# How to import LinkedIn URLs for AI scoring
> Upload a CSV of LinkedIn profile URLs, describe your ideal customer, and score each lead from 0 to 100 so low-fit prospects never reach your outreach.
- HTML: https://getkinlead.com/ai-sdr-csv-import-workflow
- Markdown: https://getkinlead.com/ai-sdr-csv-import-workflow.md

**Short answer:** To qualify LinkedIn profiles, upload a CSV with one column of direct profile URLs. An AI scoring system compares each profile against your ideal customer profile and removes anyone below the fit line you set.

## Prepare your LinkedIn URL list

AI scoring starts with a list of profile links. The cleaner the file, the fewer surprises you get after import. A CSV with one column of direct LinkedIn profile URLs is the safest format. Leave out notes, old company names, email addresses, and anything else that shares the same column as the URL.

Most import errors come from links that are not direct profiles. A company page, a post URL, or a search result link will not score as a person. Open a few links before you upload. If the page asks you to sign in, that is normal. If it shows a company or a post, fix the link.

Use a header row so the importer knows which column holds the URL. Keep one URL per row. Remove blank rows and duplicates. If you have a person twice, keep the profile that looks current.

1. **Clean your spreadsheet.** Open your list in Google Sheets or Excel and create one column with a header named LinkedIn URL.
2. **Verify the links.** Paste the full profile links, such as linkedin.com/in/username. Remove rows with empty cells or links that point to a company page.
3. **Export to CSV.** Save the file as a CSV. Keep the header row and one URL per row.

```
LinkedIn URL
https://www.linkedin.com/in/jane-cole
https://www.linkedin.com/in/marcus-lee
https://www.linkedin.com/in/priya-shah
```

## Define the customer you actually sell to

The score is only as useful as the description behind it. If you tell the system you sell to marketing people, it will return marketing people. That includes interns, agency owners, and people who have not touched a budget in years. Write the profile of the person who can say yes to your offer.

Start with the last three customers who were easy to close. What did they have in common? A title, a company size, a type of team, a reason to buy now. Put those details into plain language. You do not need a formal ICP document. A short paragraph works if it is specific.

Separate must-haves from nice-to-haves. If you sell a tool for agencies, agency experience is a must-have. If you sell to SaaS, a background in product-led growth is a nice-to-have. The AI can weigh the difference when you say which parts matter most.

Write the description as if you are briefing a new salesperson. Name the role, the company type, the team, and the problem you solve. If you cannot write one clear paragraph, the scoring will not fix that.

| | Criteria | Too vague | Specific |
| --- | --- | --- | --- |
| Job title | Sales | VP of Sales or Head of Revenue |
| Company size | Small to medium | 11 to 50 employees |
| Industry | Tech | B2B SaaS for fintech |
| Buying signal | Growth | Series B, new sales leader, or hiring SDRs |

## Set a fit line you can defend

A fit line is the score a lead must reach to stay in the list. The number itself matters less than the rule behind it. Choose the line before you look at the results, then check whether the leads above it look like people you want to contact.

On a broad list, a high line saves time. You will ignore people who match one detail but miss the rest. On a small list built from a webinar or an old customer list, a lower line can make sense. You already know these people have some connection to your world.

Check the line against real profiles. Pull ten leads from above the line and ten from below. If the ones below look just as good, lower the line. If the ones above include people who cannot buy, raise it. The goal is a list you trust enough to work.

Write down the rule you used. A line of 70 means one thing when you require seniority and company type, and another when you only require an industry keyword. The number alone will not help you repeat the result.

**Weak fit line**

- Keeping every profile that mentions your industry
- Treating a job title match as enough
- Letting freelancers through when you sell to agencies

**Fit line with rules**

- Requiring the right seniority and company type
- Dropping profiles that only share one keyword
- Reviewing a sample before the first campaign

> Pick the fit line before you see the scores. Changing it after you look at the names turns scoring into a way to justify the list you already wanted.

## Run the import and check the scores

Once the file and the customer description are ready, run the import. Map the URL column, attach the description or website you use as the scoring profile, and let the system score each lead. A 0 to 100 score is a sorting tool. Treat it as a way to decide who gets attention first.

Read the top and bottom of the list. Open five high scores and five low scores. Look for obvious misses. A founder with the right company size but an unusual title might score low. A person with the right title at the wrong kind of company might score high. Fix the description if you see the same mistake twice.

Keep a short note on what you changed. If you raise the fit line or add a must-have, write it down. The next import will be easier to judge, and you will not keep re-litigating the same choices.

Check the import count before you move on. If you uploaded 200 URLs and only 120 appear, find out why. Missing rows are usually blank cells, duplicate links, or links that point to a company page.

1. **Upload the CSV.** Import the file and map the column that holds the LinkedIn profile URL.
2. **Attach the scoring profile.** Use the customer description or website that explains who you want to reach.
3. **Set the fit line.** Choose the minimum score a lead needs to stay in the list.
4. **Review a sample.** Open five high scores and five low scores. Adjust the description if the same mistake appears twice.

## Use the scored list in outreach

A scored list is still a list. The work starts when you decide what to say to the people at the top. Use the details that made them score well. If the score came from their role, their company type, and a recent change, mention one of those things in the first line.

Do not send the same message to everyone above the line. The score tells you why someone fits. It does not tell you what they care about this week. Read the profile before you write. A short, specific note beats a clever template.

Keep the first message small. Ask for a conversation, not a demo. If the person replies, move the conversation to the next step. If they do not, leave them alone for a while. A clean list still needs judgment about how people respond.

Track what happens after the first message. If people at the top of the list reply and people just above the line do not, your fit line may be too low. If nobody replies, the problem is probably the message or the offer, not the score.

- **you:** Saw you moved into revenue ops at Northwind last month. We help teams with the handoff between sales and onboarding. Worth a short chat?
- **them:** Maybe. We are looking at that handoff now. What does the first call cover?
- **you:** We can cover your current handoff steps and where deals get stuck. Fifteen minutes, no deck.

## Doing this with Kinlead

Kinlead lets you import a CSV of LinkedIn profile URLs into lead groups. You can describe the customer you want in plain text or point the scoring at your website. Each lead gets a score from 0 to 100, and leads below the fit line are not kept.

Any campaign can point at a lead group, so the scored list becomes the audience for outreach. You can also export the group to CSV. The [lead groups and scoring](/features/lead-groups-and-scoring) page shows how groups and scores fit together.

- Import a CSV of LinkedIn profile URLs.
- Score leads from 0 to 100 against your description or website.
- Keep leads in groups that any campaign can use.
- Export a group to CSV.

## Related

- [Lead groups and scoring](https://getkinlead.com/features/lead-groups-and-scoring.md)
- [AI LinkedIn outreach](https://getkinlead.com/features/ai-linkedin-outreach.md)
- [Book a demo](https://getkinlead.com/demo.md)

## Common questions

**Can I import leads without a Sales Navigator account?**

Yes. A CSV of LinkedIn profile URLs does not need a Sales Navigator seat. The import only needs the profile links.

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

They are not kept in the scored list, so they do not reach your outreach. You can still keep the original CSV if you want to review them later.

**Do I need email addresses in the CSV?**

No. Scoring uses the LinkedIn profile URL to read the person's role, company, and experience. Email addresses are not part of the score.

**Can I change the fit line after I see the scores?**

Yes. Change the line and review a sample again. If you move the line, check that the leads above it still match the customer you described.

**How large a batch should I import?**

Start with a batch you can review. There is no useful number that fits every account, so choose a size where you can open a sample and check the scores before you run outreach.

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