Loading tutorials…
Loading tutorials…
A dedicated IP is the most-recommended SendGrid feature that's wrong for 80% of accounts. Below 100K sends/month, you'll hurt deliverability — the IP can't build reputation on low volume. Above 250K/month sustained, it pays off. Here's the decision framework and full warmup playbook.
Who this is forSendGrid Pro+ accounts sending 100K+/month consistently and considering a dedicated IP. Or accounts that already provisioned one and are watching deliverability tank because warmup was rushed. If you're under 100K/month, this tutorial mostly tells you NOT to do it.
What you'll need
Step 1
Volume threshold: 100K+/month sustained. Engagement threshold: 25%+ open rate. Business justification: a clear reputation story you can't get from shared IPs.
Volume test: are you sending 100K+ emails/month, every month, for the next 12 months? If volume is spiky or seasonal, shared IPs (SendGrid's pool) outperform dedicated.
Engagement test: is your open rate consistently 25%+? Open rate is the proxy for engagement; an IP with low engagement traffic builds a bad reputation, fast.
List quality test: bounce rate <1%, complaint rate <0.05%, unsubscribe rate <0.5%? Lower numbers mean reputation will compound positively over time.
Why provision: do you have a specific reason a shared IP doesn't work? Examples: regulated industry needing isolated infrastructure; very high-volume sender (1M+/month) where shared pool throttles you; sustained sends where you want full reputation control.
If you can't check all 4: stay on shared IPs. SendGrid's shared pool reputation is genuinely good for most use cases. The dedicated IP myth ("dedicated = better deliverability") is often false at moderate volume.
Step 2
SendGrid → Settings → IP Addresses → Add IP. Or open a support ticket for Pro plans (varies by plan tier).
SendGrid → Settings → IP Addresses.
Click Add IP. Some Pro+ plans include 1-2 dedicated IPs; additional cost $40-80/IP/month on most tiers.
For Premier or volume-negotiated contracts, you may need to request via Twilio support: include monthly volume estimate + use case.
Choose IP pool: assign the new IP to a pool (you can create multiple pools — e.g., "marketing-pool" and "transactional-pool" to isolate reputation).
After provisioning, the IP shows as Warming Up (auto-detected for the first 14-30 days). SendGrid throttles automatically during this window.
Document the IP address. Add it to your DMARC reports' allowed sources. Update DMARC monitoring tool (dmarcian, Postmark DMARC) to recognize it.
Step 3
Week 1: 500-1K sends/day to most-engaged 10%. Week 2: 1K-5K. Week 3-4: 5K-25K. Week 5-6: 25K-100K. Scale to full volume Week 7+.
Critical rule: warmup volume must equal what the IP can sustain post-warmup, scaled gradually. Sending 100K/day in Week 7 means warmup must scale toward 100K/day by then.
WEEK 1 (Days 1-7): 500-1,000 sends/day. Target the most-engaged 10% of your list (opened or clicked in last 14 days).
WEEK 2 (Days 8-14): 1,000-5,000 sends/day. Expand to engaged 25%.
WEEK 3 (Days 15-21): 5,000-15,000 sends/day. Expand to engaged 50%.
WEEK 4 (Days 22-28): 15,000-50,000 sends/day. Expand to engaged 75%.
WEEK 5-6 (Days 29-42): 50,000-100,000 sends/day. Full engaged 90-day list (anyone who opened or clicked in 90 days).
WEEK 7+ : full operational volume. Continue avoiding 180+ day inactives — they tank reputation at any phase.
Spread sends across the day — don't blast all 50K in 10 minutes on Day 22. Spread over 4-8 hours minimum.
Step 4
Gmail Postmaster Tools + SendGrid Stats + Microsoft SNDS. Watch for early-warning signals daily.
Gmail Postmaster Tools (postmaster.google.com): add and verify your sending domain. Watch Domain Reputation, IP Reputation, Spam Rate daily.
IP Reputation should climb from "None" → "Low" → "Medium" → "High" over weeks 2-4. If it goes High → Medium → Low, slow down volume immediately.
Spam Rate target: <0.1%. If it ticks above 0.3% in any day, PAUSE sending and diagnose before continuing warmup.
