Cold Email Sequencer Integrations

When a cold email sequencer asks you to "connect your mailbox," it is asking for two things: permission to send mail as you, and permission to read the inbox so it can detect replies. Those two permissions can be granted three ways - OAuth, SMTP plus IMAP with an app password, or a native API integration - and the right choice depends on which sequencer you are using and which provider hosts the mailbox.

OAuth is the modern path. The sequencer redirects you to Google's or Microsoft's consent screen, you grant the requested scopes (typically gmail.send and gmail.readonly for Google, Mail.Send and Mail.Read for Microsoft Graph), and the sequencer receives an access token plus a refresh token. The sequencer never sees your password. You can revoke access at any time from Google Workspace admin or Microsoft 365 admin without breaking other connected apps.

SMTP plus IMAP with an app password is the legacy path that still matters. You generate a 16-character app-specific password from the provider's security settings - for Google Workspace, this requires 2-Step Verification to be enabled - and paste it into the sequencer along with the SMTP host (smtp.gmail.com or smtp.office365.com), the SMTP port (587 with STARTTLS or 465 with SSL), the IMAP host (imap.gmail.com or outlook.office365.com), and the IMAP port (993 with SSL). StackMail provisions app passwords per mailbox automatically when you export to a sequencer that needs SMTP - you never have to click through Google's app password UI manually.

Native API integrations are rare in cold email but show up in CRM-bundled sequencers like Close and HubSpot Sequences. The CRM has its own backend connection to the mailbox provider and tracks sends via webhooks rather than SMTP. From your perspective the setup looks identical to OAuth.

For reply detection specifically, IMAP is the protocol that matters. The sequencer polls the inbox every 5-15 minutes and matches incoming messages against tracked threads using the standard Message-ID and In-Reply-To headers. OAuth-based connections use the Gmail API or Microsoft Graph as a functional equivalent of IMAP - same outcome, modern transport. Either way, StackMail mailboxes ship with both SMTP and IMAP enabled because we provision standard Google Workspace and Microsoft 365 seats with no non-default restrictions.

Most cold email infrastructure providers either bundle a sequencer (Instantly, Smartlead, Lemlist) or tie their mailbox quality to a specific sequencer's warmup engine. That is good for the provider's lock-in metrics and bad for you. If you sign up for an Instantly mailbox plan and a year later you decide Smartlead's reply-handling is better for your team, you face a forced migration - new infrastructure, new mailboxes, new domains, and another 14-21 days of warmup.

StackMail does not sell a sequencer and we are not building one. We provision the mailbox once, in your own Google Workspace or Microsoft 365 tenant, and you point it at whichever sequencer fits this quarter's needs. Want to test Smartlead alongside Instantly? Provision 10 mailboxes, push five to each, run for two weeks, decide. Want to migrate your entire fleet from Reply.io to Apollo? Disconnect on the Reply side, connect on the Apollo side, no infrastructure change. The mailbox keeps its sender reputation because reputation lives on the Google or Microsoft tenant - not on the sequencer's warmup database.

Practically this means three things. One, you can A/B test sequencers without paying a migration tax. Two, if a sequencer raises prices or degrades, you switch in a day. Three, your mailboxes outlive your sequencer choices - they are infrastructure, not feature flags inside someone else's product.

The most common StackMail customer scenario: you just provisioned 30 mailboxes across 6-10 domains and you need them in Instantly tonight. Doing this per-mailbox would take 45-90 minutes (30 mailboxes times 90 seconds each, plus tab-switching overhead). The bulk-connect workflow takes about 5 minutes wall-clock.

1. Select the mailboxes in the StackMail dashboard. Use the workspace filter or select-all to pick the 30 mailboxes you just provisioned.
2. Click Export to Sequencer and choose Instantly. StackMail validates that every selected mailbox is past the warmup gate (default 14 days) and surfaces a warning for any that are not.
3. Choose connector type: OAuth bulk push (where Instantly will deep-link each mailbox owner to the consent screen) or SMTP/IMAP with auto-generated app passwords. Most agencies pick SMTP/IMAP for bulk because it does not require the mailbox owner to interactively click through OAuth on each account.
4. Confirm and run. StackMail generates app passwords, formats the payload Instantly's API expects, and POSTs each mailbox into the Instantly workspace you authorised. Instantly validates SMTP and IMAP credentials in parallel; failures surface back to your StackMail dashboard with a one-click retry.
5. Verify in Instantly. Open Instantly, navigate to Email Accounts, and confirm 30 mailboxes show as Connected with green status. Set daily limits and warmup policies in Instantly per your campaign needs.

