You just spent 45 minutes writing a single email. You polished the subject line. You rewrote the CTA three times. You added a PS because someone told you PSs increase clicks.
Here is what the highest-performing email teams do instead: nothing.
They do not write the email. They write a rule that produces it.
The rule takes 5 minutes to write. It runs forever. And it outperforms your 45-minute draft on day one — because the rule is doing something your draft cannot: deciding whether the email should exist at all.
The Mechanism: Decision Rules
A decision rule is an if/then statement. That is it.
IF [condition] AND/OR [condition]
→ THEN send [email type] with intent [X]
Here are three examples:
| Rule |
Condition |
Action |
| Activation guide |
User completes setup AND 24 hours since last login |
Send "next step" walkthrough |
| Upgrade prompt |
User reaches 5 feature uses AND is on free plan |
Send "you have outgrown free" offer |
| Re-engagement |
User has not logged in for 7 days |
Send "here is what you are missing" recap |
Each of these rules replaces an email draft. Not one draft — every draft this email will ever need for every user who matches the conditions.
The rule does the deciding. The email is just the output.
Why a Rule Beats a Draft
Two teams receive the same AI writing tool.
- Team A uses it to write emails faster. They generate subject lines, body copy, and CTAs in seconds. Their throughput doubles.
- Team B uses it to write decision rules. They define conditions, triggers, and segments. The AI populates the copy into the rules.
After six months:
- Team A is sending more emails. Their open rates are flat. Their unsubscribe rate is climbing.
- Team B is sending fewer, more targeted emails. Their open rates are up. Their revenue per send is up. Their system requires less manual work every week.
The difference is not the tool. It is the mechanism.
Team A optimised writing. Team B optimised deciding. Writing is a cost centre that scales linearly. Deciding is an asset that compounds.
How to Write a Decision Rule
You need three things: a condition, a trigger, and an intent.
1. The condition
What must be true about the user?
- Behavioural: "completed setup," "used feature 5 times," "has not logged in for 7 days"
- Attribute-based: "on free plan," "enterprise account," "signed up via referral"
- Temporal: "day 3 since signup," "30 days since last purchase"
Bad condition: "subscriber on our list." That is everyone. It does no deciding.
Good condition: "subscriber who completed setup, is on the free plan, and has not logged in for 24 hours." That is a specific state that triggers a specific action.
2. The trigger
When does the rule evaluate?
- Immediately after an event (user completes setup → check rules → send or wait)
- On a schedule (daily batch check for users matching conditions)
- At a milestone (day 1, day 7, day 30 since signup or last action)
3. The intent
What is the single outcome this email exists to drive?
- Activate the user on a feature
- Convert them to paid
- Prevent churn
- Collect feedback
If you cannot name the intent in one word, the rule does not have a clear purpose. The copy will drift.
Before and After: The Welcome Email
A typical welcome email starts as a draft:
"Welcome to [product]! We are excited to have you. Here are three features to get started..."
It goes to every new signup. The same message for everyone.
Now convert it into decision rules:
| Rule |
Condition |
Email |
| 1 |
User completed setup within 10 minutes |
"You are fast. Here is an advanced workflow." |
| 2 |
User did not complete setup within 24 hours |
"Stuck? Here is a 2-minute video walkthrough." |
| 3 |
User completed setup AND invited a teammate |
"You brought your team. Here is the collaboration guide." |
| 4 |
User completed setup AND has not logged in for 3 days |
"You left something unfinished. Pick up where you stopped." |
| 5 |
User is on day 7 AND has not completed setup |
"Most people miss this step. Try this one thing." |
The draft is gone. Five rules now produce five different emails for five different user states. The copy in each email can be mediocre — the targeting makes it perform.
The Test
Open your email platform. Look at your last 10 sends.
For each one, ask: could I write the decision rule that produced this?
- If yes, you have a system. The rule is an asset.
- If no, you have a campaign. The email was a one-off.
Most teams find that 8 or 9 of their last 10 sends were campaigns, not system outputs.
That is the gap.
The Bottom Line
Stop asking what the email should say. Ask what rule should produce it.
A draft serves one moment. A rule serves thousands.
Write fewer emails. Write better rules.
Related Articles
