Email and Social Extractor: 3 Plays to Get Noticed Before You Pitch

An email and social extractor turns a bare domain into every reachable channel a company has, without leaving Google Sheets. Contact data, phone numbers and up to twelve social profiles, pulled from the website you already have in column A.

Most prospecting files stall at the same place. You have a list of domains. You do not know who to contact, and you do not know where to reach them. So you open twenty tabs, scroll through footers, guess at LinkedIn URLs, and paste half of it into a CRM that was already half empty.

This playbook shows the workflow the other way round: start from the domain, let Derrick pull the channels, and end with a file that is ready for a multichannel sequence. Three plays, one feature, and the cost of each step written down so you can decide before you burn credits.

What an email and social extractor actually returns

The feature is called Email & Social Extractor from Website. You give it a website. It reads the pages that carry contact information and returns, in the same row of your sheet:

• the generic and published email addresses it finds on the site
• the phone numbers published on the site
• up to twelve social links: LinkedIn, X, Instagram, Facebook, TikTok, YouTube, Telegram, Pinterest and others

It costs 2 credits per line and it is a paid feature, billed per request rather than per result. That matters for planning: a row that returns nothing still costs you the call. Run it on a cleaned list of domains, not on a raw export with duplicates and dead sites.

Two things it is not. It is not an Email Finder, which resolves a named person at a company and charges 5 credits only when it finds the address. And it is not a full crawler: if you want to read what a site says about itself, page by page, that is the Website Crawler. The extractor has one job, which is finding the channels.

Play 1: know who you are talking to before the call

The problem is familiar. You have a call booked and all you have is a domain. You spend the first four minutes of the meeting working out whether they are an agency, a marketplace or a product company, and the prospect can hear it.

The workflow

1. Put your list of websites in a Google Sheet, one per row.
2. Run Email & Social Extractor from Website. Among the social links it returns, you get the company's official LinkedIn URL, which is the one thing you cannot reliably guess from a domain.
3. Feed that LinkedIn URL into Enrich Companies to pull the firmographics.

What lands in your sheet

• the real company name, which is often not the domain
• the tagline and description the company uses about itself
• industry and company type
• headcount, as a number rather than a bracket
• country and headquarters

Enrich Companies costs 1 credit per company and is available on the free plan. It reads the company's LinkedIn page, so it needs your LinkedIn session connected through the Derrick Chrome extension. That is worth saying plainly, because it is the step most people skip and then wonder why their column is empty.

Why it works

Knowing the sector, the size and the self-description changes the first sentence of your call. A twelve person agency and a four hundred person software company do not have the same problem, and they do not respond to the same opening. You are not personalising for the sake of it. You are removing the part of the conversation where you ask questions the internet could have answered.

Play 2: fill your CRM with every channel that is actually reachable

Most CRMs hold a domain and, if you are lucky, one email address. Nobody knows whether the account is active on LinkedIn, whether they run a YouTube channel, or whether their community lives on Telegram. So follow-ups default to email, and email alone.

The workflow

1. Put the account domains in a sheet connected to Derrick.
2. Run Email & Social Extractor from Website on the column.
3. Push the result into your CRM through your usual integration, an export, or the Derrick API.
4. Optionally, build a single field that summarises which channels exist, with tags such as linkedin, youtube or telegram.

That last field is the one your reps will actually use. It answers a question they ask every day, which is not "what is their LinkedIn URL" but "where can I reach these people at all".

Before you run it on ten thousand rows

Two credits per line adds up, and the feature bills per request. Three habits keep the cost honest:

• Deduplicate on the domain first. Three subsidiaries of the same group often share one website. Find Duplicates is free and unlimited on every plan.
• Filter to the accounts a rep is actually going to work. Enriching a segment nobody will call is the most common way to waste a monthly allowance.
• Verify what you extract before you send. An address published on a website can be years old. Email Verification costs 1 credit per address and is cheaper than a bounce.

Play 3: warm up strategic accounts before you pitch them

Cold outreach into a named account has a structural problem. The first time your name appears in front of the buyer, it is attached to a request. Nothing before it, no context, no recognition.

The social links you just extracted are the cheapest fix available, because they tell you where that company is already publishing.

