my product is in the ballpark, I should still make a dent. I release it, people hear about it, some people like it and start using it, they tell their friends, people link to it in forums, put screenshots in blog posts, etc. I would think that if my product is as good as the market leader, through a process something like what I just described, I would get my slice of the pie.
Furthermore, I would expect my slice to be proportionate to the quality of my product. If the product is a little worse than the competitors, maybe my slice is 5%. If it's a little better, maybe I get 30-40%. If it's way better, maybe it's 75%.
And maybe that perspective is too optimistic. That's certainly possible. But it can't be too off the mark, right? Maybe if the product is a little worse I end up with 1% instead of 5%. Maybe if it's only a little better I get 5-10% instead of 30-40%. Maybe if it's way better I get 50% instead of 75%.
Although it's still a little counterintuitive to me, I have to say that my perspective is different now. The perspective I described above is some version of "build it and they'll come" (BIATC). I now thing that BIATC is pretty wrong.
I'm still not quite sure what the mechanism is, though. I suppose that it takes a lot to actually get word of mouth to happen. I suppose people don't actually do that much comparison shopping, and instead lean heavily towards things that are popular, have social proof, and that they stumble across organically, eg. in blogs.
I feel like I may have overestimated BIATC due the stuff I hear from YC folks. They talk a lot about focusing heavily on the product, as opposed to marketing it. I at least get the impression from them that making something people want is what it's all about, and that if you manage to do so, you'll have success. Here's Paul Graham in How to Start a Startup:
It's not just startups that have to worry about this. I think most businesses that fail do it because they don't give customers what they want. Look at restaurants. A large percentage fail, about a quarter in the first year. But can you think of one restaurant that had really good food and went out of business?
Restaurants with great food seem to prosper no matter what. A restaurant with great food can be expensive, crowded, noisy, dingy, out of the way, and even have bad service, and people will keep coming. It's true that a restaurant with mediocre food can sometimes attract customers through gimmicks. But that approach is very risky. It's more straightforward just to make the food good.
Using the example of restaurants, I actually can think of a lot of restaurants with great food that don't do so well. Hole-in-the-wall-type places with awesome food, but that are never too busy. And on the other hand, I can think of a lot of trendier restaurants that have terrible food but a lot of customers.
Still, I don't want to completely discount BIATC. I think it can be true in some situations. I just think those situations have to be pretty extreme. You need to 10x your competition. You need to be solving a hair on fire problem. You need to be building a painkiller, not a vitamin. You need your users to really, really love you. When you totally blow the competition away, or when you truly solve an important problem that hasn't been solved yet, yeah, when you build it, they'll come. Maybe that's what YC is trying to convey.
Customer acquisition is hard
Of course, I didn't actually just build my app and sit there expecting people to come to me. I did try to acquire customers.
It just didn't work. Despite the fact that I have a product that people say all of these nice things about, I only managed to get, let's say three paid users and 100 free trial sign ups. That amount of traction seems very much not in line with the quality of product I have, which makes me think that customer acquisition is very hard. Or maybe just that I'm still bad at it.
Here's what I tried, how it went, and what I learned:
This has always been Plan A. Find some people who already have huge followings eg. on YouTube, and piggyback off of that by offering them revenue share. Should be pretty simple. It's free money for them, and the product is good, so why wouldn't they want to do it? Especially when I ended up offering them 50%.
Well, I'm still not quite sure what the answer to that is.
Maybe they know how small the market is and how little money they'd make? Perhaps. But still, it's so low effort for them to throw a link in the description. And they're often already using poker software in their posts/books/videos, so it isn't any extra effort for them to use mine. In fact, mine is a better fit because mine is the only web app, and they could link to simulation results.
In talking to them, the response I usually get is that they've been meaning to check the app out but just haven't had time. This makes no sense to me because as poker professionals, I'd think that they are already spending time studying with software, so why not substitute mine in? Maybe they don't actually study? Maybe they don't want to spend the time messing around with a product they're unfamiliar with? Who knows.
