How to Set Up Apollo.io to Import, Enrich and Score Leads

Coding Liquids tutorial cover featuring Sagnik Bhattacharya for How to Set Up Apollo.io to Import, Enrich and Score Leads.
Coding Liquids tutorial cover featuring Sagnik Bhattacharya for How to Set Up Apollo.io to Import, Enrich and Score Leads.

To set up Apollo.io for your first 500 leads, define your buyer as filters, build a CSV with names plus a company website, email or LinkedIn address, import it with enrichment switched on, create a score for fit, then push the best-rated rows to your CRM. Check the credit estimate before you click Import.

The Complete Excel Course with AI Integration

From formulas and pivot tables to Power Query, DAX, VBA and AI with Copilot, ChatGPT and Claude — hands-on from the first lesson.

By Sagnik Bhattacharya and Virginia Thorn, Primetime Emmy Award winner

Buy now

Two details catch most first-time users. CSV enrichment on import is a paid-plan feature, and credits are spent on email and phone reveals, so a list of 500 can cost anywhere from 500 credits to several thousand depending on whether you ask for mobile numbers. The sums come first below, because they decide which plan you need.

How many credits does a 500-lead Apollo import cost?

A 500-row list costs about 500 credits if you enrich emails only and far more if you add mobile numbers, because Apollo's pricing page states 1 credit for an email and 8 credits for a phone number.

Apollo's pricing page lists three paid plans, billed annually per seat: Basic at $49 a month with 30,000 credits a year, Professional at $79 with 48,000 and Organization at $119 with 72,000 (minimum three seats). The free plan gives 900 credits a year. The same page gives the headline credit costs: an email is 1 credit, a phone number is 8 credits, and data enrichment is 1 to 8 credits depending on what you ask for.

Here is what that means for a 500-row list. These are illustrative sums using the page's stated costs; the estimate Apollo shows on the import screen is the figure to trust on the day.

What you ask Apollo to findCredits for 500 rowsShare of Basic's yearly 30,000
Emails only500About 1.7%
Emails plus mobile numbers for every row4,50015%
Emails for all, mobile for the best 100 only1,300About 4.3%

The third row is the sensible default. Mobile numbers cost eight times what an email does, and a first outreach round rarely needs them. Enrich emails for everyone, score the list, and only then spend phone credits on the top slice.

Credits are pooled across the team, granted at the start of each billing cycle, and expire at the end of the cycle without rolling over. Admins can set per-user credit caps, which is worth doing before anyone else on the team gets login details. If the cost per lead is your main question, the Apollo plan breakdown goes through each tier; the free plan and its 900 credits has its own page.

How do you turn your ideal buyer into Apollo search filters?

Write your buyer as one plain sentence, then turn it into filters, as a six-person architecture practice that wants more work from residential property developers would. The owner does not need 500 random contacts. They need 500 people who commission new-build or conversion schemes and are the person who signs off the architect's fee. That sentence turns into filters you can type into Apollo's People search:

  • Job titles: Development Director, Head of Development, Land and Planning Director, Managing Director, Founder.
  • Company size: 11 to 200 employees. Below 11, there is rarely a development function; above 200, an architect is usually picked from a framework list.
  • Industry keywords: residential development, build-to-rent, conversions, housebuilder.
  • Exclusions: your existing clients and your competitors' company names, added as a list to exclude.

Write the buyer in one line before you filter, because it stops the list drifting: "Directors at residential developers with 11 to 200 staff who appoint architects." If a row fails that line when you read it, delete it before it costs a credit.

The same exercise for a four-fee-earner law firm selling to small landlords gives different filters, and a marketing agency selling to e-commerce brands gives different ones again. The pattern holds: title, size, industry words and an exclusion list. Apollo's help says its AI auto-scores draw on your company website, account information and prospecting activity, so a clear company website and a focused search history give it more to work with.

What should an Apollo CSV contain so every row matches?

You may already have a list: a trade-show export, an old CRM extract, a spreadsheet of past enquiries. Apollo's CSV import help sets out the rules. An Apollo CSV needs at least one of four columns: Company Name, Company Website, Contact LinkedIn URL or Contact Email. Apollo recommends adding First Name, Last Name, Contact Email, Contact LinkedIn URL, Company Name and Company Website, because more data makes it easier for Apollo to recognise a contact. Enrichment may not work when Apollo does not have enough to find your people.

