- The 100,000-emails-per-month rule of thumb is oversimplified. Consistency of sending volume matters more than absolute volume in the shared-vs-dedicated decision.
- Shared IPs at well-managed ESPs deliver better than cold dedicated IPs for most senders because reputation is already established.
- A dedicated IP is a liability unless you have consistent volume, tight list hygiene, and the operational discipline to warm it properly.
- The real risk of shared IPs is neighbor contamination: a bad sender in your pool can degrade your delivery even if your own practices are flawless.
- Hybrid approaches (dedicated for transactional, shared for marketing or vice versa) often outperform committing fully to either model.
The shared-vs-dedicated IP question is one of the most frequently asked and worst-answered questions in email deliverability. The standard advice, "get a dedicated IP once you send 100,000 emails per month," is oversimplified to the point of being misleading. Many senders at that volume get worse results on dedicated IPs than they would have on well-managed shared infrastructure, and many senders below that volume benefit from dedicated IPs because their sending pattern demands it.
This guide replaces the volume heuristic with a real decision framework. By the end, you will know which model fits your specific sending pattern, and you will understand the tradeoffs you are accepting either way.
What Actually Changes Between Shared and Dedicated
Both shared and dedicated IP models deliver mail through the same SMTP infrastructure using the same authentication (SPF, DKIM, DMARC) and the same content processing. What changes is who controls the IP reputation.
On a shared IP, the ESP manages the reputation across all senders in the pool. Your sending behavior contributes to pool reputation, but it does not exclusively determine it. A well-managed pool has strict sender acceptance criteria, aggressive enforcement of spam complaint thresholds, and active monitoring for compromised accounts. A poorly-managed pool has none of these, and every bad sender drags the whole pool down.
On a dedicated IP, you own 100% of the reputation. Your behavior is the only input to how receivers judge your IP. This is a double-edged sword: you cannot be hurt by bad neighbors, but you also cannot be helped by good ones. A new dedicated IP starts with zero reputation and must be built from scratch.
The Real Cost of Dedicated IPs
Dedicated IPs have a sticker price (typically $30-$60 per month at most ESPs) but the real cost is much higher. The hidden costs:
Warmup Time and Volume Constraints
A new dedicated IP requires 4-6 weeks of careful IP warmup before you can send at full volume. During warmup, you are limited to progressively larger daily volumes to build positive reputation signals without triggering spam filters that treat volume spikes from unknown IPs as suspicious.
Typical warmup schedule:
| Week | Max Daily Volume | Typical Target |
|---|---|---|
| 1 | 50-200 | Most engaged subscribers only |
| 2 | 500-1,000 | Recent active subscribers |
| 3 | 2,500-5,000 | Broader active segment |
| 4 | 10,000-25,000 | Full active list |
| 5+ | Production | All targeted sends |
If your sending is urgent (a product launch, a seasonal campaign), warmup is not a viable path. You will either rush the warmup and damage the IP, or you will miss the window entirely.
Operational Overhead
A dedicated IP requires ongoing operational attention. Someone needs to monitor bounces, complaints, and reputation trends specific to the IP. Someone needs to maintain suppression lists aggressively because one bad send can damage a reputation that took weeks to build. Someone needs to respond quickly to reputation alerts from blocklist monitoring, Gmail Postmaster Tools, and Microsoft SNDS.
For a team with dedicated email operations staff, this is manageable. For a team where email is a secondary responsibility of a marketer who has 10 other things to do, it is a gap that opens up deliverability problems nobody notices until customers complain.
Volume Consistency Requirement
Dedicated IPs punish inconsistent sending. Receivers expect a dedicated IP to have stable volume patterns. If you send 50,000 emails on Monday, nothing on Tuesday through Saturday, and 50,000 again on the next Monday, you are creating reputation volatility that hurts delivery. A shared pool with thousands of senders averaging out produces stable volume patterns automatically; a dedicated IP does not.
