The Send Test: Teardown of Three Instagram Captions
Objective: Given three synthetic-realistic Instagram post specimens (a Reel script, a carousel caption, and a hashtag-heavy caption), identify the structural defects that would sink sends-per-reach, hashtag relevance, and caption front-loading, and separate them from plausible-looking non-defects.
You're reviewing draft Instagram content for Coinbase's consumer app team before it gets scheduled. Three drafts got flagged in standup as 'feels off' but nobody could say exactly why.
Score each specimen against the lesson's sends, hashtag, and caption rules. Distinguish a real defect from a stylistic choice that just looks unusual.
Before you start
What you'll need
Free path (everything below is enough to finish)
No account friction, sufficient for a 3-row teardown log
The process
Specimens to review
Score this Reel caption against the lesson's sends and hashtag rules. List every defect you find, ranked by severity.
Reel caption: 'What would YOU check first when auditing a website? Comment below and let's discuss! 👇 #digitalmarketing #seo #marketing #socialmediamarketing #contentcreator #smallbusiness #entrepreneur #marketingtips #growthhacking #onlinemarketing #digitalstrategy #brandbuilding #marketingagency #contentstrategy #socialmediatips'
Specimen: synthetic, realistic
This carousel has a structural design flaw independent of its content quality. Find it, and note anything else worth flagging.
Carousel, slide 1: 'Our biggest audit finding this month.' Slide 2: '...and it's about crawl budget, which we mentioned yesterday.' Slides 3-5: three data points, one per slide. Slide 6: caption ends with no closing line.
Specimen: synthetic, realistic
Average watch time on Reels is under 10 seconds. Evaluate whether this script's structure fits that reality.
Reel script: 0:00-0:04 'Hey guys, welcome back to the channel, today we're covering something big.' 0:04-0:12 topic A (budget). 0:12-0:20 topic B (bidding). 0:20-0:28 topic C (creative). Caption includes a raw link: 'Full guide here: https://example.com/guide'
Specimen: synthetic, realistic
Final deliverable
A scored teardown log covering all three specimens, each defect tagged by severity, plus one rewritten line per specimen that fixes its critical defect.
See a reference example
Snowflake teardown log (excerpt) SPECIMEN 1, Reel caption CRITICAL: comment-bait CTA, no forwardable line FIX: 'The one crawl-budget check most audits skip, save this before your next one.' MODERATE: 15 hashtags (cap is 5) SPECIMEN 3, Reel script CRITICAL: 4-second generic intro before the hook FIX: open on-screen at 0:00 with 'Your crawl budget is being wasted on pages Google will never rank.'
Success criteria
You're done when you can:
- Identifies the critical defect in all three specimens without prompting
- Correctly separates at least 4 of the 6 distractors from real defects
- Each proposed fix ties directly to the lesson's sends, hashtag, or Reel-structure rule it violates