SendGrid → Stats → IP Stats: watch bounce rate, block rate, complaint rate per-IP. If bounce >2% or complaint >0.05% on the new IP, you're warming too fast.
Microsoft SNDS (sendersupport.olc.protection.outlook.com/snds/): track Outlook/Hotmail reputation. "Red" status = warmup is failing for Microsoft providers.
Daily check during warmup. Weekly during normal operations.
Step 5
Engagement quality > volume during warmup. Every send during warmup should target subscribers likely to open + click.
Build engagement segments BEFORE warmup starts. In Marketing Campaigns: "Engaged 14d," "Engaged 30d," "Engaged 90d."
Week 1 sends: ONLY to Engaged 14d. These are your most active subscribers — they open + click reliably, building positive reputation signals.
Week 2-3: expand to Engaged 30d. Still engaged but slightly less recent.
Week 4-5: expand to Engaged 90d. Full active list.
NEVER send to "All Contacts" or to 180+ day inactives during warmup. Even one such send can spike complaints + bounces and undo a week of warmup.
Marketing emails work best for warmup (drive opens + clicks). Transactional alone can work but signals are sparser — engagement is what builds reputation.
Step 6
SendGrid offers Automated IP Warmup on some Pro+ plans — it throttles your sends to the warmup schedule. Useful if you can't enforce manually.
SendGrid → Settings → IP Addresses → click your IP → enable Automated Warmup.
Automated Warmup throttles outbound sends per the schedule above. If you try to send 50K on Day 5, SendGrid spreads it across hours OR rejects excess as 4xx.
Pros: prevents accidental over-sending during warmup. Useful for teams without strict send-scheduling controls.
Cons: less flexibility for engagement-specific scheduling. May queue legitimate marketing sends.
Recommendation: use Automated Warmup AS A SAFETY NET while still planning manual engagement-based warmup. The manual discipline is what builds reputation; Automated just prevents catastrophic mistakes.
Step 7
Warmup is done when: IP Reputation is High in Postmaster Tools, Spam Rate <0.1%, sustained at target volume for 14 days.
Around Day 35-42, check Gmail Postmaster Tools: IP Reputation should be Medium or High.
Spam Rate consistently <0.1% over the previous 14 days.
Bounce rate <1%, complaint rate <0.05% over the previous 14 days.
Sustained target volume (100K/day or whatever your operational rate is) for 14 days.
If all 4 conditions met → warmup complete. Continue normal operations + weekly Postmaster Tools monitoring.
If conditions not met: extend warmup by 1-2 weeks at current volume. Diagnose any anomalies (which campaigns drove complaints? which segments bounced?). Fix before continuing.
Common mistakes
Provisioning a dedicated IP at <100K/month volume
What goes wrong: The IP can't build reputation on sparse volume. Mailbox providers treat low-volume new IPs as suspicious. Deliverability drops 20-30% vs the shared IP you came from. After 60-90 days of damage, most accounts end up returning to shared.
How to avoid: Don't provision a dedicated IP unless you're sustaining 100K+/month with 25%+ open rate. Below that, SendGrid's shared pool reputation is better than a new dedicated IP. Wait for the volume threshold.
Compressing warmup into 2 weeks
What goes wrong: Mailbox providers' reputation algorithms need 4-6 weeks minimum to register a new IP as legitimate. Compressing tanks the IP for 60-90 days of recovery, sometimes permanently.
How to avoid: Follow the 4-6 week schedule. 100K/day target requires the full 6 weeks. Compressing saves 4 weeks of warmup and costs 12 weeks of recovery.
Sending to inactives during warmup ("might as well clean them up")
What goes wrong: Inactives bounce + complain at 10-20x the rate of engaged subscribers. One inactives-included send during warmup can spike complaint rate, undo a week of progress, and signal to ISPs that you're sending unwanted mail.
How to avoid: Engaged segments only during warmup. Save list cleanup for after warmup completes — use a re-engagement automation on a separate sender or shared IP, NOT on the warming dedicated IP.
Not monitoring Gmail Postmaster Tools daily
What goes wrong: Reputation damage shows in Postmaster Tools 24-72 hours before it shows in SendGrid stats. Without daily monitoring, you find out 3-7 days late and have already shipped another 3-7 days of damage.
