Build vs Buy Software Decision: My One Rule (Simply Explained)

The advice everyone gives you is solving the wrong problem

Every consultant gives you the same speech about whether to build software or buy it. Add up what it costs to build and maintain over three years. Compare that to a monthly subscription. Run the numbers, and about 90% of the time the math says buy.

For years, that was good advice. I gave a version of it myself.

The problem is what that math cares about. It cares about cost and the fact that good software builders are expensive and hard to find. When a developer costs $180,000 a year, of course renting someone else's software wins.

But that's not what I actually care about. I care about owning the part of my business that makes it mine. And that's a different goal, which gives you a different answer.

Here's the rule I run now, across my own DTC fashion brand and every client I build for. I happily pay for the basic plumbing. I refuse to rent the part that runs my actual business. Not out of stubbornness. Because the software that decides how my business works is the part I should never hand to a landlord.

The one line I draw: plumbing vs. your business

The whole decision comes down to a single question. Is this thing basic plumbing, or is it your actual business?

Think of it like running a restaurant. You rent the building. You buy electricity from the power company. You don't try to generate your own power or pour your own concrete. That's the plumbing. Everybody uses it, nobody wins by owning it.

But the recipes? The way you train your staff? The prices on your menu? That's your business. That's what makes you different from the place across the street.

What's worth renting (the plumbing)

This list is short and obvious once you see it:

• The database where information lives
• The servers that run the software
• The AI itself (the thing that reads and writes like a person)
• The banking system that moves money
• The email service that actually delivers mail to inboxes
• File storage

Nobody wins by building their own version of these. Renting them is the right answer every time. I pay for them gladly and never feel bad about it.

What's worth owning (your business)

This is anything that says how my business runs. My prices. My tax rules for my products. My shipping process. The way I group my customers.

When I rent that, I'm renting somebody else's guess at how my own business should work. And paying a monthly fee for the privilege.

Here's the test. If a software subscription is just a thin layer of business rules wrapped around plumbing I could rent directly, then those rules are the part I want to own. The company is charging me a markup to apply logic I understand better than they do.

Why AI changed everything

Let me be honest about why "buy" used to win, because the reason was real.

Building this stuff used to be brutal. Tax rules that change in every state. Payment systems with their own weird formats. A thousand little problems that only show up at 2am when something breaks. That difficulty was real, and the software company absorbed it for you. You paid them to make it go away.

AI didn't make that difficulty disappear. What AI changed is that one person can now handle it.

I built a full bookkeeping system in a weekend. I wrote a sales-tax tool in an afternoon. My toolkit is now 22,000 lines of code, and a big chunk of it is the exact kind of work that used to need a small team and six months.

So the cost of building dropped by about ten times. Meanwhile the benefit of owning it stayed exactly the same. When one side of the scale drops that much and the other side holds steady, the answer flips.

But I won't oversell it. AI does not make upkeep free. When you own the logic, you own the headaches forever. The day a state changes its tax rule, that's your problem now, not a vendor's. That tradeoff doesn't disappear. It just becomes one you can actually afford.

Five subscriptions I cancelled this year

This isn't theory. I cancelled five subscriptions this year, and every one followed the same pattern. I kept paying for the plumbing. I took back the business.

My bookkeeping app: I kept renting the database. I took back the rules for how my business does its accounting. I was paying monthly for a pretty screen on top of something I could build better myself.

My sales-tax service: I still pay for the up-to-date tax rates, because keeping current with thousands of jurisdictions is real plumbing. But applying those rates to my products is now my own code. They were charging me per sale to do math I could do myself.

My bank tool: I pay the bank. I stopped paying for the extra service that guessed how to categorize my transactions. My rules know my business. Theirs guessed.

My email marketing platform: I pay for email delivery, because getting mail to inboxes is genuinely hard and takes years to earn. But who gets which email, and when? That's about my customers. I understand them better than any platform charging me per contact.

My warehouse setup: I rent the warehouse and the shipping. I am not building warehouses. But the rules for how orders get handled and flagged? That's mine now.

Five cancellations, one pattern. Name the plumbing, keep paying for it. Name the business logic, take it back.

When you should still just buy it

This isn't "build everything." That would be just as dumb as "buy everything." So here's when to buy.

Buy when it's genuinely plumbing. Don't build your own database or payment system. There's zero upside in owning those.

Buy when the rules aren't yours and getting them wrong is dangerous. Some compliance tools fit here. If a specialist carries the legal blame and updates the day the law changes, you're not buying software. You're buying someone else's responsibility for getting it wrong. That's worth paying for.

Buy when you can't build and maintain it yet. This is the honest one. If you can't direct AI to build and keep the thing running, the old math still applies to you exactly. The whole argument for building rests on having that ability. Don't skip the prerequisite.

Buy when the software really is the product. Some tools are deep, genuinely hard things. Pay for those.

Try this with your own subscriptions this quarter

Here's the exercise. It takes an afternoon and pays for itself.

List every subscription you pay for, with the monthly cost next to each. Then ask one question for each: is this plumbing, or is it just my own business rules wrapped around plumbing I could rent directly?

Most businesses have three to five where the answer is obvious the moment you ask. Circle those.

Then the honest gate: can you actually build and maintain the replacement? Because owning the logic means owning the headaches forever. If you can direct AI to do it, the math has flipped in your favor. If you can't, leave the subscription alone for now.

That gate is exactly where I come in. I draw this line across a client's whole stack, find the subscriptions worth replacing, and build the owned pieces myself. For my own brand, that meant 3,000+ hours saved a year and 42% less manual work, mostly from refusing to rent logic I could own.

You can run the first pass yourself this quarter. The candidates will be obvious. The build is where it gets real.

Want to explore what AI could do for your business?

Book a free 30-minute strategy call. No pitch deck, no sales team, just a real conversation about your operations and where AI fits.

Book a Discovery Call