The Smartlead bulk-connect flow is identical in shape. For sequencers without a public bulk API (Lemlist, Outreach, Salesloft), StackMail exports a sequencer-formatted CSV that ingestion tools accept in one click - still much faster than per-mailbox manual entry.

Reply detection is the part that breaks most often when infrastructure is misconfigured. Here is exactly how it works end to end so you can debug it when it does not.

When the sequencer sends an outbound message via SMTP, it stamps a unique Message-ID header (e.g. <abc123@instantly.ai>) and stores a record in its own database that ties that Message-ID to the campaign, the contact, and the cadence step. The recipient's mail client reads the message, composes a reply, and the reply's headers include In-Reply-To: <abc123@instantly.ai> and a thread-spanning References: chain. This is RFC 5322 standard behaviour - it works the same way regardless of which sequencer or mailbox provider you are on.

The reply lands in the StackMail mailbox's inbox folder via Google's or Microsoft's normal inbound delivery path. The sequencer's IMAP poller (or Gmail API watch, depending on the integration) checks the inbox every 5-15 minutes, sees the new message, parses the In-Reply-To header, looks up the Message-ID in its database, and marks the original campaign step as Replied. The sequencer then either pauses the cadence (default for cold outreach) or routes the reply to a human-handled inbox.

Two failure modes are common. First, if the recipient's mail client strips the In-Reply-To header (rare but happens with some Outlook configurations), the sequencer falls back to subject-line matching with the "Re:" prefix - less reliable but usually sufficient. Second, if the sequencer's IMAP poll interval is too long, replies show up in the dashboard hours after they arrived. StackMail mailboxes do not have any quirky IMAP configuration - they respond to standard polls in under a second - so any delay you see is on the sequencer side.

This is the single biggest StackMail advantage and the reason being sequencer-agnostic matters. Sender reputation - the thing that determines whether your mail lands in primary inbox or spam - is computed by Gmail and Microsoft based on the IP range, the authenticating domain, the SPF/DKIM/DMARC alignment, the message volume curve over time, and the recipient engagement signals. Notice what is not in that list: the sequencer.

When you switch from Instantly to Smartlead (or any other pair), the underlying mailbox is unchanged. Same Google Workspace or Microsoft 365 tenant. Same IP range. Same SPF/DKIM/DMARC records. Same domain reputation built up over the last 30, 60, 90 days. The receiving servers (Google's and Microsoft's spam filters) do not know or care which sequencer composed the message - they evaluate the sender, the headers, and the content.

What does reset is the sequencer's internal warmup history. Instantly tracks per-mailbox warmup metrics in its own database; if you disconnect and reconnect, that history is gone from the Instantly UI. But it was always a UI metric - the actual deliverability lives on the Google/Microsoft side and persists.

Practical recommendation: if you are evaluating a new sequencer, run it on 10-20% of your mailboxes for 2-3 weeks alongside the incumbent before flipping the rest. This gives you a real comparison without disrupting active campaigns. Once you decide, the cutover is a single dashboard action - disconnect from old, bulk-connect to new, done.

StackMail mailboxes are real Google Workspace and Microsoft 365 accounts. Any sequencer that accepts SMTP-with-app-password connections - which is essentially every cold email tool ever built - works with our mailboxes out of the box. We do not have to build a named integration for it; you just paste the SMTP host, port, email and app password into your sequencer's custom-SMTP form.

For the named-integrations list above, what we add is a CSV export pre-formatted for that sequencer's import (different sequencers expect different column names and orderings), a one-click bulk-push API where the sequencer offers one (Instantly, Smartlead), and deep-link helpers into the sequencer's OAuth consent screen so you do not have to navigate the sequencer UI manually.

If you use a sequencer we have not named and you would like a polished integration, send a note to support and tell us which sequencer plus your typical mailbox count. We typically add a connector inside two weeks if it is a tool we see across multiple customer requests. In the meantime, the custom-SMTP fallback works today.