Pay particular attention to numbers, customer names, product capabilities, case-study outcomes, comparisons, and objection responses. Remove or rewrite anything unsupported, outdated, confidential, or too broad.
Before you start
- Identify the company segment and persona the content belongs to.
- Gather current approved product, marketing, sales, and case-study sources.
- Assign subject-matter owners for product, claims, and brand review.
- Confirm which sections are AI-drafted and which are already approved.
- Plan another review after major segment, product, pricing, or positioning changes.
Review the Knowledge DB by section
| Section | Review for | High-risk content |
|---|
| ICP Profile | Correct company and persona context. | Overly broad fit or outdated targets. |
| Pain Points & Value Propositions | Evidence, relevance, and persona specificity. | Unverified pain or guaranteed outcomes. |
| Products & Services | Current capabilities and availability. | Roadmap items presented as live features. |
| Case Studies | Permission, customer identity, numbers, and context. | Invented results or confidential details. |
| Objection Handling | Accuracy, tone, and approved response boundaries. | Unsupported comparisons or legal claims. |
Approval is content ownership
AI-drafted content is a starting point. An approver takes responsibility for the accuracy and intended use of the approved version.
- Verify every number against an approved source.
- Use customer names only with required permission.
- Describe capabilities as available only when confirmed.
- Keep segment and persona context explicit.
- Remove internal notes, links, and unresolved hypotheses.
- Record the owner and date of review.
Steps
- **Select the segment and persona. **Confirm which audience the content will influence.
- **Review the ICP Profile. **Compare it with approved scoring and positioning logic.
- **Review pain points and current solutions. **Remove generic, speculative, or sensitive claims.
- **Review Products & Services. **Confirm feature availability, scope, and wording with an authoritative source.
- **Review Case Studies and Objection Handling. **Validate permission, facts, numbers, comparisons, and claim limits.
- **Edit or delete unsafe content. **Do not approve text merely because it sounds convincing.
- **Approve sections deliberately. **Approve individual sections; use Approve All Sections only after a complete pass.
- **Schedule the next review. **Repeat after changes to segments, product, pricing, positioning, or evidence.
Review and safety checks
- The correct segment and persona are selected.
- Every number and customer claim has an approved source.
- No roadmap item is described as generally available.
- Pain points and objections are specific and respectful.
- Internal links, notes, and confidential details are absent.
- Each approved section has an accountable owner and review date.
- Downstream generated copy still receives human review.
Troubleshooting
| Problem | Solution |
|---|
| AI text contains convincing but incorrect numbers | Do not approve it. Remove or replace every number without an authoritative source. |
| Messages use the wrong pain points | Check company segment, persona, approval status, and downstream template context. |
| A product description is outdated | Update it from the current approved source and review dependent sections. |
| Approve All Sections was used too early | Reopen the affected scope and review each section; remove unsafe approval where supported. |
| A case study contains confidential details | Remove them and confirm permission and approved wording before restoring any customer reference. |
Frequently asked questions
What does approved mean in the Knowledge DB?
The content has been reviewed and may be used as a trusted input, subject to downstream message review.
Is AI-drafted content safe to approve without edits?
Only after a human verifies every claim, number, name, capability, and segment-persona context.
When should I use Approve All Sections?
Only after a full review of every visible section for the selected scope.
How often should the Knowledge DB be reviewed?
After meaningful changes to segments, product, pricing, positioning, evidence, or customer permissions.
Why are the wrong pain points appearing in messages?
The persona or company segment may be wrong, the section may be outdated, or the template may lack context.
Can internal Jira or Confluence notes remain in approved content?
No. Keep internal sources and unresolved editorial notes out of content that can feed customer-facing output.
Keep reading
Set up copywriting, prompts, and templates
Use Copywriting settings for shared style and delivery rules, and use Templates for reusable prompts, replies, and Deep Research instructions. Each campaign still needs the correct list, signal, persona, language, and sender context.
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.