When Shared IPs Win
Shared IPs are the right choice for a larger segment of senders than the conventional wisdom suggests. Specifically, you want a shared IP if any of these apply:
You Send Below 100K Emails Per Month
At low volume, you cannot generate enough engagement signals on a dedicated IP to establish strong reputation. A shared IP with thousands of other senders produces the steady volume and diverse engagement patterns that receivers use to build trust. Your own mail benefits from that pool-level reputation.
Your Volume Is Inconsistent
If you send 500,000 emails during a launch week and then nothing for three months, a dedicated IP is going to be a liability. Receivers see the volume spike, cannot correlate it with a history of good behavior on that IP, and filter accordingly. A shared IP absorbs your spike into the pool's existing traffic and you get treated the same as the other senders who are sending steadily.
You Do Not Have Email Operations Discipline
If your team does not have someone actively monitoring deliverability metrics, reading DMARC reports, and responding to reputation signals, a dedicated IP will degrade silently over time. Shared infrastructure at a reputable ESP has this monitoring built in. Use it.
Your Traffic Is Mixed (Transactional + Marketing)
Most small-to-mid senders run transactional and marketing through the same pipes. On a shared IP at a good ESP, the pool typically segments this automatically. On a dedicated IP, you need to manage separate IPs for each traffic type, which doubles the operational burden.
When Dedicated IPs Win
A dedicated IP is the right choice when you have specific, well-defined needs that shared infrastructure cannot meet.
You Send Consistently Above 500K/Month
At this volume, your own sending dominates what receivers see, and you benefit from full reputation control. You also have enough engagement data to measure your deliverability independently of pool dynamics. The 100K threshold commonly cited is a lower bound; 500K per month is closer to the point where a dedicated IP reliably wins over a shared one.
You Are in a Regulated Industry
Financial services, healthcare, government, and legal contexts often require isolation of sending infrastructure for compliance or audit reasons. A dedicated IP provides a clean boundary that makes compliance easier to document.
You Need Isolation from Reputation Risk
If you have been burned by a bad neighbor on a shared pool (which happens more often than ESPs like to admit), a dedicated IP eliminates that risk permanently. Your reputation becomes entirely your own.
You Need a Branded PTR Hostname
On a shared IP, your PTR record points to the ESP's generic hostname (sendgrid-ipool-01.sendgrid.net or similar). On a dedicated IP, you can usually specify a branded hostname (mail.yourdomain.com). This is a small but real trust signal, particularly for B2B contexts where recipients may inspect email headers.
Branded PTR hostnames on dedicated IPs also make header inspection more professional. When a receiver looks at your Received: headers and sees mail.yourdomain.com instead of mx-outbound-37.espname.net, the impression of legitimacy is measurably higher. For B2B senders where email is the primary business channel, this matters.
The Hybrid Approach
Many sophisticated senders split traffic across shared and dedicated infrastructure based on traffic type. Common patterns:
Dedicated for Transactional, Shared for Marketing
Transactional email (password resets, order confirmations, shipping notifications) has predictable, steady volume and very low complaint rates. It is ideal for a dedicated IP because you get consistent good signals. Marketing email has spikes and higher complaint rates, which are better absorbed by shared infrastructure.
Dedicated for Marketing, Shared for Transactional
The inverse pattern works when transactional volume is modest and marketing volume is large. A dedicated IP for high-volume marketing sends gives you reputation control, while shared infrastructure handles the smaller transactional stream without requiring its own dedicated IP.
Multiple Dedicated IPs by Traffic Type
Enterprise senders at very high volume often run multiple dedicated IPs, one per major traffic type or audience segment. This isolates reputation risk per-segment: if the marketing IP has a bad week, transactional mail is unaffected.
Practical note: Combining subdomain isolation with IP isolation gives you maximum flexibility. Use mail.yourdomain.com for marketing on IP A, and transactional.yourdomain.com for transactional on IP B. Each gets its own reputation at both the domain and IP level, which is the modern best practice.
How to Evaluate a Shared IP Pool
Not all shared pools are equal. If you are choosing an ESP for shared sending, evaluate the pool quality before committing.
