How to determine your AI lead fit line

Learn how to set a lead qualification score threshold from 0 to 100 to balance lead volume with quality when using AI for LinkedIn prospecting. A practical guide for founders and SDRs.

Get started

Short answer

Define your ideal customer profile with mandatory and preferred traits, then assign a point value to each that adds up to 100. Set a fit line threshold to discard leads that score below a specific number based on whether you need more volume or higher precision.

Start with your last five good customers

Open your CRM or a spreadsheet and list the last five customers who renewed, referred someone or said the product was easy to sell. Write down what they had in common before they bought. That list is your starting profile, not a guess about who might buy. If you do not have five customers yet, use five people who took a call and were close to buying. The point is to start from evidence you already have.

Separate the common traits into two columns: must have and nice to have. Must have traits are the ones where the deal breaks without them. If you sell scheduling software to clinics, a decision maker who runs appointments is a must have. A nice to have is a clinic that already uses a specific calendar tool. The first trait decides whether the person can buy. The second trait only changes how easy the conversation will be.

Hard requirements protect you from wasting sends. A lead who misses a hard requirement is a zero even if every other box is checked. A founder with the right title at a bakery is not a fit for a developer tool, no matter how many posts they write about growth. Write the must have list first, before you think about points. If you start with weights, you will talk yourself into giving a weak trait more points than it deserves.

A trigger is the event that makes the problem urgent: a new funding round, a job post for a sales rep, a launch post. Triggers are not hard requirements, but they are useful for sorting leads who otherwise look the same. Add one trigger to your profile if you can name it without guessing.

Vague preferences

  • Someone who looks like they care about growth
  • People in the tech space
  • Companies that are doing well

Hard requirements

  • Job title is Founder or CEO
  • Industry is Software Development
  • Company size is 11 to 50 employees

Write the profile in plain language

Before you assign numbers, write a paragraph that describes the buyer in words a new salesperson could use on a call. Include the job title, company size, industry, the problem they have and the trigger that makes them look now. The paragraph becomes the prompt you reuse when you search for leads. It also becomes the brief you give a teammate who has to review a list for you.

Keep the paragraph short enough to paste into a search box. If it runs to a page, you are describing a market, not a buyer. A useful profile reads like a note to a colleague: who they are, what they run and why they would care this month. Leave out adjectives that cannot be checked. 'Innovative' and 'fast-growing' do not help a search. '11 to 50 employees' and 'posted a job for a sales rep' do.

If you cannot decide between two titles, write both in the paragraph and let the scoring separate them. A profile that names two titles is still clearer than a profile that says 'decision makers'.

Paste this job

Find [job title] at [company type] with [company size] employees in [industry or region]. They likely have [problem] and have recently [trigger, such as hiring a sales team or opening a new office]. Avoid [exclusions, such as agencies or students].

Example: Find founders at B2B software companies with 11 to 50 employees in the US or UK. They likely have a small sales team and have recently started posting about outbound. Avoid agencies and consultancies.

Assign weights to your attributes

Turn the profile into a score by giving each attribute points until the total reaches 100. The attribute that best predicts a closed deal gets the most points. Job title usually gets the largest share because the wrong person cannot buy, no matter how good the company looks. If you sell to a role that exists in every company, title may matter less and industry may matter more.

Company size and industry come next. A useful check is to ask which attribute you would drop last if you could only keep one. That one deserves the highest weight. If two attributes are equally important, split the points evenly rather than inventing a difference. A score is a tool for sorting, not a science experiment. It should be simple enough that you can explain it to a teammate in one minute.

A common split is 50 points for title, 30 for industry and 20 for company size. Use that as a starting point only. If your product works for any industry but only for a specific role, move points from industry to title. Write the weights down where you can see them. If you change them in your head, you will not know why a lead passed last week and failed this week.

Job title match (exact)

Attribute
50 points

Industry match

Attribute
30 points

Company size match

Attribute
20 points

Choose your fit line based on your goal

The fit line is the cutoff. Discard any lead below it. Where you set it depends on what you need this month. If you are testing a new message and need conversations, a lower line gives you more people to learn from. If your calendar is full and you only want buyers who look like your best accounts, a higher line saves time.

Most teams start too high because a perfect score feels safer. Then they get three leads and blame the tool. Start lower, review the leads, and raise the line when you see the same weak pattern repeat. The line is a working setting, not a permanent rule. A line that never changes is either too loose or too tight, and you will not know which until you look at the leads.

If you sell a high-ticket service, a higher line makes sense because one bad call costs more time. If you sell a low-ticket product, a lower line lets you test more messages before you narrow. Write down the goal next to the number. '70 because we need 20 conversations this month' is a decision you can review. '70 because it felt right' is not.

  1. 1

    Determine your phase

    Decide if you need volume (growth) or precision (high quality).

  2. 2

    Set a volume threshold

    Set the fit line at 60 to 70. This keeps leads who match primary requirements but might miss a secondary trait.

  3. 3

    Set a precision threshold

    Set the fit line at 85 to 90. This ensures every lead is a near-perfect match to your ideal customer profile.

Test and calibrate the score

Your first fit line is a guess. Audit it by reviewing a small sample of leads that barely passed and those that barely failed. Ten leads from each side is enough to see a pattern. Read their titles, companies and recent posts. Ask whether you would send a connection request without hesitating. If you would not, the lead should not have passed.

If people you would never message are passing, your line is too low or a weak attribute has too many points. If strong prospects are failing, your line is too high or a hard requirement is too strict. Move the line in 5 point increments and review again after a week of sends. Do not change three things at once. You will not know which change fixed the list.

Example

  1. You

    I'm reviewing the 65-score leads. This person is a Marketing Manager at a 10-person agency. They passed, but they're too small for us.

  2. Prospect

    Our company size weight is only 20 points. They passed because their title and industry matched perfectly.

  3. Suggested reply

    Raise the fit line to 70 or increase the weight of company size to 30 to filter these out.

Never move your fit line by more than 10 points at a time, or you risk accidentally cutting off your entire lead flow.

Keep the score honest as you learn

Update the profile when a deal closes with a customer who does not fit your current description. That is new evidence. Add the trait that made them a good customer, then decide whether it belongs in the must have column or the nice to have column. Do not change the whole profile after one unusual deal. One customer does not prove a pattern.

Also review leads that replied but were not a fit. A reply from the wrong person usually means your title or industry weight is too loose. A reply from the right person who cannot buy means your scoring is looking at the company but not the role. Adjust one attribute at a time so you can tell what changed. Keep the must have list short. Three to five items is enough. A long list of must haves will filter out good leads who do not match a detail that never mattered in a real deal.

  • Add a trait only after you see it in two or more good customers.
  • Remove a trait when it keeps letting in people you never contact.
  • Keep a note of the date you changed the profile so you can compare results.
  • Check the list against a recent deal you lost to see which requirement was missing.

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. You set the fit line, and Kinlead does not keep leads below it. Leads sit in lead groups, and any campaign can point at any group.

Scoring happens during discovery, so your lead groups only hold people who meet your standard. You can see the scoring in practice by booking a demo.

Frequently asked questions

Related

Get started with Kinlead

Prefer markdown? How to determine your AI lead fit line.md