Here is a filled-in sample of three rows from the architecture practice's file (illustrative, with "[first name]"-style placeholders instead of real people):

First Name,Last Name,Contact Email,Contact LinkedIn URL,Company Name,Company Website,Title
[first name],[surname],,https://www.linkedin.com/in/example-1,Example Homes Ltd,examplehomes.com,Development Director
[first name],[surname],name@example-build.com,,Example Build Group,example-build.com,Managing Director
[first name],[surname],,,Example Conversions,exampleconversions.com,Head of Land

Three details from Apollo's help page save a failed import:

  1. Websites as bare domains. Enter the domain only, without www. or any prefix. A column full of https://www.example.com is the most common reason account matching fails.
  2. Real names on every row. The page warns that "Most rows need a valid full name" appears when the file fails validation. Delete rows that hold only a company or a generic inbox such as info@.
  3. Quotes closed properly. "Illegal quoting in line 1" means a mismatched quote mark; fix it and save as CSV again.

Apollo says to avoid files over 10,000 rows and to split larger ones. At 500 rows you are nowhere near that, and the weekly import limit table on the same page starts at 100,000 rows for teams spending $1,000 or more a year, so a first list will not touch it. Delimiters can be comma, semicolon, colon or pipe; if your spreadsheet software uses semicolons, Apollo reads those too.

A before and after of a messy list

The practice's raw export had 612 rows from three sources. Cleaning it before import looked like this:

Problem foundRowsFix
Website typed as https://www... 240Find and replace "https://www." and "https://" with nothing; trim trailing slashes
Generic inbox only (info@, hello@)41Delete; Apollo cannot match a person from them
No name, company only33Delete, or search those companies in Apollo later
Same person twice from two sources38Sort by email, keep the row with more columns filled

That leaves 500 clean rows from 612 (612 minus 41, 33 and 38). Cleaning on a spreadsheet costs an hour; skipping it costs credits spent on rows that cannot match. If you want to automate bulk tidying, cleaning messy data with AI covers the sheet side.

If the spreadsheet side is the slow part, try The Complete Excel Course with AI Integration on Coding Liquids.

How do you import a CSV into Apollo?

To import a CSV into Apollo, go to People, then Import, then CSV, map each header to an Apollo field, set the import options and choose email or phone enrichment before starting.

You need permission to bulk-import contacts by CSV; if the button is missing, ask your Apollo admin to change your permission profile. Then, per Apollo's help:

  1. Click People, then Import, then CSV.
  2. Choose Select CSV File and pick your file.
  3. Map each CSV header to an Apollo field. If you need a field Apollo lacks, create a custom contact field first; Apollo warns you must restart the upload after creating one.
  4. Set the import options listed below.
  5. Open the Data enrichment tab and choose email, phone or both.
  6. Read the credit estimate, then click Import.

The options in step 4 are the ones owners get wrong, so here are the choices I would make for a first list:

SettingChoice for a first 500Why
StageA "Cold" or "New" stageStages decide how people are treated once they reach a sequence
If contacts already existUpdate the existing recordApollo checks first name, last name, company, email and LinkedIn URL, and this adds your data rather than creating duplicates
If owner is missingYourselfUnowned contacts are easy to lose track of
Add to a listA new list named with the date, such as "Developers import 17 Nov"Lets you filter the 500 later and keeps them apart from other prospecting
Auto-assign accountsOnCreates company records from the domain column
Create and update records in CRMOff for nowIt overrides the integration's push settings; push after you have scored

Apollo says most imports finish in a few minutes but large CSVs can take hours. Check progress under Settings, then Data management, then Imports and exports, and open Contact Import. After it finishes you can filter by the "contact CSV import" filter to see only these 500.

Should you enrich emails first and phone numbers later?

Enrich emails first and add phone numbers only for the top-scored slice, because a phone number costs 8 credits against 1 for an email. Run multi-source only on rows that came back with no email.

In the Data enrichment tab you toggle email, phone or both. A separate option, Enrich with Multi-source, uses waterfall enrichment: Apollo checks several third-party data sources in order until it finds the data. Admins control the lineup under Waterfall Settings. Apollo's credit-usage help lists waterfall enrichment as its own feature, with email and mobile on separate credit balances, so it can show up as a separate line when you review usage later.

