# How to score an imported LinkedIn list with AI
> Import a CSV of LinkedIn profile URLs, describe your buyer in plain sentences, and let AI scoring and a fit line cut the list before you send anything.
- HTML: https://getkinlead.com/ai-lead-scoring-csv-import-workflow
- Markdown: https://getkinlead.com/ai-lead-scoring-csv-import-workflow.md

**Short answer:** Import a CSV of LinkedIn profile URLs into a tool that scores each person against a written description of your buyer. Every profile gets a score from 0 to 100, and a fit line you set decides which leads stay in the list and which get dropped before any outreach goes out.

## Write the buyer description before you touch the CSV

Most imported lists get scored against nothing. Someone exports two thousand names, points a tool at them, and the tool guesses what a good lead looks like from a job title. What comes back is a ranked list of strangers with no reason behind the ranking, and you still have to open the profiles yourself.

Write the buyer description first, in plain sentences, the way you would explain it to a new SDR on their first morning. Name the titles that buy, the company shape that fits, the signals that make someone worth a message this week, and the things that disqualify a person no matter how senior the title looks.

Keep it to a paragraph. Descriptions with twenty criteria score inconsistently, because half the criteria can't be seen on a profile in the first place.

**Description that scores nothing**

- We sell to decision makers at fast-growing companies.
- Our ideal customer is anyone who needs better sales results.
- Target B2B companies of all sizes.

**Description that scores well**

- Head of Sales or VP Sales at a B2B software company with 20 to 200 employees.
- They have at least two account executives and have posted about hiring SDRs.
- Disqualify agencies, consultancies and anyone selling to consumers.

```
We sell: [one line on what the product does and who it's for]
Buyer titles: [three to five exact titles as they appear on LinkedIn]
Company shape: [industry, headcount range, funding stage, region]
Signals: [hiring, recent funding, new product launch, job change in the last six months]
Disqualifiers: [titles, industries and company types that never buy]
```

## Build a clean CSV of LinkedIn profile URLs

The importer needs one thing per row: a profile URL in the form linkedin.com/in/name. If your list came from an event, a form or a spreadsheet of names, you have to resolve each person to a URL before anything can be scored. Company page links and search result links won't work.

Cleaning takes an hour and saves you from messaging the wrong person. A row that points at a company page, a dead profile or a different person with the same name scores wrong, and you won't notice until a reply comes back confused.

1. **Collect names and companies in one sheet.** Put every prospect in a single tab with columns for name, company, title and any URL you already have. Duplicates are easier to spot in one place.
2. **Resolve each person to a profile URL.** Search LinkedIn for the name plus the company. Copy the URL from the address bar of the profile itself, not from the search results page.
3. **Drop rows you can't resolve.** If two people share a name and you can't tell which one is the buyer, delete the row. A wrong profile scores wrong.
4. **Strip formatting and export.** One column of URLs, no merged cells, no blank rows in the middle. Save as CSV with UTF-8 encoding.

## Sort your criteria into three buckets

Three buckets keep the scoring honest. Must-haves are the things that make someone a buyer. Nice-to-haves are signals that move a lead up the list. Deal-breakers remove a person no matter how good the rest of the profile looks.

Write the buckets down before you import, not after you see the scores. Deciding what counts as a deal-breaker while looking at a ranked list is how you end up keeping leads you already knew were wrong.

| | Dimension | Must-have | Nice-to-have | Deal-breaker |
| --- | --- | --- | --- | --- |
| Job title | VP Sales, Head of Sales, Sales Director | Sales Manager, RevOps lead | Recruiter, HR generalist |
| Company | B2B software, 20 to 200 staff | Services business with a sales team | Consumer brand, staffing agency |
| Activity | Posted or commented in the last 30 days | Changed jobs in the last six months | No activity for a year |
| Market | Sells into North America or Europe | Any English speaking market | Regions you can't support |

## What the AI actually reads on each profile

A profile gives the scorer a headline, the current role, past roles, the about section and whatever the person has posted or commented recently. That's the whole input. It doesn't read an email address or a phone number because those aren't on the profile.

It also can't see budget, buying authority or whether the company is about to freeze hiring. Anything you need that isn't written on LinkedIn has to come from somewhere else, or be treated as unknown rather than assumed.

This is why the description matters more than the model. Two people can score the same list with the same tool and get different results, because one described a buyer and the other described a wish.

> Score what's visible on the profile. If a criterion can't be seen on LinkedIn, it can't be scored, and a tool that pretends otherwise gives you confident numbers with nothing behind them.

## Set the fit line and read what falls below it

The fit line is the score below which a lead isn't kept. Set it high for the first batch, look at what gets cut, then move it. A line at 80 on a list of 500 people might leave you 40 leads, which is fine if those 40 are the right ones.

Read ten profiles that landed just above the line and ten that landed just below. If you disagree with where the cut fell, the description is the problem, not the scoring. Rewrite it, re-score the same CSV, and compare the two exports side by side.

Don't tune the line to hit a number you like the look of. Tune it until the leads above it are people you would actually message today.

## Re-score when the list comes back wrong

Two failures show up on the first pass. Everything scores high, which means the description was too broad and the fit line is doing no work. Or nothing scores high, which usually means the list is wrong: the export pulled in the wrong titles, or the company filter was too loose.

Fix the description before you touch the list. Change one thing at a time so you can tell which change moved the scores. Keep the export from each run so you can see which leads entered and which left.

If you change the description after a campaign has started, the leads already in the group keep the scores they came in with. Re-score the group, or build a new one and point the campaign at it.

## Doing this with Kinlead

Kinlead imports a CSV of LinkedIn profile URLs into a lead group and scores every lead from 0 to 100 against a description you write or your website. Leads below the fit line aren't kept. Any campaign can point at any group, and queued messages can be read and edited on the Actions page before they send.

Pro costs $49 a month after a 7-day free trial, and a card is needed to start the trial. There's more on [lead groups and scoring](/features/lead-groups-and-scoring).

## Related

- [Lead groups and scoring](https://getkinlead.com/features/lead-groups-and-scoring.md)
- [AI lead scoring fit line setup](https://getkinlead.com/ai-lead-scoring-fit-line-setup.md)
- [AI lead scoring from website](https://getkinlead.com/ai-lead-scoring-from-website.md)

## Common questions

**Can I score a list without a Sales Navigator seat?**

Yes. Lead search and scoring use LinkedIn's standard search, so a Sales Navigator subscription isn't required to import and score a CSV of profile URLs.

**What happens to leads that score below the fit line?**

They aren't kept in the lead group. That keeps your campaigns pointed at people who match the description rather than everyone you managed to import.

**Can I score the same list twice?**

Yes, and you should after any change to the description. Keep the export from each run so you can see which leads moved above or below the line.

**Does scoring work on company pages?**

No. Scoring reads a person's profile, so each row needs a profile URL in the form linkedin.com/in/name. Company page links have to be resolved to the people who work there first.

[Create a Kinlead account](https://getkinlead.com/signup) or [book a call](https://getkinlead.com/demo) to talk through pricing and setup.
