Content ops at a two-person startup looks nothing like content ops at a 50-person team. You can't run a weekly editorial calendar, you can't hire a copy editor, and you don't have time to fight with a CMS. What you can do is pick a few repeatable patterns and commit to them.
The minimum viable workflow
For most founders I've worked with, a sustainable loop looks like:
- Pick one post topic per week tied to a real customer question
- Draft outside your CMS (markdown, not a rich text editor)
- Review once for voice and once for accuracy, as separate passes
- Publish on a consistent day of the week, always
The trap is trying to run a ten-step editorial process with two people. The other trap is doing zero review and shipping something that sounds like a random AI draft.
What AI actually helps with
In my experience, AI is genuinely useful at three stages:
- Research compression: summarizing what competitors said so you know what to leave out
- First draft: a structured skeleton you can edit, not a blob of prose
- Consistency checks: catching tonal drift when you've written ten posts back-to-back
What it's not good at is deciding what to write about. That still comes from talking to customers.
When to invest in infra
If you're publishing weekly for three months and still shipping, it's time to invest in infrastructure. Before then, duct tape is fine. Most content ops tooling is over-engineered for the stage you're at.