OP #317: When Custom Becomes the Default

I’m alive. I promise. Beehiiv tells me it’s been 26 days since the last OP reached your inbox. Time moved faster than expected. Amazingly, I’ve received notes from many of you checking in on me - I appreciate that.

I hope you’ve been well. And if things have been a bit sideways lately, I hope there’s a path back to center not too far ahead.

Silicon Alley Sports has its first event of the year coming up: the inaugural Padel Invitational. We’re down to just a couple of tickets. If you try to grab one and find it sold out, join the waitlist. Spots do open up.

It’s shaping up to be a great first event. Plenty of people will be playing padel for the first time. For the record, I’ve played exactly once. That’s kind of the point.

If this sounds interesting, you can learn more here.

A quick note before getting into this week’s piece: we are missing the OP links section as I just have not been consuming a ton of content while I’ve been under my rock.

Instead, all of my attention has been elsewhere. And the essay below on software, vibe coding, and AI reflects that focus. It matters. Put your attention there.

But before you do, here are the few links that I am sharing.

How OpenClaw's Creator Uses AI to Run His Life in 40 Minutes (YouTube) - this has been a hot topic over the past week in the AI ecosystem. If you want to see a bit of the future, give this a view. I wont lie, it’s a bit nerdy, but a great glimpse into where we could be headed.

How Watch YouTuber’s Make Their Money (YouTube) - Let me start by saying - even if you are NOT interested in watches, this is a great video that walks through the variety of ways that YouTuber’s in general monetize their channels. If you are in marketing, content generation, media, advertising, etc - a highly relevant video. FWIW, I like Michael’s channel very much… he’s a management consultant by day who has a very practical disposition.

• High Speed Internet Boom Hits Low-Tech Snag: a Labor Shortage (WSJ)

• Manhattan’s Congestion Pricing Also Slashed Drive Times in the Suburbs, New Study Finds (Jalopnik)

• I’m Coding Again (AVC)

• Stand Together Acquires Proto to Help Bring Social Impact to Scale (StandTogether)

• Software Is Becoming Something You Speak Into Existence (Scientific American)

Until next time,
Be good. Do well.
Darren

When Custom Becomes the Default

What gave me a real leg up when I entered the working world was not pedigree or credentials. It was time spent learning how things actually worked.

In middle school, high school, and college, I taught myself HTML and tried to understand how the Internet functioned under the hood. I worked at a local computer store building PCs. I landed an internship at a leading digital agency. I was not optimizing for resume bullets. I was optimizing for curiosity.

At the time, I did not have language for it, but I was energized.

While many of my close friends were out chasing girls and beers, I was at Borders Books and Music, flipping through a new book on HTML 1.0, trying to make sense of tags and tables. I was flipping through Computer Shopper, planning my next PC build. That curiosity compounded.

That same feeling is back.

I have been relatively quiet with The Operating Partner lately, not for lack of thinking, but because life has been loud. Silicon Alley Sports launching. A full plate at work. And, quietly in the background, me formally learning how to vibe code.

Borders has been replaced by YouTube.
Books by repos.
Trial and error by shipping.

I have also been fortunate to have a close friend who has been experimenting alongside me, someone already far ahead, who has pushed, mentored, and inspired me as I have gone deeper.

Here is the stack I have been working with:

• Visual Studio Code

• Claude Code

• Neon DB

• Vercel

• Resend

• Stripe

• Claude, Gemini, and ChatGPT

So far, I have built two real projects.

The first is the Silicon Alley Sports Ticketing Platform.

Ticketing has always been a necessary evil for us. Over the years, we have used Eventbrite, Humanitix, and others, paying meaningful dollars while using maybe 5 percent of the functionality. When every dollar goes back into the community and impact, wasted software spend matters.

I wanted to see if we could do better.

From idea to launch, I built a fully custom ticketing platform in under a week.

Not just cheaper. Better.

We added gating questions at checkout to ensure buyers are actually part of the Silicon Alley community. We built trivia directly into the platform. And we are about to release “The Lineup,” player cards, and profiles published ahead of each event so attendees can learn about one another before they arrive. A lightweight social layer, purpose-built for the room.

Now, the economics.

Eventbrite typically charges 3.7 percent of the ticket price, plus a $1.79 service fee per ticket, plus credit card processing.

Ignoring the card fee, that is roughly 3.7 percent plus $1.79 per ticket.

We process about $50,000 in tickets annually.
3.7 percent is roughly $1,800.
Service fees add another $358.

That puts us around $2,200 per year just for the platform.

I would be shocked if our new platform costs more than $100 annually, all in.

That is a 20x swing.

Financially, it makes obvious sense. From a user experience perspective, it is not even close.

We went from a rack suit to a bespoke one.

Writing that sentence gave me goosebumps, because that is the real insight.

We are entering an era where bespoke software is attainable. One size fits all is no longer the default. There will always be categories where generic tools are good enough, but the center of gravity is shifting.

The cost of custom has collapsed.

A few things I have learned along the way:

1. A background in coding helps, but it is no longer a prerequisite.

2. The stack you choose today will change. That is fine. Build anyway.

3. Tools like Claude, especially with MCPs or Superpowers-style workflows, are genuinely mind-bending.

4. Prompting is a skill. Thinking clearly before you type matters.

5. Copying and pasting is a real competency. One missed character in an API key breaks everything.

6. Details matter, especially environment variables and configuration files.

7. There is no substitute for diving in. Just start.

It is hard to imagine the future when you are standing outside of it.

Once you step in, it starts to reveal itself. Clearer. Not necessarily less scary. But clearer.

And that clarity is a competitive advantage - at least right now.

Before closing, a few questions I keep coming back to:

Q1 Where are we still running “rack suit” software, like a default Salesforce implementation or a generic email marketing tool, when a purpose-built workflow could materially improve conversion, productivity, or decision-making?

Q2 If not everyone should be building enterprise software, why is learning to build on nights and weekends becoming table stakes for operators who want to see things differently, ask better questions, and call BS when a slick Figma mock gets mistaken for a real product?

Q3 In a world where Google followed Yahoo, OpenAI followed Google, and Anthropic followed the popularity of ChatGPT, does being first actually matter anymore, or is the real advantage clarity of thinking, speed of learning, and knowing when to move?