Skip to main content

Using AI Enrichment

Use the AI Research Agent to score, classify, or qualify the companies in your lists, and see how many credits each run uses.

Written by Ruthie Keith

What is AI enrichment?

AI enrichment adds new data to the companies in your list. You describe what you want in plain language, and the AI answers for each company in a new column. Use it to qualify leads, segment a list for a campaign, or pull specific information about each company.

AI enrichment returns one of four output types:

  • Score (0–10): rank companies by how well they match your criteria

  • Text: extract or classify information

  • Number: get an estimated value

  • Yes/No: qualify companies against a true/false question

What data does AI enrichment use?

AI enrichment searches Leadfeeder's own data for each company: firmographics, signals, job ads, and the company's website. For job ads, it checks the company's website and our job ads database.
​

AI enrichment doesn't search the open web. You can still ask about things like open positions, expansion plans, or team growth, and you won't pay extra for it.
​

How much does AI enrichment cost?

From 17 November 2026, AI enrichment costs 1 credit per 100 enriched companies (0.01 credits per company). You pay the same rate in Lists, Workflows, and the API.

  • Previews don’t use credits. You can run up to 25 previews per account per day.

  • Each data point in the Enrich panel shows a Free or Uses credits badge. Data points already in your list show a green check mark.

  • The panel shows the credit estimate for a run before you click Save and run. The number updates when you change which companies to enrich.

  • You only pay for companies that were enriched. Failed records aren’t charged.

  • Each run uses credits, including re-runs.

  • Enrichments you ran before 17 November 2026 aren’t charged.

  • Workflows with an AI enrichment step keep running and use credits from 17 November. Check your automations if you want to limit credit use.

  • If you don’t have enough credits, a run from Lists doesn’t start, and the panel shows how many credits you need. A Workflow stops at the enrichment step, so companies that weren’t enriched aren’t passed on to later steps (for example, to your CRM). The Workflow owner gets an email and a notification. Enrichment requests through the API are skipped.

  • AI enrichment isn’t included in the Lite and Discover plans.

Finding contacts with AI costs the same rate: 0.01 credits for each company you search, however many contacts it finds. Revealing the contacts’ details costs extra: 0.2 credits per contact for email only, or 1 credit per contact for email and phone. Adding contacts by buyer persona is free. The contacts show in your company list, not in a separate contact list.
​

For more on credits, see How are credits spent?
​

When should I use AI enrichment?

Use AI enrichment when:

  • You want to add deeper insights to your company lists automatically.

  • You have a large company list to prioritize or qualify.

  • You want to target companies for a specific sales or marketing campaign.

  • You need to qualify leads on nuanced criteria like values, growth, or focus areas.

  • You need specific information about companies in a structured format.

How it works

1. Create a company list

  • Find companies in Companies and save them as a list.

  • Filter your Web Visitors feed and build a list.

  • Create a new list in the Lists app and import companies from a CSV or Excel file.

AI enrichment works with all company lists, including CRM and Web Visitors feed lists.
​

2. Open the AI Research Agent

  1. Go to the Lists app and open the list you want to enrich.

  2. Optional: select companies first. The panel then picks them for you in step 7.

  3. Click Enrich above the table.

  4. In the Companies section, click AI.

  5. On AI Research Agent, click Configure.

The Enrich panel splits data points into Companies (General, Corporate Contact Details, Location, News & Signals, Social Media Accounts, Industry, Technology & Compliance, Financials, Custom, AI, and in Lists also Web Engagement) and Contacts. You can search for a data point by name. Workflows use the same panel.
​

Enriching from Companies: select your companies and click Enrich. Leadfeeder creates a new list for the results, so set the list name, privacy, and folder in the panel.
​

3. Describe what you're looking for

Under Describe what you are looking for, write your prompt in plain, specific language. Your prompt should match the output type you choose in the next step.
​

💡 How to write a good prompt

  • Be clear, like you're explaining it to a smart friend.

  • State what to include and exclude.

  • Use keywords.

  • If it's hard to spot, write "very likely" to help the AI focus.

  • Match your prompt to your output type (for example, for Yes/No outputs, ask yes/no questions).

✅ Good prompt examples for Score (0–10)

  • I am looking for companies which produce food or beverages - except chain bakeries. There must be clear proof that the company produces food or beverages itself. Excluded are the following company types.

    • Chain bakeries or butcher shops with direct sales - no matter if they produce themselves or not.

    • Retail businesses (no brick-and-mortar stores)

    • Delicatessens or market stalls

    • Pure wholesalers or retailers without their own production

  • I want to prioritize companies that very likely have many lease agreements and are likely required to report under IFRS 16 or Local Generally Accepted Accounting Principles (GAAP).

  • I want to prioritize organizations that are professional sports team. Excluded are amateur or recreational clubs or teams. Further excluded are operators of sports arenas and similar.

  • I'm looking for companies that are professional sports teams, but not amateur clubs, recreational teams, or venue operators.

  • As metal parts seller, I am looking for companies which are likely to purchase metal parts in bulk.

  • I am looking for companies which are very likely to operate a motorized fleet (trucks, cars, ...). The company shall very likely operate multiple vehicles. Examples are transportation, logistics, maintenance in house associations, home cleaning, delivery or pickups.

  • I'm looking for companies that operate a fleet of motorized vehicles, typically used for logistics, service, or delivery operations. These companies should manage multiple vehicles as part of their core business.

  • I want to prioritize companies that are travel agencies or tour operators, with a strong likelihood of offering international tours.

  • I'm looking for companies that are travel agencies or tour operators, and which are very likely to offer international tours.

