Build vs. Buy vs. AI
Software Arch
You're the one evaluating options. Add 'can AI do this?' to the decision tree. It changes the calculus for a lot of problems.
Platform
Platform decisions now include: build our own, buy a vendor, or use AI-assisted implementation. You own the tradeoff analysis.
Tpm
Your roadmap prioritization just got a new variable. AI can accelerate some work to near-zero. Factor that into capacity planning.
Build vs. Buy vs. AI
TL;DR
- The old choice: build it or buy it. The new choice: build it, buy it, or let AI help build it faster/cheaper.
- AI shifts the build option: what used to take 3 engineers for 6 months might take 2 engineers for 3 months. Re-run the build vs. buy math.
- David makes these calls. He needs a framework. Here it is.
Every project used to be: build (we have the skills, we want control) or buy (faster, someone else maintains). AI adds a twist: build with AI assistance can be dramatically cheaper. That changes the math.
The Updated Decision Tree
Build when:
- Core to your product or strategic advantage
- AI can accelerate implementation meaningfully
- You have the expertise to direct and validate
Buy when:
- Commodity need (auth, payments, etc.)
- Vendor has scale and security you can't match
- Build would distract from core product
AI-assisted build when:
- Implementation is standard (CRUD, APIs, dashboards)
- You have people who can direct AI and review output
- Buy options are expensive or inflexible
The new question: "Could we build this with AI in half the time?" If yes, the build option just got more attractive.
What This Changes
- Vendor evaluations: Some vendors sold "we'll build it for you." Now you might build it yourself with AI. Revisit the ROI.
- Capacity planning: Same team can take on more. Or you need fewer contractors. Model it.
- Skills: You need people who can direct AI, not just people who code from scratch. Hiring criteria shift.
Don't Overcorrect
AI doesn't make everything a build. Compliance-heavy, security-sensitive, or highly customized work may still be buy. The framework is a lens, not a rule.
Quick Check
A project used to be build or buy. AI adds a twist. What's the NEW question for build vs. buy decisions?
Build or buy. That's the choice. Build = 3 engineers, 6 months. Buy = $200K, vendor lock-in. You pick buy because build is too expensive.
Click "Build vs. Buy vs. AI" to see the difference →
Do This Next
- Pick one upcoming project and run it through the build/buy/AI-assisted lens. What's the best option now? Has it changed?
- Document your criteria for when AI-assisted build is appropriate. Share with your leads. Make it a shared framework.