For the architecture practice, this is a sensible order:

  1. Import all 500 with email enrichment on and multi-source off. Check how many rows come back with an email.
  2. For rows with no email, run a second pass with multi-source on, but only on that subset, so you pay waterfall rates for the gap rather than the whole list.
  3. Leave phone off until scoring has picked the best 100.

Apollo's import screen shows a credit estimate before you confirm, so use it to judge whether a second enrichment pass on the same rows is worth paying for, and assume re-enriching a row for a missing email or phone can cost credits again unless the estimate says otherwise. Do not click enrich twice out of impatience.

Checking what the credits bought

Open Settings, then Credits and activity, then Credit usage. The Overview tab groups spend by feature, the Usage details tab lists each event, and there is a Show refunds toggle. After the architecture practice's first pass, the check was simple: 500 rows should show about 500 credits under email, nothing under phone, and a smaller waterfall line for the retry. If phone credits appear, someone ticked the phone box.

How does Apollo score leads, and how do you create a custom score?

Apollo rates each person or company Excellent, Good, Fair or Not a fit by scoring them against your ideal customer profile. Auto-scores come from your company website, account information and prospecting activity. Per Apollo's scoring help, AI auto-scoring is available on paid plans, and you can create up to two custom scores on free and Basic plans (unlimited on Professional and Organization).

To create a custom score as an admin:

  1. Go to Settings, then Ideal customer profile, then Scoring, and click Create new score.
  2. Choose to build it with the AI assistant or manually, and choose people or companies.
  3. If you use the assistant, describe the prospects in plain words. Apollo's help recommends configuring the AI context centre first so the assistant understands your business.
  4. Review the generated criteria and adjust the filters or weighting before saving.
  5. Set it as your primary score. Apollo requires one primary people score and one primary company score once you create custom ones, and these replace auto-scores as the default.

A description the architecture practice might give the assistant (illustrative):

Create a score for residential developers with 11-200 employees.
Weight job titles of Development Director, Head of Development and
Managing Director highest. Favour companies with build-to-rent or
conversion schemes. Penalise housebuilders with 500+ employees and
anyone whose title is Land Buyer or Quantity Surveyor.

What you should expect back is a list of criteria with weights, not a verdict. Read each one. If the assistant gave company size a weight bigger than job title, move it back, because a Managing Director at a 12-person developer is a better lead for you than a junior planner at a 150-person one. Apollo shows a score column in search results, lets you hover for the reason, and offers a Show breakdown button on a profile that lists which criteria added points.

Reading the ratings against your 500

Filter the imported list by score and look at the split. A healthy first list for a sharp filter set might be 80 Excellent, 190 Good, 150 Fair and 80 Not a fit (illustrative). The rule I would apply:

  • Excellent: enrich phones, personal first-line research, first to receive outreach.
  • Good: email sequence only.
  • Fair: hold; recheck after you add more signals.
  • Not a fit: remove from the list, and read three of them to learn why the filter let them through.

If almost everything lands in Good, the criteria are too loose. If almost everything lands in Not a fit, the score is stricter than your list, or the list was bought for the wrong buyer. Both are information you can use before a single email goes out.

Why do many rows show Not a fit after scoring?

Scoring before the data is complete can mark good rows Not a fit, because the score penalises missing data rather than a poor match. Picture a professional services owner who imports 500 rows, enriches emails and scores the list immediately (an illustration, not a client story). Eighty rows show Not a fit. On inspection, 60 of them have no job title because the CSV never carried one and Apollo has not found it.

The fix takes ten minutes: filter to rows with no title, run enrichment again on those only, and re-read the ratings. The general rule is to check completeness (title, company size, industry) with a quick filter before trusting the score. If more than a fifth of a rating group is blank in key fields, repair the data first.

How do you test an Apollo list before anything is sent?

Test the list by checking 25 rows from Excellent and 25 from Good before anything is sent. Look at each record for three things: the email looks plausible for that company's domain, the title matches what the person does, and the company still operates. Delete anything wrong and note the pattern. If six of 25 are wrong in the same way, fix the cause for the other 450.

