How to send cold emails from your own Gmail or Outlook address

Yes. When you connect your Gmail or Outlook account, an AI agent can send email from your address the same way you would, so each message sits in your Sent folder and each reply comes back to your inbox. That's the main reason people want it. A cold email from a real person's mailbox looks like a real conversation, and the recipient can answer you directly instead of a no-reply address or a separate sending tool.

You can also do this without an agent. Gmail and Outlook both support mail merge, which fills a template with names and other details from a spreadsheet and sends one copy to each person. Google Workspace has a built-in mail merge in Gmail on some plans, and Outlook can send a mail merge started from Microsoft Word. Add-ons such as GMass, Mailmeteor, and YAMM run the same idea inside Gmail and add scheduled follow-ups. Sending tools such as Instantly and Smartlead connect to your mailbox too, and they rotate sends across several mailboxes when you need volume.

All of these routes send well, but none of them decides who should get an email or what it should say. A mail merge sends your template to every row in your sheet. That includes the person who changed jobs last quarter, the company that's already your customer, and the address that bounces. Each recipient gets the same paragraph with their first name swapped in, and most people can tell.

Your mailbox also has limits worth respecting. Google allows about 2,000 messages a day from a Google Workspace account and 500 from a free Gmail account, and a Microsoft 365 mailbox allows up to 10,000 recipients a day. Cold email works far below those numbers. Your address carries your domain's reputation, so a run of bounces or spam complaints can push your ordinary email into spam folders too. Before you start, make sure your domain has SPF, DKIM, and DMARC set up, which are the records that prove to Gmail and Outlook that your mail really comes from you.

A good result is a small number of emails that each deserve a reply. Every recipient fits your customer profile, every email names a real reason to write to that person now, the volume stays small enough to protect your mailbox's reputation, and the follow-ups arrive in the same thread as the first message. Kite does that work for you, from choosing the recipients to tracking the replies.

Start by telling Kite who you want to reach and why, in the Kite web app or in Slack. You can give it a list you already have:

"Email the 30 people in this spreadsheet from my Gmail and ask for a 20-minute call about our new reporting feature."

Or you can ask it to find the people:

"Find 25 operations leads at logistics companies that raised money this year, and send each one a short intro from my Outlook address."

Kite is an AI marketer. It researches your business, builds a growth strategy, and ships the work that brings you new customers. By the time you ask, it already knows your positioning, your ideal customer profile, your current customers, and your competitors from that research. That means it can propose who to target, what reason to give them, and what to ask for without a long brief. Confirm the plan or adjust it, and Kite starts.

You'll need to connect the Gmail or Outlook account you want to send from. You sign in through Google or Microsoft and grant access, and Kite then sends from that mailbox. If you don't connect a mailbox, Kite can still send from your team's own Kite email address, which every team gets. You don't need a prospecting account of your own to find new people, because Kite has its own Apollo connection built in.

Kite asks for your approval before it sends outreach. It shows you the list and the emails, and you approve them with the Approve button or a reply. Once you trust how it writes, you can give it more room to send on its own.

Kite sends from the mailbox you connected, under the name and signature you use there. In Gmail, it can read the addresses your account is allowed to send as, so when you ask it to send from an alias such as sales@ or your own name at a second domain, it confirms the account can actually send as that address first. Every email is signed by a real person you name, usually you, with your company named plainly.

Kite decides who gets an email by giving every recipient a reason, and it drops anyone it can't justify in one sentence. When you share a list, it reads the spreadsheet or CSV file and checks each person against your customer profile. When you ask it to find people, it searches Apollo for companies by industry, headcount, location, funding, and hiring, then for people at those companies by title and seniority.

Kite then confirms that each person still holds the role at that company before it writes a word, because a wrong name or an old title ends the conversation before it starts. It scores each person's fit from one to five with a one-line reason, and when the list runs past about fifteen people, it splits it into primary and secondary groups.

Kite also removes anyone who should not hear from you. It skips your existing customers and anyone who has declined or unsubscribed, based on earlier replies, the notes in your team's knowledge base, and any suppression flag in a connected CRM. It can search your mailbox for earlier threads with each person, so it knows whether you've already been in touch.

Kite labels every address with the verification result its source returned, such as verified, catch-all, risky, or unverified, so you can see which ones are likely to bounce. It never guesses an address from a name pattern, and it uses work addresses only. If you want to protect your mailbox, tell Kite to send only to verified addresses, and it will find replacements for the people whose addresses fall short.

Kite writes one email per person, built on a fact about that person rather than a template with a name swapped in. It opens with the reason for writing in the recipient's terms, ideally a recent event at their company such as a funding round, a new hire, or a launch, with its date and source. The email keeps to one ask, stays under about 120 words, carries no more than one link, and reads as plain text, the way you would write to one person.

Kite writes in your brand's voice and edits each draft so it doesn't read as machine-written. Every cold email closes with a low-pressure line such as "If this isn't relevant, tell me and I won't follow up." If you'd rather send the emails yourself, Kite can save them to your Drafts folder in Gmail or Outlook instead, where you can edit and send each one.

Once you approve, Kite sends each email separately from your mailbox, so every recipient gets their own personalized message and nobody sees anyone else's address. It sends a first batch of three to five emails, confirms that Gmail or Outlook accepted them, and then sends the rest. If sends start failing, it stops and tells you why instead of pushing on.

For each email, Kite keeps the message and thread identifiers that Gmail or Outlook returns. Those identifiers are what let it attach later follow-ups to the right conversation. When a send result is unclear, Kite marks that person as unconfirmed and tells you, rather than sending a second copy that might duplicate the first.

Kite sends up to two follow-ups, spaced days apart, and replies within the original thread so the recipient sees the whole conversation in one place. In Gmail, it replies to the original thread. In Outlook, it replies to the original message. Each follow-up adds something new, such as a different detail about the recipient's situation or a short piece of proof, rather than repeating the first email.

Kite puts the follow-ups on a schedule when it sends the first emails, and each one checks opt-outs again before it goes out. Anyone who has replied or asked to stop is left alone. After the second follow-up, Kite treats silence as an answer.

Replies land in your own inbox, where you can answer them yourself. With Gmail or Outlook connected, Kite can also watch that inbox: every two hours it looks at new mail and sends your team's owner one message about the replies that matter. The watch only reads your mailbox unless a team member allows a specific action, such as replying or organizing mail.

When a reply calls for a response, Kite handles it as an inbound lead. It reads what the person asked, checks their fit again, and drafts an answer under 100 words with one next step, such as two meeting times or your booking link. With a CRM connected, it records the person, the reply, and the next step there.

When the batch is done, Kite posts a report in the conversation where you asked. The report lists who was emailed by name, which sends failed and why, the email copy it used, and the follow-up plan with dates. After the campaign ends, Kite records which subject lines, angles, and reasons for writing got replies in your team's knowledge base, so the next campaign starts from what worked.

When Kite finishes, the emails live in your own mailbox and the record of what happened lives in your Kite conversation.