The first tournament on FairwayNation was mine.
The Sully Memorial is a real event with real golfers, real sponsors, and a real date that doesn't move because the software isn't ready. It's a memorial golf tournament for my brother, Sean P. Sullivan, who died in a helicopter crash on July 19th, 2012. This year was the 14th annual event, with a full field of players, sponsors, and volunteers. Before I sold the platform to a single organizer, it had to get me through that day without embarrassment. It did, and everything worth knowing about the product came from that.
What I thought the product was
A registration page. Golfers sign up, pay, get a confirmation. Maybe a leaderboard. I'd sketched it as a weekend project.
What the tournament needed
Sponsors. The revenue in a charity tournament isn't the golfers; it's the hole sponsors, the cart sponsors, the dinner sponsor. They need their own signup, their own tiers, their logo on the site, and an invoice their accountant can process. That's a sponsorship system, and it's half the product. I wouldn't have built it for a customer. I built it because a sponsor emailed me at 9pm asking where to send the logo.
Real payment plumbing. Money from registrations goes to the tournament, not to me. Stripe Connect does that, and it's a meaningfully harder integration than a checkout button. The alternative is me holding other people's charity money in my account, which is a problem in every sense.
Bot protection. Within weeks of the registration page going live, the form was getting hit by bots. Not attacks, just noise: fake signups, card testing, garbage in the database. A real tournament with a real organizer would have found this on the worst possible day. I found it on a Tuesday, because I was the one watching.
A website, not a form. Organizers don't want a registration link to paste into a Facebook post. They want a site: the course, the schedule, the sponsors, the story of the event. So the product grew a website builder, and the builder is what organizers see first.
Each of these was a week or more of work I would have argued a customer out of. Each turned out to be the product.
Why dogfooding isn't optional
You can interview organizers. I did. They'll tell you what they think they need, which is a registration page, because that's the part they're picturing. They won't tell you about the sponsor logo email at 9pm, because they don't remember it as a software problem. It's just tournament stuff.
Running the event turns tournament stuff into requirements. The friction you feel at 9pm is the friction every organizer feels, and none of them will put it in a feature request.
The other thing your own event gives you is a deadline that can't slip. The tournament was on a date. The software had to work by that date. That's a clarity most side projects never get.
If you're hiring me
When a client brings me a product idea, my first question is whether they've done the job the product is for. If they have, the spec writes itself from their frustrations. If they haven't, the first milestone is usually finding someone who has and watching them work, before anything gets built on a guess.
Before you sell it, run it. If you can't run it, find someone who does and stand next to them.