What this playbook is for
A one-person business usually does not break because the work is too hard. It breaks because too many tiny coordination jobs stay trapped in your head: follow-up, notes, reminders, handoffs, client prep, lead capture, invoice nudges, content recycling, and admin cleanup.
Good automation should remove recurring drag. It should not turn your business into a maze you are afraid to touch.
Quick take
| Layer | Best tools right now | What they are actually good at | What to avoid |
|---|---|---|---|
| Meeting capture | Granola, Otter | Transcripts, structured notes, action items, searchable memory | Assuming raw transcripts are the same as useful follow-up |
| Thinking and drafting | ChatGPT Projects, Claude Projects | Keeping recurring workflows, prompts, source files, and operating context together | Using AI as a substitute for defining the process first |
| Workflow automation | Zapier, Make, n8n | Moving data and actions between forms, email, CRM, docs, tasks, and alerts | Automating broken steps before you trust the manual version |
| Operating system | Notion | Task views, SOPs, content pipeline, client tracker, recurring checklists | Hiding the whole system inside one person's private pages |
Start with the right automation targets
| Good first automation | Why it is worth doing |
|---|---|
| Meeting notes -> task list | It saves follow-up time every week and reduces dropped commitments |
| New lead form -> CRM -> email alert | You stop losing inbound interest in inbox chaos |
| Published content -> social draft -> archive entry | One piece of work starts compounding across multiple surfaces |
| Invoice sent -> reminder sequence | It reduces awkward manual chasing |
| Research note -> task or doc template | It shortens the path from idea to output |
The operating model
A good solo system reuses the same input
One customer call can become:
- enhanced notes
- next steps
- CRM update
- follow-up email draft
- future content idea
Granola is especially useful when you want your own notes plus transcript context to turn into structured post-meeting notes. The point is not just recording meetings. The point is making meetings usable after they end.
Use Zapier when speed matters most
Zapier is the easiest place to start when you want app-to-app automations fast and you do not need heavy branching right away.
Use Make when you want visual orchestration
Make is stronger when the workflow has more steps, conditions, routers, or logic that you want to see clearly.
Use n8n when you want more technical control
n8n makes more sense when you want self-hosting, deeper customization, or workflows that sit closer to your own stack.
Decision rule
Start with the lightest platform that handles the actual workflow you repeat every week. Do not choose a harder tool because future-you might someday need it.
Keep humans on the judgment edges
It is fine for AI to draft a recap, summarize notes, or suggest next actions. It is not fine to let it silently send important client commitments, refund decisions, or pricing changes without review.
Use AI for:
- summaries
- first-draft emails
- task extraction
- categorization
- content repurposing
Keep human review for:
- promises
- money decisions
- legal or sensitive wording
- anything relationship-critical
A practical solo-ops stack
| If your bottleneck is... | Use this stack | Main result |
|---|---|---|
| Meetings and follow-up | Granola + Notion + Zapier | Calls become notes, tasks, and cleaner follow-up |
| Scattered admin work | Make or Zapier + Notion | Fewer manual updates across tools |
| Custom internal workflows | n8n + Notion + your existing stack | More flexible automations with tighter control |
| Recurring writing and decisions | ChatGPT Projects or Claude Projects + Notion | Less context re-explaining and better recurring drafts |
A simple weekly automation rhythm
- Notice the tasks you did three times this week.
- Pick one that has a clear trigger, a clear output, and low downside if it fails.
- Run it manually one more time and document the exact steps.
- Automate only that version.
- Review the output for two weeks.
- Only then add branches, AI drafting, or extra destinations.
What to standardize first
| Priority | Asset | Why it matters |
|---|---|---|
| 1 | Trigger definition | Every workflow needs a clean starting event |
| 2 | Destination system | The result has to land somewhere visible and reliable |
| 3 | Template output | Summaries, emails, tasks, and notes should follow a stable pattern |
| 4 | Review checkpoint | You need a human check before trust gets too high |
| 5 | Error recovery note | When something breaks, you should know what to fix fast |
Common mistakes
- Automating a workflow before writing down how it should work manually.
- Letting AI summarize everything the same way regardless of context.
- Sending customer-facing automation with no review step.
- Creating five automations that all update the same field in different ways.
- Forgetting that the purpose is less coordination overhead, not more tooling.
Checklist
Operator note
The best solo automation usually feels boring. That is a compliment. It means the business keeps moving even when you are not manually pushing every small piece yourself.