How to avoid: Postmaster Tools every morning during warmup. Microsoft SNDS too. Set a calendar reminder. 5 minutes daily for the warmup window catches issues before they cascade.
Failing to update DMARC/SPF when adding the new IP
What goes wrong: Sends from the new IP fail SPF alignment if SPF doesn't include SendGrid's mechanism (which covers all SendGrid IPs). DMARC reports show failures. Eventually mailbox providers downrank.
How to avoid: SPF: `include:sendgrid.net` covers ALL SendGrid IPs (shared + dedicated). No additional SPF changes needed when adding a dedicated IP. But verify the existing SPF record includes sendgrid.net.
Provisioning, then never sending — IP reputation never builds
What goes wrong: Provisioning a dedicated IP and only sending 10K/month means reputation never builds past 'unknown.' Mailbox providers treat the IP with suspicion indefinitely. Deliverability worse than shared.
How to avoid: Don't provision unless you'll sustain target volume. Use shared IPs at <100K/month. If volume grows: warmup over the 4-6 week ramp. Reputation requires sustained volume to maintain.
Recap
Done — what's next
How to set up SendGrid domain authentication — DKIM CNAMEs, SPF, DMARC, and link branding
Read the next tutorial
Hand it off
Dedicated IP warmup is the email-marketing task with the highest cost-of-error. Rushed warmups cause 60-90 days of deliverability damage that bleeds 20-40% of email revenue. A specialist who has run 20+ warmups will run the 4-6 week discipline with daily Postmaster monitoring and segment-by-segment scaling. Typical engagement is $1,200-2,500 for the full warmup at $14-16/hr.
See specialist rates
Almost never. Transactional volume is usually sub-100K/month and mixed with marketing on the same domain. SendGrid's shared pool handles transactional excellently. Dedicated IPs add cost + risk without measurable benefit for typical transactional volume.
You can, but you shouldn't. A bad marketing campaign poisons the IP for transactional too — order confirmations land in spam alongside the promo. Use SendGrid's IP pool feature to assign different IPs to marketing vs transactional, OR keep transactional on shared.
Don't accelerate to catch up. Resume at the scheduled volume for the day you missed (treat it as if you sent normally). Mailbox providers don't care about your calendar; they care about steady volume growth. Missing a day extends warmup by ~1 day.
$40-80/month per IP on Pro and higher plans. Some Pro plans include 1-2 IPs in the base price. Premier negotiated contracts often include multiple IPs at lower per-IP cost.
Maybe. Pause sending for 7-14 days, then resume at a lower volume than before (e.g., back to Week 2 levels) for another 14 days. If Postmaster Tools shows recovery, continue. If not, you may need to abandon this IP and provision a new one (yes, expensive lesson). This is why warmup discipline matters.
Yes — Settings → IP Addresses → IP Pools. Create a 'test pool' with 1 dedicated IP. Send experimental campaigns through this pool first. If a campaign damages the test pool, your main marketing pool isn't affected. Adds cost but isolates risk.
SendGrid
Gmail and Yahoo's 2024 bulk sender rules made domain authentication non-negotiable above 5K sends/day — and strongly recommended below it. SendGrid hides the link-branding step that most accounts skip, leaving every click flagged 'via sendgrid.net' in Gmail's clip warning. Here's the full auth stack.
SendGrid
Open rate dropped from 28% to 14%. Bounces jumped. Customer support is forwarding 'we never got the email' tickets. The instinct is 'subject lines' or 'content' — usually it's deliverability infrastructure. Here's how specialists diagnose SendGrid deliverability without guessing.
SendGrid
Suppressions are the dull, unglamorous part of email that decides whether your account survives the year. Gmail and Yahoo's 2024 bulk sender rules made one-click unsubscribe mandatory at scale, and SendGrid's suppression groups are how you respect that without losing transactional sends. Here's the full setup.
Mailchimp
Deliverability is the silent killer of email revenue. Open rates dropping from 28% to 16% feels like 'the algorithm changed,' but it's almost always one of three things: authentication, warmup, or list hygiene. Here's the playbook.
SendGrid
DIY SendGrid is the right call until it isn't. The signal isn't 'sending more emails' — it's that the cost-of-mistakes finally outweighs the cost-of-hiring. Here's the honest framework for when that line is crossed.