The Budget Call: Auditing an Instagram Format-Performance Export
Objective: Given a real 20-post Instagram export mixing single images, carousels, and Reels with reach, saves, sends, and engagement rate, classify posts by format and recommend where next quarter's content budget should go.
You're the social media analyst at Instacart, the on-demand grocery delivery marketplace. You've pulled a 20-post export covering the last quarter and have one quarter's content production budget to allocate across formats.
Sort by format, compute engagement rate per post, and flag which format is the losing bet before a single new post gets briefed.
Before you start
What you'll need
Free path (everything below is enough to finish)
No account friction, handles a 20-row export without any paid analytics tool
The process
1 step
Step 01 of 01
The lesson's format-performance data shows single images down double digits year-over-year on reach, interactions, and engagement rate, while carousels and Reels have split the format landscape between them.
13 carousel rows, 5 Reel rows, 2 single-image rows in the export. Reach, saves, and sends-per-reach all trail on the single-image rows. Where does next quarter's budget go?
Procedure
- Import and freeze row 1
- Add a formula column: engagement_rate = (likes + comments + saves + shares) / reach
- Sort descending by engagement_rate, then group by the `format` column
- Compute the median engagement_rate per format group
- Flag any single-image row scoring below the carousel median
FORMAT MEDIANS (20 posts) Carousel (13 rows): 0.51% engagement, 9.2 avg saves Reel (5 rows): 0.47% engagement, 3.1 avg saves Single image (2 rows): 0.31% engagement, 1.0 avg saves FLAGGED Post #14 (single image, produce recipe card): 0.22% engagement, below every carousel row
Healthy
Budget shifts toward carousels and Reels; single-image posts get reserved only for time-sensitive announcements.
Unhealthy
Continuing to brief single-image posts at the same rate as carousels because 'that's what we've always made'.
What this means
Format is the first filter, not caption quality; a well-written caption cannot fix a format the algorithm structurally under-serves.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| Content calendar still has single images as the default weekly post | Convert the weekly single-image slot to a 4-6 slide carousel using the same source asset | 30 min |
| No one has checked whether Reels or carousels lead on saves this quarter | Add the engagement_rate formula column to the standing weekly reporting sheet | 5 min |
Final deliverable
A format-performance scorecard ranking single image, carousel, and Reel by median engagement rate and saves, with a budget-shift recommendation for next quarter.
See a reference example
Adyen, Q3 Instagram format scorecard (excerpt) CAROUSEL, 13 posts Median engagement: 0.51% | Median saves: 9.2 Recommendation: increase to 3x/week REEL, 5 posts Median engagement: 0.47% | Median saves: 3.1 Recommendation: hold at 2x/week, strongest on reach SINGLE IMAGE, 2 posts Median engagement: 0.31% | Median saves: 1.0 Recommendation: reserve for same-day announcements only
Success criteria
You're done when you can:
- Correctly groups all 20 rows by format and computes a median engagement rate per group
- Flags the underperforming single-image row(s) by name, not just by format average
- Recommendation ties a specific budget shift to the data, not a generic 'do more video' statement