# Read two numbers, then decide

Story / Test and publish / Reading the numbers. Checked 2026-11-13.

You know whether to fix the opening, fix the body, repeat or drop.

- Use when: A post that has settled, about 72 hours after it went out.
- Phrase: compare reach and engagement with the account's own medians, never with another account
- Page: https://directorchair.ai/publish/read-two-numbers
- JSON: https://directorchair.ai/api/publish/read-two-numbers.json

## Watch out

- Reach without engagement: a fake-news opener had 490,000 plays at 1.0 percent likes and 0.18 percent saves, and led nowhere. (measured; Director Chair field notes on short-form video (one TikTok account, 233 posts and 27 measured clips, read 30 Oct 2026))
- Engagement without reach: the three weakest posts reached 7,600 to 11,700 views yet were saved at 0.9 to 3.4 percent, as much as the hits. (measured; Director Chair field notes on short-form video (one TikTok account, 233 posts and 27 measured clips, read 30 Oct 2026))

## The four outcomes

| | Engagement at or above the median | Engagement under half the median |
|---|---|---|
| **Reach 3 times or more** | A winner. Post the next one on the same promise within 48 hours | The opening works, the body or payoff does not. Keep frame 0, fix what follows |
| **Reach under half** | The body works, the opening did not stop people. New frame 0, re-cut, post again | Drop the idea |

Reach is views divided by the median of the account's ten posts before it. Engagement is likes, saves and shares per view against the account's own medians.

## Prompts

### Any chat model

```
Give me a step-by-step plan I can follow this week, for a short video about [your product], for [your audience]. The technique is read two numbers, then decide: you know whether to fix the opening, fix the body, repeat or drop. How it works: compare reach and engagement with the account's own medians, never with another account. It is right for: a post that has settled, about 72 hours after it went out. Reach without engagement: a fake-news opener had 490,000 plays at 1.0 percent likes and 0.18 percent saves, and led nowhere. Engagement without reach: the three weakest posts reached 7,600 to 11,700 views yet were saved at 0.9 to 3.4 percent, as much as the hits. Be concrete: numbers, seconds and settings, no theory. Ask me one question first if something you need is missing.
```

## Goes well with

- Outlier ratio: https://directorchair.ai/publish/outlier-ratio
- Feed a breakout: https://directorchair.ai/publish/feed-a-breakout
- Hook rate and hold rate: https://directorchair.ai/publish/hook-rate-and-hold-rate

## Source

Own words from Director Chair field notes on short-form video (one TikTok account, 233 posts and 27 measured clips, read 30 Oct 2026)

Text CC BY 4.0, Director Chair (directorchair.ai). Reference clips and their card texts belong to their makers.