How to group LinkedIn leads for targeted AI campaigns
Learn how to segment LinkedIn leads by fit score, persona, and intent to run targeted AI outreach campaigns that avoid generic messaging and increase replies.
Get startedShort answer
Segmenting LinkedIn leads means dividing your audience into smaller groups based on shared traits like job title, industry, or behavior. This lets you adjust AI messaging to match the specific pain points of each group so your outreach doesn't sound generic.
Define your lead segments
A common mistake is to put everyone with the same job title in one bucket. A marketing manager at a 10 person startup and a marketing manager at a 10,000 person company do not buy for the same reasons.
The useful split is by pressure. Job title alone is a weak filter. A founder worries about runway and whether the next hire pays for itself. A VP worries about hitting a number this quarter and getting the team to adopt something new. One message cannot carry both.
List the buyer types you actually talk to. For each one, write the sentence they would say on a bad day. That sentence becomes the center of the segment.
Generic approach
- One message for every marketing manager, regardless of company size.
- A subject line that names the job title and nothing else.
- One value proposition for the whole list.
Segmented approach
- A group for marketing managers at seed stage startups and a separate group for large companies.
- A first line that names the pressure that role feels this quarter.
- A proof point from a company of similar size.
Identify your segmentation criteria
LinkedIn gives you enough information to separate leads by company, role, and behavior. You do not need a perfect data set. You need a reason to put two people in different groups.
A fit score is the fastest way to sort. You describe the customer you want, score each lead from 0 to 100 against that description, and set a line below which leads are not kept. The score does not replace judgment, but it stops you from spending a campaign on people who were never going to buy.
Behavior is the strongest signal. Someone who commented on a competitor's post about a missing feature is warmer than someone who only has the right title. Put behavior-based leads in their own group so the first message can refer to what they did.
Firmographics
- Criteria type
- Company size, industry, region
- Examples
- Whether the company can buy at all
Personas
- Criteria type
- Job title, seniority, team size
- Examples
- Tone and level of detail
Intent
- Criteria type
- Competitor post comments, recent shares
- Examples
- Which leads get a message this week
Fit score
- Criteria type
- 0 to 100 match against your customer profile
- Examples
- Which leads stay in the list
Build your lead lists
Turn each segment into a separate list before you write any message. If you search once, export 500 names, and try to sort them later, you will make mistakes. Run one search per segment and keep the list small enough to check.
Names should tell you what is inside. 'SaaS founders, seed stage' or 'Enterprise HR, North America' is enough. Avoid clever labels that only make sense to you three weeks later.
Check the first 20 rows by hand. If the list has people who do not match the segment, fix the search before you add more. A clean list of 40 people is more useful than a mixed list of 300.
- 1
Run a targeted search
Filter for one role and one company type, such as VP of Sales at software companies.
- 2
Isolate the profiles
Export the results to a CSV or save them to a dedicated lead group.
- 3
Repeat for the next segment
Run a new search for founders at seed stage software companies and save them separately.
- 4
Check the fit
Read the first 20 profiles. If more than a couple do not belong, tighten the filters and start again.
Write targeted prompts for each group
AI writes what you ask for. If the prompt says 'be professional and helpful', you get text that could go to anyone. Give the model the segment, the pressure, and the one thing you want the person to do.
For a founder group, the prompt can mention time and cash pressure, and whether the fix is worth a week of attention. For a VP group, mention the quarter, the team, and the reporting line. The product is the same. The reason to reply changes.
Read the first few drafts before you let them send. If a draft could go to the other segment without edits, the prompt is still too broad.
Example
- You
I saw you commented on Northwind's post about lead quality. Most seed stage founders I talk to hit that when they hire their first two reps. Is it a problem for you this quarter?
- Prospect
It is. We get replies but they are mostly students and agency owners.
- Suggested reply
That usually means the list is too broad. I can send the three filters I use to keep students and agency owners out. Want me to?
Write the pressure first, then the product. If the first sentence could go to another segment, rewrite it.
Paste this job
Segment: seed stage SaaS founders Pressure: lead quality drops when they hire the first reps Signal: they commented on a competitor post about bad leads Goal: a 15 minute call Tone: direct, no pitch Write a connection note under 200 characters that opens with why they became a lead.
Match the message to the segment's pressure
Segments fail when the only difference is the first name. If the founder message and the VP message both say 'I help companies grow', you have not segmented anything.
Change the proof for each group. A founder wants to know that another founder with a small team got a result without adding headcount. A VP wants to know that a peer at a similar sized company fixed the same process. Use one example per segment, not a list of logos.
The call to action can change too. A founder may take a 15 minute call. A VP may want a short document first. They are different buying processes.
A segment needs a different reason to reply. Changing only the first line leaves the message generic.
Keep segments small and separate
One lead should sit in one active campaign at a time. If the same person gets a founder message and a VP message in the same week, they will notice. Pick the segment that fits the person best and let the other campaign run without them.
Review lists monthly. People change jobs and companies change size. A fit score from last quarter may be wrong now. Remove anyone who replied or booked. Remove anyone who said no.
LinkedIn does not publish exact daily or weekly limits. Keep an eye on pending invitations and how often people accept, and slow down when either gets worse.
One lead, one active campaign at a time.
Doing this with Kinlead
Kinlead finds people on LinkedIn who match the customer you describe, or your website, and scores each lead from 0 to 100 against that profile. Kinlead does not keep leads below your fit line.
You can keep leads in lead groups or import a CSV of LinkedIn profile URLs. For high-intent leads, the Steal Customers agent watches competitor company pages and optional founder or employee profiles, reads comments on their recent posts, and turns commenters who look like buyers into leads. Each lead keeps the post URL, what the post was about, and what they commented.
Point any campaign at any group. Kinlead's AI writes connection notes and follow-up messages from the campaign goal, tone, and your sales playbook. You can read and edit queued messages on the Actions page before they send, and you can pause any campaign.
- 1
Find and score leads
Describe your customer or use your website. Kinlead scores each lead and drops anyone below the fit line.
- 2
Create lead groups
Save leads to a group by persona, or let Steal Customers build a group from competitor post commenters.
- 3
Launch a campaign
Point a campaign at one group. Choose the goal, tone, and playbook so the messages match that segment.
- 4
Handle replies
Kinlead puts replies from every campaign in one inbox, sorts them by intent, and shows interested replies first. You get an email when someone replies.
Frequently asked questions
Related
Get started with Kinlead
Prefer markdown? How to group LinkedIn leads for targeted AI campaigns.md