The workflow

1. Take the enriched file from play 2 and filter to accounts that have three or more active channels. Those are the companies with someone whose job includes being visible.
2. Assign a rep to engage lightly for a few days: read what they publish, comment where you have something useful to add, subscribe where subscribing is public.
3. Log those touches in a simple pre-warm column so nobody does it twice.
4. Start your sequence after your name has appeared in their feed, not before.

Why it works

You are not buying attention, you are borrowing familiarity. A message from a name the prospect has seen twice reads as a follow-up. The same message from a stranger reads as an interruption. The extractor is what makes this scalable: without it, finding the channels for two hundred accounts is a week of manual work, which is exactly why most teams never do it.

One caveat worth stating. This play degrades badly if it is automated into noise. Commenting on a hundred posts a day with generic praise is visible, and it is visible to the buyer first. Keep the account list small enough that a human is genuinely reading.

What it costs, row by row

The three plays use different combinations of features, and the credit cost is not the same. Here is what each one bills for a single company, so you can size a batch before you launch it.

Play 1 costs 3 credits per company. Play 2 costs 2 credits per company, or 3 if you verify what comes back. On the free plan, 100 credits per month is roughly 33 companies fully qualified, which is enough to test the workflow on a real segment before committing to anything. Paid plans start at 9 euros per month, and the per credit price falls as the volume rises.

The number that surprises people is the difference between per request and per result billing. The extractor charges for the attempt. Email Finder charges only on success. That is why the sequencing above matters: extract on a deduplicated list, then spend the expensive credits only on the rows that came back with something.

Where it breaks, and what to do about it

No extractor returns a full row every time. Four failure patterns account for most of the empty cells, and three of them are fixable from your side.

The site publishes nothing

Plenty of companies deliberately publish no address and no phone number, and route everything through a form. The extractor will still return the social links, which is usually the useful half. Treat the empty email column as information: this account is reachable on LinkedIn, not by email, and your sequence should reflect that rather than fight it.

You get a generic address, not a person

What sits in a website footer is almost always a role address: contact, hello, sales. That is fine for a small company where the founder reads it, and close to useless in a four hundred person organisation. Rule of thumb: below roughly twenty employees, the generic address is a real channel. Above that, use it to confirm the domain pattern and go find the named person instead.

The domain is wrong before you start

Redirected domains, parked domains and country variants all produce empty rows that look like a feature failure. They are a list problem. Deduplicating and normalising the domain column costs nothing and removes most of it.

The company changed and the site did not

A published address survives the person who owned it. This is the one you cannot detect by looking at the row, which is the entire argument for verifying before you send rather than after your bounce rate tells you.

Run the same workflow outside Google Sheets

The three plays above describe the Derrick sidebar in Google Sheets, which is where most teams start because the file already exists there. The same enrichment is available on two other surfaces, and the right one depends on where the work happens.

• Google Sheets sidebar. Best when you have a list in front of you and you want to see the columns fill in. This is the surface every play above uses.
• Derrick MCP. Best when the question starts in a chat. Connect Derrick to Claude or ChatGPT and ask for the channels of a company inside the conversation, without building a sheet first. Available from the Standard plan.
• REST API. Best when the enrichment should happen without anyone opening anything: a new account created in the CRM triggers the extraction and the fields are populated before a rep ever sees the record.

The credit cost is the same on all three. What changes is who does the triggering.

What this playbook unlocks

• You know who you are addressing: LinkedIn URL from the extractor, firmographics from Enrich Companies.
• You know where to reach them: emails, phone numbers and up to twelve social channels in the same row.
• You show up before you ask: a short warm-up pass on the accounts that publish.

None of this requires a new tool in the stack or a line of code. It requires the domain you already have and a sheet you already opened. If your outbound is currently one channel wide because nobody had time to find the others, this is the shortest path out of it.

Related playbooks in this series: reading a prospect's website at scale, detecting the technologies a company runs, scoring accounts on what their users say about their tools, and finding phone numbers in bulk.

One more thing. We send one email every 2 weeks with the prospecting workflows we actually run, credit costs included. Subscribe to the Derrick newsletter if you want the next playbook before it is published here.