The 30-Reels-a-month pipeline guide
This guide is based on the 1-hour-a-month AI workflow I used to grow a winery from $15K to $80K in monthly revenue featuring Destiny Lalane's Instagram workflow, published on the AI Lab by ActiveCampaign.
Get the quick-start guide
This is the compressed build version of Destiny Lalane’s Reels pipeline. The full post wires Google Sheets, Claude, Zapier, Descript, and Buffer into a one-click system. This guide gets you to your first batch of on-brand captions and a manual posting flow in a single focused session, so you can prove the concept before automating the rest.
What you’ll build
A Google Sheet that, with one prompt to Claude, writes 30 unique Instagram Reels captions tuned to your business. You’ll edit the ones you like and schedule them by hand. It’s the same engine that helped Wanderlust grow reservation revenue from $15,000 to $80,000 a month, minus the full automation plumbing.
Time commitment
- Quick version (this guide): 1 to 2 hours
- Full version (see original post): a few hours of setup, then about 1 hour a month
The build (minimum viable version)
The build (minimum viable version)
Step 1: Write your brand guide—20 minutes
Open a blank doc and answer three questions: who your customers actually are, what offers you push, and which offers make the most money. Write the answers as plain bullets, the more specific the better. “Tiki Tuesday happy hour all night, $5 off all tiki drinks” works; “great drinks and a fun atmosphere” doesn’t.
Quick version: Skip the polished document. Ten specific bullets in a note is plenty for your first batch.
You’re doing it right if: Someone who’s never visited your business could describe what makes it different after reading your bullets.
Step 2: Generate captions with Claude—20 minutes
Open Claude, paste your brand guide, and ask it to write 30 Instagram Reels captions, each with a headline, caption, and hashtags. Tell it to vary the hooks and lean on your highest-margin offers. The full system batches this into three Claude calls so the captions don’t fall into repetitive patterns, so ask for them in three groups of ten if you notice repetition.
Quick version: Run it straight in the Claude chat window. You don’t need the Apps Script, the Web App, or Zapier to get usable captions today.
You’re doing it right if: You can read ten captions in a row and none of them feel like copies of each other.
Step 3: Review, approve, and schedule—30 to 40 minutes
Read every caption. Edit anything off-brand, cut the weak ones, and keep the 15 to 30 you’d actually post. Then open Buffer, create a draft for each keeper, attach the matching phone clip, and set a posting date and time.
Quick version: Post manually to your channel for the first week instead of building the Buffer bulk-upload CSV. Prove the captions land before you automate scheduling.
You’re doing it right if: You have a queue of approved Reels with videos attached and dates set, ready to go out without you touching them again.
You’re done when…
You’re looking at a batch of approved, on-brand Reels captions paired with clips and scheduled to post. You produced them in one session from a brand guide and a single AI workflow, not by writing each one from scratch.
When you’re ready to go further
- Build the Google Sheet command center with the Apps Script so the whole team can run it
- Add Zapier as the orchestrator so one button fires three Claude calls and writes captions back automatically
- Connect Descript (Enterprise) to auto-create video projects, then bulk-upload to Buffer with a CSV
- Use what goes viral as feedback: feed the winning hooks and offers back into your brand guide
Related
More data from the AI Lab.