  • I want to prioritize companies that are manufacturers of fire protection systems and fire safety equipment, explicitly exclude resellers or distributors, the company must be a direct manufacturer.

  • I am looking for companies which satisfy all of the following.

    • They offer a (commercial) venue for smaller live events, such as concerts and standup or similar

    • There is a bistro or cafeteria or bar where small snacks are sold, like a sandwich, cakes, etc.

  • I want to prioritize companies which satisfy the following criteria:

    • They are primarily either a fitness center OR a sports center OR a bouldering gym OR a climbing gym.

    • There is a clear indication for a permanent, commercial bistro or cafeteria or bar where small snacks are sold, like a sandwich, cakes, etc. It is not enough that food is sold only to special external events or similar.

✅ Good prompt examples for Text output

  • What is the primary industry sector this company operates in?

  • Classify this company as either: 'Enterprise', 'Mid-Market', 'Small Business', or 'Startup'.

  • What is the company's main product or service offering?

  • Identify the company's target customer segment.

✅ Good prompt examples for Number output

  • Estimate the approximate number of employees at this company.

  • Estimate the number of office locations this company has.

  • Estimate how many years the company has been in business.

✅ Good prompt examples for Yes/No output

  • Does this company operate in the healthcare industry?

  • Is this company B2B focused?

  • Does this company have international operations?

  • Is this company venture-backed?

🚫 Avoid

  • Vague prompts like "Good companies."

  • A prompt that doesn't match the output type (for example, asking for a score but choosing Yes/No).

4. Choose what to generate

In the Generate dropdown, pick the output type:

  • Score: returns a value between 0 and 10. Use it when the answer isn't strictly yes or no and you want to see how strongly a company matches your criteria.

  • Text: returns a short written answer. Use it for explanations, summaries, classifications, or short insights.

  • Yes/No: returns Yes or No. Use it when your criteria are clearly true or false.

  • Number: returns a numeric value. Use it for counts, estimates, ranges, or other measurable values.

5. Preview results

  1. Click Generate preview. Previews don't use credits, and you can run up to 25 previews per account per day.

  2. Check the sample results. Each result shows the AI's answer with a short explanation.

  3. To adjust, change your prompt or output type and click Re-generate preview.


​

6. Choose where to save the results

Under Save results to, open Custom field and create a new field or select an existing one. The results show as a column in your list. Give the field a clear name that says what it contains.
​

Examples of good field names:

  • "Fit Score" (for Score output)

  • "Industry Category" (for Text output)

  • "Employee Count Estimate" (for Number output)

  • "Is B2B" (for Yes/No output)

7. Choose which companies to enrich and run

Under Run this enrichment on, pick one option:

  • Selected companies: only the companies you selected in the list.

  • All companies: every company currently in the list.

  • First 10 companies: a cheaper first pass. Change the number to enrich more or fewer companies, and run the rest later.

In Lists, you can enrich the whole list in one run. If you set a number in First 10 companies, you can enrich up to 50,000 companies. When you enrich from Companies, you can select up to 50,000 companies.
​

Each option shows its credit estimate. The total for the run shows at the bottom left of the panel.
​

  1. Click Save and run.

  2. Track progress in the Notification Center.

  3. Refresh the list to see the results.

Tip: Run a new prompt on the first 10 companies before you enrich the whole list.
​

8. View results

  • After you refresh, your enrichment column shows the results and explanations in your table.

  • Results display according to the output type you selected.

9. Sort, filter, or adjust

  • Open the three-dot menu on your enrichment column and choose Sort ascending or Sort descending.

  • Filter the list by your enrichment field.

  • Use Edit AI enrichment to change your prompt, output type, or both. Each run uses credits, including re-runs.

  • Use Run AI enrichment for all companies to enrich every company again, or Run AI enrichment for companies without a value to enrich only companies that don’t have a result yet, such as new additions.

  • Use Rename field to give the column a new name.

  • Use Remove this column to delete a column you don’t need.

Manage columns

  • Use column settings to add or remove fields as needed.

  • Column changes apply per list.

Additional notes

  • Output fields are saved at account level, so you can reuse them across lists.

  • Deleting an output field removes it from your entire account.

  • You can run AI enrichment several times with different prompts and output types to create several enrichment fields.

  • You can export your list to a file, and the export includes your AI enrichment results.

  • Different output types suit different purposes:

    • Use Score for prioritization and ranking

    • Use Text for categorization and classification

    • Use Number for quantitative analysis

    • Use Yes/No for quick qualification

Best practices

✔ Do this

💡 Why it works

Be specific

AI needs clear criteria

Filter your list first

Combine firmographics with behavior

Write naturally

No special syntax needed

Keep it concise

Better accuracy

Match prompt to output type

Ensures you get the right format

Test a new prompt on the first 10 companies

Each run costs credits

Use Score for prioritization

Best for ranking companies

Use Yes/No for qualification

Best for simple yes/no criteria

Use Text for categorization

Best for classification tasks

Use Number for estimates

Best for quantitative values

❌ Don't do this

🚫 Why it fails

Expect perfect results

Results are directional, not absolute

Run huge lists and expect instant results

Large jobs take time

Use the wrong output type

Mismatched output types give poor results

Who is AI enrichment for?

  • SDRs & AEs: research, enrich, and prioritize large lead lists quickly.

  • Marketers: segment audiences for better campaigns with various data points.

  • Partnership teams: find and qualify companies that match partner criteria.

  • Operations teams: extract structured data to enrich CRM records.
    ​

--

Questions, comments, feedback? Please let us know by contacting our support team via the chat or by sending us an email at support@leadfeeder.com.

Did this answer your question?