← Decision Lab
02Architecture

Build / Buy / Bend

Do you need to own this capability, or just have it?

  • Build vs. buy
  • Differentiation
  • Ownership
  • Speed to value
  • Optionality

Interactive prototype

Open in a new tab →
https://build-buy-bend.lovable.app

If the prototype doesn't load here, open it in a new tab.

The situation

A venture firm wants institutional memory: every founder meeting, CRM record, diligence doc, and past investment, searchable and trustworthy.

They have a small technical team, an existing document stack, and a partner who wants it working before the next fund close.

The obvious answer

Build a custom retrieval platform. It's the firm's proprietary knowledge, so it should be proprietary software.

The question underneath it

The real question is where differentiation actually lives. Storage, search, and permissions are commodity. Judgment about what a good answer looks like is not.

What I'm paying attention to

  • Differentiation

    Build only where the capability is the moat.

  • Time to value

    A bent stack shipping in three weeks teaches more than a build shipping in nine months.

  • Ownership cost

    Custom software is a permanent staffing decision, not a one-time project.

  • Switching cost

    Buying is fine until the data model belongs to the vendor.

  • Optionality

    Bending keeps the decision reversible while you're still learning the problem.

My take

Bending is the option that gets skipped, and it's usually the right first move. Extending tools you already run gets you a working answer, real usage data, and a much sharper spec for whatever you build later.

Build where you differentiate. Rent the commodity layer. Bend the stack while you're still learning.

Related writing

Build vs. Buy Is Missing a Third Option

Substack essay — Coming Soon

Office Hours with Chanel

Working through a version of this in your own company?

Private AI office hours for leaders who need to turn messy AI ideas into clear strategy, useful workflows, or safe internal agents — with senior-engineer judgment in the room.