Building in the open in 2024

I ran the last one for a decade, that’s a whole story. A hell of a story. But a different story. I’ll tell it to you when I can, but not right now.

The one before that was an e-commerce site that did pretty well but I didn’t love it. Before that were two service businesses. The first one I did for the love of the game, the second one was an attempt to make people stop asking me to fix their computer by charging them outrageous prices, which backfired horribly when they were eager to pay. None are relevant except to say I’ve been around the block and have the scars to prove it.

When it was time to get back out there, I wanted to use all I’ve learned to do better. Before I talk about what those lessons produced, I’m going to talk about what those lessons were. Cause before effect, after all.

One thing I wanted to do better this time was pattern matching - making the startup look the way that the industry and investors “expect” a startup to look. My last startup was an awesome idea with awesome tech (still is, but like I said, another story), but that one didn’t match patterns. It didn’t match investor patterns, industry buying patterns, patterns of existing, immediate, recognized and admitted needs. Because it didn’t “look” right to anyone, everything about it was way harder than necessary.

The “make it look right” approach runs the risk of building a cargo cult, imitating the trappings of something but without understanding the essence of that something, but then again, a thing that looks like a knife is going to make a better knife that a thing that looks like a bowling ball, so sometimes just sharing apparent similarities can get you pretty far, even if it doesn’t get you all the way there. Like how mimicking someone’s accent makes it easier for them to understand you.

For this one, I wanted to adopt every tool, method, and pattern that I knew “the industry” wanted to see to minimize the friction from development, go-to-market, scaling, adoption, and that would make investment optional (and, therefore, available if desired) instead of necessary (and, therefore, largely unavailable).

That required establishing some expectations for successful patterns I could match against.

What patterns am I matching to? Here’s a general sketch of my pattern matching thought process:

Software first and software only. It’s the easiest industry to start a business in, lowest startup costs, and easiest customer acquisition.

I wanted to build software for an element of the industry that’s actively emerging (and therefore has room to grow) and part of an optimistic investor thesis (and therefore has a cohort of people who are intent on injecting capital into the market to help it grow).

It needs to fills a niche that is underexplored (low competition) and highly potent (lots of opportunity), while being aligned to recognized and emerging needs within the industry (readily adopted).

I wanted it to have evidence supporting the business thesis that proves the demand exists, but demonstrates that the demand is unanswered (as of yet) by sufficient or adequate supply.*

I wanted the lowest number of dominoes to line up and tip for everything to work correctly - the more dominoes in the line, the less likely the last one will fall.

I wanted to implement modern toolsets for everything, wherever possible.

I wanted to obey the maxim, “When there’s a gold rush, don’t mine the gold, sell the picks and shovels.”

Whatever I chose would need to produce cash flow almost immediately with minimal development time or go-to-market delays, because the end of ZIRP killed the “trust me bro” investment thesis predominant over the last 15 years.

I wanted to match to YC best practices, not because YC can predict what will definitely work, but because they’ve churned through so many startups in the last 15 years that they have a good sense of what will definitely not work.

And I wanted to build client-centric, because if my intent is to to produce cash flow immediately, we need to get clients immediately, and if we need to get clients immediately, we need to focus on what clients need right now.

Extra credit: What’s the difference between a customer and a client?

*Note: Competition is awesome! Competition is validating and not scary, because competition proves a market exists. But competition, especially mature competition against an immature startup, makes it harder to break into a space. A first mover advantage isn’t everything, but seeing demand before it’s sufficiently supplied is a great advantage if you’re capital constrained or otherwise unproven. Think about how much money the first guy to sell fidget spinners or Silly Bandz made versus how much money the last guy to order a pallet of each made. Finding demand that exists already but is as of yet insufficiently satisfied is a great place to start.

What opportunity spaces are most relevant? The industries and markets I chose to observe were:

AI, because if I’m following a theme & pattern for today, it’s AI.

Fintech, because cash is king, and fintech puts your hands on cash flow.

Crypto/blockchain, because that’s the “new” fintech (or maybe the “old-new” fintech?), and crypto creates powerful incentives and capital formation strategies, along with a lot of flexibility for transaction systems.

Tools, particularly unmet demand in tools, that enable these industries.

If you wanted to do some brief and simple homework, you could map each of those bullets to several of the numbered list items preceding them.

The reasoning was pretty simplistic - AI is what people want to build and invest in now, while fintech and crypto/blockchain are what people were building and investing in for the last major investment thesis. That means that there’s demand in the market for AI and AI-adjacent startups, while there’s a glut of underutilized and highly developed tools within fintech and crypto/blockchain, with a lot of motivated capital behind the adoption. When someone is thinking “I built this thing and not enough people are using it”, and you then build something that uses it creates a great way to find allies.

This rationale harnesses technology that is being built and financed now (which means it needs tools and support methods, and a lot of other “picks and shovels”), while leveraging technology that was recently built and financed and is eager for more widespread adoption of the existing toolkits, which makes it suitable for using to build the AI-adjacent tools that are in demand now. It’s like two harmonics producing constructive interference - it makes two waves into one larger wave, which gives me more momentum to surf against.

This was a learning process, and I iterated against my general paradigm repeatedly as I learned more. Neither of us have the patience to go through that in excruciating detail, so I’ll cover the highlights in my next post.

Extra credit answer: A customer gets a product, a client gets a service.

Challenge: Is software a product or a service?

Thanks for reading, more to come soon.