My takeaway here is that if the deal you're offering people is only marginally beneficial, you might just end up with no partners.
YouTube and blogging
I think I have produced solid content with my YouTube channel and blog. People have told me that. But I've gotten only a minuscule amount of hits.
Again, my initial thinking was what I described in BIATC. That the amount of hits I get should be at least roughly proportional to the quality of my content.
Nope! That didn't happen! In retrospect, this makes sense. How are people supposed to discover your YouTube videos? YouTube recommends videos that are already popular, and chooses the videos that are already popular to put high in their search results. Similar with blogging. It's a chicken-egg problem that I have yet to figure out.
There are a ton of huge companies built off of ad revenue: Google, Facebook, Instagram, Twitter, Reddit, etc. So then, that gives me the impression that paid ads are a huge thing that everyone is doing. And if they're doing it, they're doing it because it works for them, presumably. So paid ads have always been something that I assumed to work well.
Nope! The first place I learned this was in Julian Shapiro's guide. He said something that made it actually seem pretty clear in retrospect. If you were able to get an ad channel working profitably for you, you'd just be able to scale up with that channel, acquiring customers profitably, and making a ton of money. If every company were able to do that, entrepreneurship would be pretty easy.
Since reading his stuff, I started to hear other people say the same things, that it's really hard to do profitably and most companies never manage to do so. I still decided to give paid ads a shot, because it's a low risk high reward thing, but it turned out not to work for me.
I spent a pretty good amount of time doing direct sales. Some in person networking at a poker meetup, playing poker at the casinos, meeting people in coffee shops, Ubers, etc. Then I also spent some time DMing people on poker forums, and emailing poker players and coaches. I didn't go too crazy because I don't want to be spammy, but I definitely did some.
None of it really worked though. I think my takeaway mirrors the BIATC stuff, where you need to have something really valuable to actually get peoples attention.
I again found this surprising. I figured that when you actually are in a conversation with someone, they'd sort of give you the benefit of the doubt. But no, I didn't find that to be true.
Refer a friend
There was a point where I was offering $20 for every friend you refer. Pretty good, right?! Seemed pretty generous to me, and like something people would want to take advantage of. But no, that didn't happen. Still confusing to me, but I guess the lesson again is that it really takes a lot to get peoples attention.
Sharable simulation links
(I talk more about this in this post.)
Poker players spend a lot of time in forums discussing hands. They'll say things like, "you only have 33% equity, so you should fold". My app lets you link to a simulation that shows that you only have 33% equity. For non-users, it's in readonly mode. For users, when they could click the link and play around with the assumptions.
I thought this would be huge. Making it easier for people to discuss hands in the forums. But no, it did nothing. I think a big part of it is that most people in the forums approach it very casually and don't want to spend a lot of time on their comments. They just want to add their two cents, and then leave.
The app was even free for a long period of time. I would have thought that this would attract a ton of users, but no, that didn't happen. More evidence that it takes a lot to actually get people to look in your direction, and that customer acquisition is hard.
People are lazy and irrational
Here's what I mean. It's really true that if you're a poker player and are actually trying to get better, that you should have some sort of poker software to study with. Coaches and professionals say this all the time. But people are still too lazy to do so.
Hell, they are just too lazy to study in general. I've said this before, but they prefer passive things. Watching a video and having someone "tell you the answers", even though that sort of passive study never works, regardless of the field. Eg. with math, you have to actually do a lot of practice problems yourself to get it to stick.
I've also heard a lot of coaches complain about students who pay them $100+/hr not do the homework that they are assigned. The coaches beg and plead, but the students just don't want to do the work.
That's the lazy part. I guess the irrational part plays into that, but what I really had in mind is that poker software is pretty easily a +ROI investment, but people still don't care. And if poker players aren't persuaded by +ROI investments, then I'm not sure what other demographic would be persuaded.
There are definitely things that I'm forgetting. Hopefully I'll add to this post as I remember them.
If any readers out there have any advice, I'm all ears! I want to try to learn as much as I can from this.