Do not send from a new domain straight away. The domain authentication tutorial covers the sender setup that has to be in place before sending; it is a separate piece of work from list building, and a list sitting ready for two weeks costs nothing. Because these are business contacts you did not collect yourself, read how Apollo handles data sources and opt-outs and keep your own suppression list, so that anyone who asks not to be contacted stays out of later imports.

How do you move the best Apollo rows into your CRM?

For a first list, push only the Excellent and Good rows into your CRM and tag them with the list name, leaving Fair and Not a fit inside Apollo.

You have two routes. If you connected a CRM such as Salesforce, HubSpot or Pipedrive, the import screen has a Create and update records in CRM tick-box. Apollo's help pages I read do not say whether scores travel into your CRM, so after a small test import, open one record in your CRM and see whether a score appears before you rely on it. Or you can export the list. Apollo's help pages disagree on whether exporting saved contacts costs export credits, and its credit-usage page says export credits may apply only to certain plans or legacy configurations, so open the export screen and read what it says before you confirm.

My suggestion for a first list is to push only Excellent and Good rows, tag them with the list name, and leave Fair and Not a fit inside Apollo. That keeps the CRM from filling up with 500 contacts nobody has spoken to. Add a CRM field such as "Source = Apollo import, 17 Nov" so you can report on the batch later. If you would rather automate the transfer, the Zapier route to a CRM does that with duplicate checks, and the Apollo versus HubSpot comparison helps if you have not picked a CRM.

If you want help fitting this into how your sales process actually runs, I offer one-to-one AI consultations for small businesses.

A seven-point checklist for the day of the import

  1. One-line buyer description written and filters saved.
  2. CSV has names plus at least one matching column, domains without www., no generic inboxes.
  3. Plan confirmed as paid if you want enrichment on import.
  4. Credit estimate read; phone enrichment off unless budgeted.
  5. Duplicate rule set to update the existing record; list named with the date.
  6. Score created, set as primary, and checked for blank-data rows.
  7. 50-row spot check done before any push to the CRM or any sequence.

A 500-row list done this way takes an afternoon, spends a known number of credits, and leaves you with a ranked list rather than a pile of addresses. If you want to compare it with a lighter approach for a tiny business, the startup prospecting setup covers the day-to-day search habit rather than a one-off import.

Questions owners ask before the first import

Can I import and enrich on Apollo's free plan?

You can import a CSV on the free plan, but Apollo's CSV help page says enrichment during import is for paid plans only. The free plan includes 900 credits a year, so for a 500-lead list, treat the free plan as a place to test the import and the score, not to enrich the whole file.

Does importing a CSV use credits?

Apollo's import screen shows a credit estimate based on your row count and enrichment settings, and its help page says enriching costs credits. With enrichment switched off, check that the estimate reads zero before you click Import. Do not assume; the estimate on the screen is the figure that counts.

What if Apollo cannot match some of my rows?

Apollo's help says enrichment may not work when it has too little data to recognise a contact. Add Contact Email, Contact LinkedIn URL, Company Name and Company Website where you have them, then re-import the unmatched rows as a small second file rather than re-running the whole list.

How many custom scores can I create?

Apollo's scoring help says free and Basic plans can create up to two custom scores, while Professional and Organization plans have unlimited custom scores. AI auto-scoring is listed for paid plans only. One people score and one company score is enough for a first 500 leads.

How many credits does a 500-lead Apollo import use?

It depends on what you ask Apollo to find. On this tutorial's illustrative sums, emails only for 500 rows is 500 credits, while emails plus mobile numbers for every row is 4,500. Apollo's import screen shows a credit estimate before you click Import, and that estimate is the figure to trust on the day.

How do I import leads into Apollo from a CSV?

Prepare a CSV with at least one of Company Name, Company Website, Contact LinkedIn URL or Contact Email, then go to People, Import, CSV, pick the file and map each header to an Apollo field. Set the import options, choose email or phone enrichment in the Data enrichment tab and read the credit estimate before you click Import.

Further reads

Sources: Apollo pricing page and help-centre articles on CSV import, credit usage and scores, read on 17 Nov 2026.

Have a specific process in mind?

Describe the job in your own words on a 1:1 call and I'll show you what AI you can use for it with the tools you already pay for.

Book a 1:1 call with me