You launched a campaign and some prospects do not match your target audience. The fix is not to panic, and it is usually not to be harsh with those people.
The goal is simple:
- Pause the campaign.
- Understand why the wrong leads entered.
- Treat already-contacted people politely.
- Tighten the persona for the next mission.
- Use feedback to train the AI on what good and bad looks like.
Quick Decision Tree
| Situation | What to do |
|---|---|
| Leads are queued but not contacted yet | Pause the campaign, review them, and remove them from this campaign if needed. |
| Leads were already contacted | Do not be rude, do not block by default. Reply politely and stop future follow-ups if needed. |
| The persona is missing geography, industry, job title, or seniority | Fix the persona, then create a new mission so the new filters apply cleanly. |
| The source is broad, like followers or event attendees | Keep the source, but add a tight persona and enable Exclude Not a Fit. |
| The AI scoring, status, message, MBTI, or microsite feels wrong | Check our dedicated guide: Train the AI with Thumbs Up / Thumbs Down Feedback |
Why Out-of-Scope Leads Appear
Most cases come from one of these causes:
1. Persona filters were missing
The most common issue is a persona that is too broad. For example, the campaign uses AI Lead Finder, but the persona does not specify the right:
- Geography
- Industry
- Job title
- Seniority
- Company size
If your target market is France, but the persona has no geography filter, the AI may find relevant-looking people in Mexico, the USA, Serbia, or any other country.
2. The lead source is broad by nature
Some sources collect people who showed a signal, but not necessarily buyer intent:
- Company followers
- Competitor followers
- Event attendees
- Post commenters
- People from a large imported list
These sources can be useful, but they need a strict persona on top. Many people follow or connect randomly. Some are partners, students, recruiters, competitors, or simply curious.
3. The list is stale
A spreadsheet from a few months ago may include people who changed role, company, location, or seniority.
4. The AI made an interpretation mistake
Sometimes the AI scores a lead as a fit, assigns a status, writes a message, or creates a rationale that you disagree with.
That is exactly why the feedback feature exists: every correction helps your agent learn what your team considers accurate.
1. Pause the Campaign
Stop new actions before changing anything.
- Open Campaigns.
- Open the campaign with the wrong leads.
- Switch the campaign status to Paused.
Pausing stops new connection requests, messages, and follow-ups while you investigate.
⚠️ Pausing does not undo actions that already happened. It only stops future actions.
2. Find What Is Actually Out of Scope
Do not just label the leads as "bad." Identify the exact reason they are out of scope.
Ask:
- Are they in the wrong country or region?
- Are they in the wrong industry?
- Is the job title too junior, too senior, or unrelated?
- Is the company size wrong?
- Did they come from a broad source like followers or commenters?
- Did the AI scoring rationale misunderstand the persona?
Then open the campaign or inbox and use filters such as title, seniority, location, industry, company size, lead tag, or status to separate:
- Leads still waiting for an action
- Leads already contacted
- Leads that are actually relevant
- Leads that are out of scope
💡 Write the problem in one sentence, for example: "The source is correct, but geography was missing, so we contacted people outside France." This makes the fix much clearer.
3. If the Lead Was Already Contacted, Stay Polite
If someone already received a connection request or message, do not block them by default and do not be mean.
Even if they are not the perfect buyer, they are still a person who accepted or interacted with you. A simple, polite message is usually enough:
"Hello, thank you very much for adding me. I hope you'll enjoy my content."
That is it.
If they might be loosely relevant in the future, you can keep the relationship warm. Later, if it makes sense, you may ask for a referral to the right person. But do not treat a wrong-fit lead as a problem to punish.
Use blocking or blacklisting only when you are sure you never want to contact that person again, or when the contact is clearly inappropriate.
4. Fix the Persona Before Launching Again
If the issue came from persona filters, update the persona with the missing criteria.
Review:
- Geography — country, region, or city
- Job title — include and exclude keywords
- Seniority — founder, C-level, director, manager, IC
- Industry — keep it narrow
- Company size — match your actual buyer
- Source quality — followers and commenters need stronger filters
Important: create a new mission when the persona changes
If you change the persona with new filters, create a new mission to make sure the new criteria are applied effectively in the campaign.
Editing the existing campaign helps for future configuration, but a new mission is the cleanest way to ensure the lead generation uses the corrected persona from the start.
5. Use "Exclude Not a Fit" for Broad Sources
If the source is broad, enable Exclude Not a Fit after tightening the persona.
This is especially useful for:
- Competitor followers
- Company followers
- Event attendees
- Post commenters
- Large lists with mixed quality
The feature helps exclude prospects that the AI identifies as not matching your persona before they receive a message.
💡 Exclude Not a Fit only works well if the persona is precise. If the persona is vague, the AI has less signal to decide what "not a fit" means.
6. Update Messages if Needed
If the campaign is already launched and the message needs a small correction, you can edit the message content.
For example, if the first message forgot to say "Bonjour," open the campaign, edit the sequence, and update the message. Future messages will use the corrected text.
⚠️ Some campaign timing settings may not be editable once the campaign is running, but message content can usually be corrected for future steps.
What Good Looks Like
After fixing the issue, you should have:
- A paused or cleaned campaign
- A clear explanation of why the wrong leads appeared
- A polite approach for people already contacted
- A tighter persona
- A new mission if the persona filters changed
- Exclude Not a Fit enabled when using broad sources
- Feedback submitted on wrong scoring, status, messages, MBTI, or microsite examples
- A habit of validating good AI decisions, not only reporting mistakes
This way, you avoid repeating the same issue, you keep the relationship respectful, and you help your agent learn from real campaign situations.