Enrich Lead Lists in Google Sheets

You run outbound for a small software team, and 300 fresh leads land in your sheet every Monday. Each row holds a name, a company, and a website. Nothing else. Looking all 300 up by hand costs about 6 hours a week, so most rows never get a real first line.

Paid enrichment tools fix that if you have the budget and someone to own the integration. Formulas fix it if you have neither. Here is the first one.

What comes back

Now drag it down 300 rows and go do something else. One prompt, 300 lookups, a title column you never typed. That is the pattern for the rest of this post: one column at a time until the bare export is a research sheet. New to the add-on? The getting started guide takes five minutes.

Quick verdict

What the paid tools do well

They hand you a database before you have a single name. Apollo's homepage puts its own at 240M+ contacts and 30M+ companies, with filtering by job title, seniority, company size, industry, and location. If your target list does not exist yet, that is worth paying for.

They plug into the CRM. Apollo lists integrations with Salesforce and HubSpot. ZoomInfo describes exporting straight into your CRM. Nobody has to babysit a spreadsheet.

They also send. Apollo includes built-in email sequencing and multi-touch campaigns, so research and outreach live in one place. Clearbit went further and joined HubSpot outright, and it does the normalized-role, corporate-hierarchy, and IP-intelligence work inside that platform.

Apollo says over 600,000 companies use it. ZoomInfo's about page says 30,000+ customers. These are real products doing a real job.

Where they fall short

The cost tracks the record count. Apollo meters usage in credits with a monthly cap, and its own page describes a 10,000 credit monthly limit on non-paying Unlimited accounts. Your allowance runs dry the week the conference list arrives.

The fields are whatever the vendor stores. Your playbook might need "does their careers page mention Python" or "what did they announce last quarter". A database has no column for a question you invented on Tuesday.

Prices are often not on the page at all. ZoomInfo's pricing page publishes no numbers and asks you to request a demo. Clearbit's site publishes none either. Small teams tend to want a number before a call.

And the freshness is the vendor's refresh cycle, not today. A formula that reads the company's homepage right now has no refresh cycle to wait for.

How SheetMagic handles the same job

Same sheet, one column per question. Every column below is a formula you drag down.

Read the company's own homepage

The website column is the most underused field in any lead export. Pull the page in, then ask questions of the text.

What comes back

Then two AI columns read that text instead of the whole internet.

What comes back

Swap the prompt and you get another column from the same text. "Name any software tools mentioned on this page" gives you a rough tech-stack signal for free, because the page is already in column G.

Pull the LinkedIn company facts

When your export has company LinkedIn URLs, one formula covers the whole list.

What comes back

Put this on its own tab. It spills a wide table, so it needs empty space to the right and below. Then bring the two fields you care about back to the main sheet with a plain VLOOKUP on company name.

When you only have the domain

Sometimes the row has a company and no person. Search by role instead of by name.

What comes back

That covers the gap paid tools are hired for. You are searching by title and domain, and the result lands in the sheet rather than in an export queue.

Add an intent signal

A title is not a reason to email. News and hiring are.

What comes back

Hiring is the other tell. A company posting three sales roles is building a team, and that is a better opening line than a job title.

What comes back

Back on the main sheet, =COUNTIF(Jobs!$B:$B, B2) turns that tab into a single hiring-signal number per row.

Check the emails before anyone sends

Bad addresses cost you domain reputation, which is more expensive than the research.

What comes back

Drop the undeliverable rows, route the role-based ones to a different sequence, and send to the rest.

Score the row before you trust it

Enrichment is only useful if you know which rows are thin. Native Sheets handles this.

What comes back

Sort by that column and your best 80 rows are at the top. Those get a hand-checked first line. The 3-of-9 rows go back in the pile.

Side by side on one real workflow

The task: 300 event leads, each with a name, a company, and a website. You want title, what they sell, employee count, a news hook, and a validated email.

By hand. Search the person. Open LinkedIn. Open the website. Skim it. Search the company name plus "funding". Guess the email pattern. Type six cells. Repeat 299 times. Call it 70 seconds a row on a good day, which is about 6 hours.

With a paid database. Upload the list. Match records. Choose fields. Spend export credits. Download. Paste into the sheet. Fast, and limited to fields the vendor stores. Any question outside those fields still gets researched by hand.

With formulas. Paste seven formulas into row 2. Drag down. Run three background scrapers on their own tabs. Wait. That is about 20 minutes of setup and a wait you can spend on calls. The 8th question you think of on Thursday is a new prompt, not a new tool.

When to still use a paid enrichment tool

Buy the database when the list does not exist yet. If your job starts with "find me 5,000 heads of ops at logistics companies with over 200 staff", that is a database query. Apollo's filters by title, seniority, size, industry, and location exist for that, and a spreadsheet is the wrong shape for it.

Buy it when the CRM is the system of record and sales ops needs enrichment to happen without anyone opening a sheet. That is an integration job.

And buy it when you need one tool for research plus sequencing plus dialing. Apollo bundles sending. SheetMagic does not send email.

Formulas win when the list already exists, when the fields you need keep changing, and when you would rather pay for a month than for every record.

Copy this template

Fourteen columns. Four you paste in, nine that fill themselves, one that scores the row.

Three helper tabs carry the wide tables. LinkedIn!A1 holds =LINKEDINCOMPANY(Leads!D2:D301). Jobs!A1 holds your JOBSEARCH call. FindLeads!A1 holds your FINDLEADS call. All three spill, so leave those tabs empty.

Columns J through M read from those tabs, so they show "add LinkedIn tab" or "add Jobs tab" until the tabs exist. Columns E through I and N work the moment you import. Run VALIDATEEMAIL on column M once it fills.

Download the template CSV. Import it into a blank sheet (File, Import, Upload) and the formulas are ready to drag down.

Writing the actual email is the next column over, and prompt templates has patterns for it. If you want the persona-level view of this whole workflow, the lead generation page walks through it.

What to watch for

• Background formulas look broken for about a minute. VISIT, LINKEDINCOMPANY, FINDLEADS, JOBSEARCH, and VALIDATEEMAIL all show a loading placeholder while a worker does the job. It is not frozen. Leave it alone.
• Spilled tables need empty space. A scraper that returns 50 rows and 12 columns needs those cells free, or Sheets shows #REF!. That is why the wide scrapers go on their own tabs.
• Web search only works on some models. The webSearch=TRUE argument in AITEXT needs an OpenAI GPT-4o or newer model, a current Claude model, or a Perplexity Sonar model. Gemini and Mistral will ignore it.
• Everything scraped is public. These formulas read pages anyone can open. No logged-in data, no private profiles, and no guaranteed direct dials.
• Keep Formula freezes results. Once a result lands it becomes static text with the formula kept in the cell, so a recalculation does not re-run and re-bill 300 rows. That is the setting you want on for work like this.

Try it on 20 rows first

Take the top 20 rows of last week's list and run columns E through I. That is four formulas and a few minutes, and it tells you whether your prompts are asking the right questions before you spend a credit on all 300.

Then add the three helper tabs and let the rest of the list fill in while you make calls.