Questions to ask (or answer from observable behavior):
- What are the acceptance criteria for new senders? Pools that accept anyone with a credit card are worse than pools that require list hygiene verification, volume disclosure, and content review.
- What is the complaint threshold for suspension? Pools that tolerate 0.5%+ complaint rates are worse than pools that cut senders at 0.1%.
- Are pools segmented by traffic type? An ESP that runs separate pools for transactional vs marketing, or for opt-in vs cold outreach, delivers better than one that mixes everything together.
- What is the published deliverability for the pool? Independent benchmarks for most major ESPs are available. Look for shared-pool inbox placement of 85%+ as a baseline.
- Does the ESP publish IP reputation data? ESPs that are transparent about their IP reputations (blocklist history, Google Postmaster Tools integration, SNDS participation) tend to be more disciplined than those that hide this information.
How to Run a Dedicated IP Properly
If you decide a dedicated IP is right for you, the operational disciplines that make it work:
- Warm up slowly and carefully. Follow the 4-6 week schedule. Do not skip steps. Start with your most engaged subscribers, not your full list.
- Monitor reputation daily during warmup, weekly after. Google Postmaster Tools, Microsoft SNDS, and blacklist checkers should be part of a daily review until the IP is established.
- Maintain consistent sending cadence. Dedicated IPs prefer steady daily volume over weekly blasts. If your natural cadence is weekly, consider splitting the send across multiple days.
- Aggressive list hygiene. Remove hard bounces immediately. Cull inactive subscribers more aggressively than you would on a shared pool. Every bad send directly damages your reputation.
- Configure branded PTR. Set up
mail.yourdomain.comPTR for the dedicated IP and verify FCrDNS. Do not leave the ESP's default hostname. - Monitor post-warmup reputation for drift. Reputations can degrade slowly over months as list quality changes. Monthly inbox placement testing catches drift before it becomes a crisis.
Migrating Between Models
Senders sometimes need to switch between shared and dedicated infrastructure. The migration is not trivial either direction.
Shared to dedicated: Start the dedicated IP warmup while continuing to send through shared infrastructure. Gradually shift traffic to the dedicated IP over 4-6 weeks. Do not abandon the shared IP until dedicated reputation is established.
Dedicated to shared: This is actually harder than the reverse, because moving an established sender into a less-controlled environment can expose you to neighbor reputation risk. Test with a subset of traffic first, monitor delivery metrics carefully, and keep the dedicated IP available for fallback.
The general deliverability improvement principles apply either way: consistent sending, tight list hygiene, and ongoing reputation monitoring are the things that actually move the needle, not the specific infrastructure model.
Frequently Asked Questions
The commonly cited 100,000 per month threshold is a lower bound. Realistically, 500,000 per month sent consistently across 5-7 days per week is where dedicated IPs reliably outperform shared infrastructure. Below that, consistency matters more than volume, and a well-managed shared pool usually wins.
Technically yes, but it almost never makes sense. Cold email has high complaint rates and unpredictable volume patterns, which are the opposite of what a dedicated IP needs to thrive. Most reputable ESPs also prohibit cold outreach on dedicated IPs to protect their broader IP range from reputation damage. Cold email is better suited to specialized sending infrastructure designed for that workload.
At a well-managed ESP, bad senders are detected quickly through complaint rate monitoring and suspended before they can significantly damage pool reputation. At a poorly-managed ESP, bad senders can degrade pool reputation for days or weeks before action is taken. This is why ESP selection matters more than shared-vs-dedicated: a well-managed shared pool outperforms a poorly-managed dedicated setup in most cases.
Multiple dedicated IPs make sense above roughly 2 million emails per month, or when you need to isolate reputation between traffic types (transactional vs marketing, for example). Below that volume, a single dedicated IP can handle the traffic, and splitting across multiple IPs dilutes reputation signals without providing meaningful benefit.
No. A dedicated IP guarantees reputation control, which is different from reputation quality. If your sending practices are poor, a dedicated IP makes your problems more visible and harder to hide behind pool averages. Dedicated IPs reward disciplined senders and punish undisciplined ones. Shared infrastructure partially insulates both ends of that spectrum.