Test every new template across different segments and languages. A consistent message is useful only when its facts, tone, CTA, recipient, and sender behaviour are appropriate.
Before you start
- Confirm the approved brand voice and prohibited wording.
- Choose allowed languages and who may act as a sender.
- Review the approved Knowledge DB and research inputs.
- Define channel-specific message and delivery constraints.
- Prepare representative test leads from different segments and languages.
Understand the shared controls
| Control | Use it for | Check |
|---|
| Message Style | Tone, greeting, sign-off, subject patterns, and structure. | The style fits the brand and channel. |
| Content Rules | Required facts, wording blacklist, alternatives, and CTA pool. | Rules do not force unsupported claims. |
| Delivery settings | Message limits, send-delay defaults, and sender constraints where available. | Values match current workspace and channel policy. |
| Templates | Reusable prompts, replies, and Deep Research prompts. | The template includes segment, signal, language, and evidence context. |
| Campaign context | List-specific purpose and next action. | The template is adapted rather than copied blindly. |
Prefer control over superficial consistency
Shared rules should reduce obvious mistakes without making every message identical. Maintain several approved frameworks and CTA options, then give the campaign a concrete reason to use one.
- Keep greetings and sign-offs appropriate to each channel.
- Blacklist inaccurate, insensitive, or off-brand wording.
- Offer approved alternatives instead of only banning words.
- Use CTAs that match the signal and relationship.
- Set one explicit output language for each workflow.
- Keep limits and delays aligned with channel safeguards.
Steps
- **Define shared style. **Set greeting, sign-off, subject patterns, tone, and structure.
- **Configure content rules. **Maintain the wording blacklist, alternatives, evidence rules, and CTA pool.
- **Review delivery defaults. **Set message limits, send delays, and sender constraints only where supported and approved.
- **Create reusable templates. **Use placeholders for segment, persona, signal, language, and verified evidence.
- **Configure Deep Research separately. **Keep research source-backed instead of hiding broad research inside a message prompt.
- **Test several contexts. **Generate examples for different segments, signals, stages, channels, and languages.
- **Review every output. **Check facts, tone, CTA, recipient, sender, exclusions, and channel constraints.
- **Version approved changes. **Record owner, purpose, test set, and review date.
Review and safety checks
- The output language matches the lead and campaign.
- Facts are supported by approved research or Knowledge DB content.
- The wording blacklist and alternatives work as intended.
- The CTA fits the relationship and signal strength.
- No private notes, credentials, or internal links appear.
- Sender, limits, and delays match channel safeguards.
- Messages remain subject to human review.
Troubleshooting
| Problem | Solution |
|---|
| All messages sound the same | Use several approved frameworks and CTAs, then add specific segment and signal context. |
| The wrong language is generated | Check AI output language, lead language, campaign template, and separate-language campaign requirements. |
| A prohibited phrase still appears | Confirm the blacklist is saved and applied, then test a new output and tighten the template. |
| A message invents a product claim | Remove it, verify the Knowledge DB and source, and restrict the template to supported evidence. |
| Delivery settings conflict with a campaign | Use the stricter approved safeguard and resolve the configuration before activation. |
Frequently asked questions
What belongs in Copywriting settings?
Shared style, content, wording, CTA, language, and supported delivery rules.
What belongs in Templates?
Reusable prompts, replies, and Deep Research instructions that campaigns adapt to a list, persona, signal, and language.
Can a template replace message review?
No. Review the final facts, recipient, sender, tone, CTA, language, limits, and delays.
How do I stop every message sounding identical?
Maintain several approved structures and CTAs, and include the specific signal, segment, and relationship context.
Why is the wrong language used?
Check configured output language, lead data, campaign template, and whether mixed-language audiences need separate campaigns.
Should send limits be stored in a writing template?
Keep operational limits and delays in supported delivery settings and campaign controls, not in prompt prose alone.
Keep reading
Review and approve the Knowledge DB
Review every Knowledge DB section before approving it for research or copywriting. Approval means the content is trusted enough to influence customer facing output, so persuasive wording is not a substitute for accurate evidence.
Write effective Deep Research prompts
Write Deep Research prompts around a decision: qualify an account, verify a signal, identify the relevant role, or find a factual conversation angle. A focused prompt produces more useful results than an open ended request to research everything.
Understand P1, P2, P3, and DQ
Use P1, P2, P3, and DQ to make consistent review decisions—not as automatic truth. A priority is useful only when the team shares a definition and checks company fit, persona fit, source evidence, and the next action together.
