Dharmesh Shah, co-founder of HubSpot, joins Sam and Shaan to tell the story of building Wordplay — a Wordle-inspired word game he launched in 48 hours with his 11-year-old son. The project grew to 9.5 million users and was generating $90,000/month via Google AdSense before he turned the ads off. They discuss viral growth mechanics, what the project could sell for, and the lessons Dharmesh learned about cold-start problems and community building.
Speakers: Sam Parr (host), Shaan Puri (host), Dharmesh Shah (guest, HubSpot co-founder)
Building Wordplay With His Son [00:00:00]
Sam: You built a thing called Wordplay — is it wordplay.com?
Dharmesh: It’s wordplay.com. Right. By the way, you buy domain names — you bought the hustle.com for us, you bought me copythat.com. Thank you for that. I forgot to say thank you in person.
Sam: So you have wordplay.com. Did you tweet that it was getting seven million users a week, or three — what was the number?
Dharmesh: So I just went and checked the numbers. Just to kind of place the timeline in frame: November and December of 2021 is when Wordle took off. January is when it really hit — I think close to its peak. That’s roughly when the New York Times acquired Wordle. A couple things were kind of bothering me — I think it was a great outcome for Josh Wardle, but there were things that were limitations in the game, and I was an avid player.
Now it’s like, okay, the New York Times has bought it — there’s no way anything’s going to change for years, if ever. That was thing number one.
Thing number two: it had been a while since I had launched a simple, free thing publicly. I hadn’t gone through that process, and there were things I wanted to learn — Vercel, Next.js, TypeScript, things like that.
And then that intersected with thing number three, which was the dominant variable in the equation: my son was taking a Python programming class. He’s 11. He’s going through this process, and for him it’s still very abstract. He’s learning programs, he enjoys everything involving technology and screens, but it was abstract.
So on a Saturday night I said, okay, let me build something. Python happens to be my language. Build something so he can see — because he plays Wordle, he knows the game. I started on Saturday night with a deadline for Sunday: I’m going to launch something with him tomorrow. We’re going to have Google Analytics on it, I’m going to launch it by tweeting it. He knows what social media is, so I’m like, we’ve got this thing that didn’t exist yesterday — I wrote the code over the course of last night in Python, the language that you’re learning — and now we’re going to launch it, and you can see users coming in.
What I want him to have in his head is that when he encounters problems that are solvable with software — which is a lot of problems — it’s now tactile for him. It’s like, I don’t have the skills right now to do what dad just did, but I know it’s possible, because he did it in the span of 24 hours. So if I want to go build a video game, or build some tool for my school, or build some social network for whatever — it feels more approachable. He’s got years to go before he’ll get to that level, but anyway.
Those three things sort of came together, and we launched it on that Sunday — I think in February or so. Since then, 45 million games have been played, 9.5 million people have come through it. At any given moment right now, if you look at Google Analytics, there should be two or three thousand people playing.
How Wordplay Got 9.5 Million Users [00:05:30]
Sam: How did it get so much traffic? Where did the traffic come from?
Dharmesh: Well, it helps that I have a social media following — I’ve got a million followers on LinkedIn, 300,000 on Twitter. There’s that.
Sam: That’s a lot. But that’s not — I wouldn’t have thought that was enough for that. How many users do you think you got from that in the early days? Was it big right off the bat? The first week — was it big?
Dharmesh: The first week was probably maybe 50 to 100,000. So it wasn’t a lot, but it was something.
But then once you get into it — this is the power of iteration — one of the things I found missing in the original Wordle was that it was a single-player game. You played once a day and that was it.
Two big changes we made in Wordplay: one was unlimited play — you didn’t have to play just once a day. And two, it allowed you to challenge your friends. You could say, “I just played this word, I solved it in four turns, here’s my score” — and send a simple link to anyone or a group. This happens tens of thousands of times. Someone will take a link from a game they’re proud of and post it to their WhatsApp group, wherever, and say “I invite you to beat my score.”
That’s the viral element. As more people play, you can track it as a Google Analytics event — how many times are people clicking that share button and issuing challenges. Once you get into something with a compounding effect on the user base, the base grows.
The thing I’m fighting now is that interest in Wordle and related games is waning — everything has its peak and its ebb. But it’s been an interesting exercise.
