Smartlead Deliverability Repair Protocol

The protocol in one paragraph

Cold email deliverability is a reputation problem dressed up as a copywriting problem. When replies collapse, operators tweak subject lines and burn another week. The actual failure is structural: an unaligned DMARC record silently spam-folds every send from a domain, a mailbox too young to have a reputation, a list with 10–15% invalid addresses, or volume that outruns what the inbox can absorb. This protocol repairs all four in a fixed order because order matters — authenticating a domain you are about to torch with a bad list just makes the damage auditable. Run it as a checklist, not a menu.

Step 0: isolate the blast radius

The single most important decision is where you send from. Cold outreach carries inherent risk: some recipients mark you spam, some addresses are dead, some ISPs throttle new senders. None of that risk should touch the domain your customers use to email you, your investors use to reach you, or your app uses for transactional mail.

Never send cold email from your primary business domain. Set up separate sending domains and rotate across them. If one develops a reputation problem, the rest of your operation is untouched.

The 2026 pattern is 2–3 mailboxes per sending domain. A 30-mailbox operation therefore needs 10–15 domains. Buy them as close variations of your main domain (e.g. reach-yourcompany.com, getyourcompany.com), configure DNS for each, forward them to your main site so they look legitimate, and connect them all to Smartlead. Smartlead's unlimited-mailbox model makes this cheap to scale — the constraint is domains and verification, not seats.

Authentication: SPF, DKIM, DMARC aligned

Every sending domain needs all three, and they must agree with each other. A mismatch between the domain in the From header and the domain DKIM signed is itself a red flag.

SPF authorizes which servers may send on the domain's behalf. For Google Workspace mailboxes that is:

v=spf1 include:_spf.google.com ~all

DKIM adds a cryptographic signature. Your provider gives you a TXT record at a selector host (e.g. google._domainkey.yourdomain.com) containing the public key — paste it verbatim, do not reformat the long string.

DMARC ties them together and tells receivers what to do on failure. Start at p=none so you can watch reports without blocking mail, then move to p=quarantine once the reporting shows clean authentication:

v=DMARC1; p=none; rua=mailto:[email protected]

The trap: a half-configured DMARC record that points at quarantine or reject while SPF/DKIM are still misaligned will spam-fold everything. Smartlead's pre-flight checker catches most of this — run it. And set up a custom tracking domain; the shared default tracking domain carries other senders' reputations, which is not yours to inherit.

Warmup: the 21-day floor

New domains and new mailboxes have no reputation. Sending cold from them is the fastest route to spam. Smartlead's peer-to-peer warmup network generates realistic send-and-reply traffic across its pool and runs continuously, even after campaigns start.

The data is unambiguous: campaigns launched before 14 days of warmup underperform identical campaigns launched at day 21+ by wide margins, especially against Microsoft-hosted recipients. The realistic on-ramp for a new deployment:

• Days 1–3: infrastructure. Buy domains, configure DNS (SPF/DKIM/DMARC, start DMARC at p=none), create mailboxes, connect to Smartlead.
• Days 3–21: warmup plus list preparation in parallel. This is when you source, enrich, and verify — it costs zero calendar time because it runs alongside warmup. No excuse for launching day 21 with an unverified list.
• Days 21–30: controlled launch at 10–15 cold sends per mailbox per day while warmup continues. Ramp per-mailbox volume by roughly 20% per week toward a 25–40/day cruising altitude.

The mistake that ends careers here: stopping warmup after launch. Warmup generates ongoing engagement signals that sustain reputation. Stop it and those signals stop accumulating; deliverability erodes within six to eight weeks. Keep it running indefinitely.

Size your bounce budget before you send

This calculator turns the thresholds above into hard numbers for your setup. It tells you the steady-state daily volume your mailbox count supports, how long the ramp takes, and — most usefully — how many invalid addresses your list can contain before you cross the 2% bounce danger line.

Volume ramp and per-inbox caps

Volume is the lever operators abuse first because it feels like progress. It is also the lever ISPs watch most directly. Keep per-inbox cold volume in the 30–50/day safe band; the 50–100 band is defensible only on well-warmed, established inboxes; past 75 on Google Workspace you are asking for a throttle.

List hygiene: verify before send

This is the number-one mistake, period. Teams spend weeks perfecting warmup while sitting on lists with 10–15% invalid addresses. Verification costs pennies per email. Bounces cost you the domain.

Verify every new list before the first send — whether it came from a conference scan, a scraped export, or a purchased file. Verification tools classify addresses into valid, invalid, catch-all, and risky. Send only to verified-valid. Maintain bounce rate below 2%; a hard bounce means the address does not exist and is permanent, so purge hard bounces and never re-attempt them.

The cheapest credit is the one you never spent. The cheapest bounce is the invalid address you removed before the send, not the domain you rebuilt after the blocklist.

Monitoring: watch the signals, not the open rate

Open rates are unreliable — Apple Mail Privacy Protection inflates them and open-tracking pixels suppress them. Reply rate is the clean signal; positive reply rate is cleaner still. But reputation monitoring is what tells you a problem exists before replies collapse.

• Google Postmaster Tools — domain reputation (high/medium/low/bad), IP reputation, delivery errors, and spam rate near real time for Gmail recipients.
• Microsoft SNDS — complaint and spam-trap data for Outlook recipients.
• DMARC reports — review them weekly for the first month, then monthly. If you set DMARC to p=none and never read the reports, you are monitoring nothing.

Check complaint rate weekly on both platforms. Above 0.1% is the warning line; at or above 0.3% Gmail starts routing you to spam. A sudden shift to "low" or "bad" domain reputation in Postmaster Tools explains an inbox-placement problem faster than any copy audit.

Where each lever actually moves the needle