DM Script Scorecard
Before you send one script a thousand times, score it ten ways once. One point per criterion — be harsh, because prospects will be.
| Criterion | The test |
|---|---|
| Personalized first line | Could this exact sentence be sent to anyone else? If yes, 0 points. |
| Clear reason for contact | The prospect immediately understands why you, why them, why now. |
| About them, not you | First 2 sentences focus on the prospect, not your offer or bio. |
| One idea only | A single message, a single thread — no feature dumps. |
| Low-friction ask | A question they can answer in one line — not 'hop on a call?' |
| No fake familiarity | No 'hey friend!', no pretending you know them. |
| Credible sender profile | Your profile confirms you're real and relevant before they reply. |
| Human tone | Reads like a person typed it. Contractions welcome. Corporate speak isn't. |
| Honest intent | No bait-and-switch — the conversation goes where the opener implied. |
| Respectful exit path | Easy for them to say 'not interested' without friction or guilt. |
How to use it
- Draft your opener and score it against all ten criteria.
- Rewrite anything scoring 0 — the fixes are usually one sentence.
- Test at 8+ on a small batch and measure the actual reply rate.
- Feed that rate into the calculator to see what the script is worth at full campaign volume.
Common mistakes
- Grading your own script generously — have someone else score it
- Optimizing the pitch when the personalization is the problem
- Testing five changes at once so you learn nothing
- Scaling volume on a script that never earned it in small batches
Frequently asked questions
What's a good score before launching?
8+ out of 10. Below that, iterate the script — sending a 5/10 opener a thousand times just industrializes mediocrity.
How does script quality show up in the forecast?
As reply rate and positive reply rate. Score two script variants, test both on a small batch, then plug each one's actual rates into the calculator to see the revenue difference.
Should I A/B test openers?
Yes — in small honest batches (50–100 each), changing one element at a time. Track reply rates per variant and let the numbers pick the winner.
Turn script quality into forecast numbers.
Model what a better reply rate is worth before you scale.
Forecasts are estimates based on user-provided assumptions. Results are not guaranteed.
Related: Follow-Up Sequence · Reply Rate Calculator