The $90,000/Month Discovery [00:08:30]
Dharmesh: One of the things I hadn’t done in a while is ask: okay, what’s traffic actually worth? Let’s say I had to make a living on this thing or make the project profitable. So I put Google AdSense on it — easiest thing to do — and discovered that if I weren’t bothered by having two ads on the main game board page, if I were solving for monetization, it was $90,000 a month.
Sam: In Google AdSense? Wow. I thought you were going to say way less.
Shaan: I thought he was going to say like five grand a month.
Sam: Ninety grand a month — and your only costs are hosting, right?
Dharmesh: Yeah, and my time.
Sam: First of all, Shaan — go to wordplay.com and then go to the About page. His About page is incredibly well written. Wonderful copywriting, wonderful voice. It says — or actually it says, “Why would I do this? That’s a really good question. My wife asked me the same question.” And then he goes on to explain why. You’re quite a good writer for also being an amazing engineer. Typically those two don’t go together.
Shaan: Did you take my writing course? Because I’m not throwing around credit here with my alumni.
Sam: I like to say I taught Dharmesh everything he knows — even before he met me, I somehow taught him all that good stuff before he built HubSpot.
What It’s Worth and Whether He’d Sell [00:11:30]
Sam: What do you think this could sell for right now? And is this the biggest side project you’ve ever created?
Shaan: Our buddy tried to buy it. He tried to buy it off you. Did he really?
Dharmesh: He tweeted out, “I’d like to buy this off you, I’ve built mobile games before” — he was semi-serious, at least. I don’t know if he was or not, but I did not take him up on it.
Sam: What do you think you could sell it for? If you cared, what do you think?
Dharmesh: Two or three million dollars, maybe. Right now, let’s say ad revenue would go down to $50,000 or something like that — I haven’t turned Google AdSense back on in a while — that’s $600,000 a year. Apply a 5x multiple, and you’re somewhere between two and ten million dollars would be my guess.
Sam: Was posting on LinkedIn and Twitter the only way you got users, or did you do some other weird hacks?
Dharmesh: I had my blog. I’m a big believer in the flywheel — there are all these little things attached to the flywheel. LinkedIn helps a little, puts in a little energy. Twitter does a little bit. My blog, the email goes out, there are things I sort of do. Yeah. It helps.
But the hardest part is the cold-start problem — how do you get your first hundred to a thousand people? And then it’s a matter of tracking retention on those users: do they come back?
The other interesting thing about Wordplay as a lesson learned: when it started, the average time a user spent according to Google was roughly four minutes. In the early weeks. Now that number is at 14 minutes — the average time a user spends across however many thousand users come in on a given day.
Biggest Side Project Ever Built [00:15:00]
Sam: Is this the biggest side project you’ve ever created?
Dharmesh: Depends on how you measure it. Technically no, not in terms of raw traffic. I’ve built graders and things like that before that have done better.
Sam: What side project have you built outside of HubSpot that gets more traffic than this?
Dharmesh: A few people really did side projects. This was my first one that was clearly — because it was a game — something I couldn’t bring into the HubSpot umbrella. I like to have all my eggs in that basket, but given the circumstances here, it didn’t really make sense.
Even now I’m thinking about doing a marketing and sales related edition of Wordplay — like the entire list of words is MarTech companies: do you know the players in the space? And then you can get sponsorships or whatever.
Website Grader was the kind of project where it was a free, low-friction thing that people could come to. This is why my son and I were talking yesterday — I was looking at Flutter, the new development environment. Dart is the language but Flutter is the app framework for building mobile apps across iOS and Android. It’s what the cool new kids are learning. And he’s like, “No — why would I ever want to build a mobile app? That’s something people have to download. I have to get approval from Apple to post it. I want to just do it immediately.”
He’s right. It’s like, okay.
Lessons Learned and What Comes Next [00:18:00]
Dharmesh: Anyway, that’s the Wordplay story. It’s been a lot of fun. It’s hard to know with these things where the road leads. Wordplay itself may or may not connect back to other things. But the things I’ve learned through it — it’s like, okay, here’s what it takes to build this kind of thing.
And I’m using it to play around with community. Can I take the Wordplay user base — what are things they care about — and build a community around Wordplay using existing community platforms? Can I see if that’s a thing? If you could aggregate people with like interests, that in itself has value.
I wanted to send you something that kind of self-runs. We’ll see.