Table of Contents
- Shared vs Dedicated IP at a Glance
- The rows that actually decide it
- How Shared IP Reputation Spreads
- When a Dedicated IP Hurts Delivery
- Why monthly totals mislead
- Which Sending Situation Fits Each IP
- The verdict by sending situation
- A migration sequence that holds up
- What We Check First on an IP
- Questions Before You Change IP
- Is a dedicated IP worth the extra fee
- How long before a dedicated IP behaves well
- Can we move back to a shared IP
- Our volume is seasonal, what then
- Common Mistakes When Choosing an IP
- When to Bring in an Expert
Do not index
Do not index
Your campaign is landing in spam, a migration is on the calendar, or someone in the growth channel has just proposed moving off the shared pool and onto a dedicated IP. All three situations lead to the same question, and the honest answer isn't "dedicated is better." A shared IP is one address used by several senders, so its reputation and its sending quota are pooled; a dedicated IP belongs to one organization, which removes neighbor risk and simultaneously removes the history a mature pool has already earned. Neither choice guarantees inbox placement, and neither repairs a bad list, broken authentication or a sending stream that shouldn't exist.
Table of Contents
Shared vs Dedicated IP at a GlanceThe rows that actually decide itHow Shared IP Reputation SpreadsWhen a Dedicated IP Hurts DeliveryWhy monthly totals misleadWhich Sending Situation Fits Each IPThe verdict by sending situationA migration sequence that holds upWhat We Check First on an IPQuestions Before You Change IPIs a dedicated IP worth the extra feeHow long before a dedicated IP behaves wellCan we move back to a shared IPOur volume is seasonal, what thenCommon Mistakes When Choosing an IPWhen to Bring in an Expert
Shared vs Dedicated IP at a Glance
Launch week goes fine until Tuesday. The newsletter that landed in Gmail's Primary all spring still lands. Then a new dedicated address goes live, and by Wednesday the same campaign is returning 421 and 450 deferrals, while password resets still leaving on the old shared pool arrive without a hitch. Nothing in the list, the copy or the authentication changed. Only the tenancy did.
That distinction is the whole decision. A shared IP is one address used by more than one sender, so every tenant feeds a single IP reputation and a single sending quota. A dedicated IP carries one organization's mail, so you stop inheriting a neighbor's complaints and you also lose the pool's accumulated sending history. Neither one guarantees inbox placement.
Criterion | Shared IP | Dedicated IP |
Tenancy | Many unrelated senders on one address | One organization |
Who owns reputation | Pooled, any tenant's complaints affect the rest | Yours alone, including a thin or new history |
Quota | Shared, exhaustion can stop every domain on that IP | Your sending only |
Neighbor and blocklist risk | Inherited from the pool | Isolated, but your own listing is yours to clear |
Daily volume needed | Pool history can carry lower or bursty volume | Needs steady daily mail to build positive signals |
Who mitigates an IP block | Usually the ESP | You, or your ESP if they still operate the address |
What it costs | Normally included in the plan | Often a paid add-on per address, billed monthly |
Best fit | Uneven volume, healthy pool, limited ops ownership | Predictable volume, stream separation, active monitoring |
Read the table across, row by row, rather than down a column. Teams often fixate on the reputation row and skip the one that decides the outcome: whether they can put mail on the wire every single day.
The rows that actually decide it
Tenancy is measurable, and the measurements aren't subtle. A large-scale Internet measurement study of malicious websites found shared IPs hosting an average of 2,624.18 second-level domains, against 796.08 on dedicated IPs, and dedicated addresses stayed active for an average of 2.3 years against 1.2 years for shared ones. Shared infrastructure turns over much faster than dedicated infrastructure. That turnover is why a pool can be excellent in March and mediocre in June, and why "the pool has history" is only useful if the history is current.

That buying decision is not permanent. Plenty of programs start on a well-run pool, outgrow it, and move to a dedicated address once daily volume and monitoring capacity exist. The mistake is treating the address as a status symbol rather than an operational unit.
Either way, verify that SPF and DKIM align with your sending domain and that a DMARC record exists, because tenancy doesn't compensate for authentication that doesn't line up.
How Shared IP Reputation Spreads
Your reputation on a shared address is partly written by strangers. When another tenant pushes complaints past a provider's threshold, the filtering penalty lands on the address, which means it lands on every domain sending through it. Your copy, your list and your authentication stay exactly the same while your inbox placement absorbs someone else's behavior.
The failure mode is rarely dramatic. Opens slide on the promotional stream over ten days. The welcome series lands in Spam for the first time in a year. The list wasn't bought, the authentication didn't break, and the subject lines are the ones that worked all winter. What changed was a neighbor: a complaint spike, a blocklist listing, or a volume surge from a tenant three accounts down the same rack.
The second shared resource is the quota. When a shared address hits its provider-enforced sending ceiling, every domain on that address can stop sending, including the password resets that have nothing to do with the campaign that caused the surge. A promotional blast at nine in the morning can silence a login flow by twenty past.
Treat this as correlation until proven otherwise. Legitimate senders share infrastructure with other legitimate senders all day, and it works. The shift is in the diagnosis: the question moves from "what's wrong with our mail" to "who else is on this address," and a pool-level answer requires a pool-level investigation.
Blocklist exposure is common enough to be an assumption. Across the sending domains observed by Mailwarm, MailAdept's parent company, listings on at least one DNS blocklist are routine, and some of those listings sit on shared infrastructure that unrelated senders depend on.
Going dedicated removes the neighbor. It does not remove the work. Complaint rates, feedback loops and engagement still decide placement, and now every signal on the address is yours to fix or yours to live with. That is why a diagnosis comes before any IP change: if a campaign suddenly misfires, rule out your own copy before blaming the pool, which is what a spam words checker is for. Diagnosing a neighbor when the trigger was a phrase in your subject line wastes a week.
Ignoring pool reputation has a concrete cost: a healthy campaign inheriting someone else's block, support answering missing-mail tickets, and revenue mail sitting in spam while the dashboard cheerfully reports delivered.
When a Dedicated IP Hurts Delivery
A dedicated IP creates more risk when the sender cannot produce consistent positive signals. Treat it as an accountability tool, not an inbox upgrade. It removes the neighbor, then puts the full burden of complaint rate, authentication health, deferrals and sending rhythm on you.
The hard case is steady volume. A sender moving 3,000 to 5,000 messages every day, in roughly the same pattern, to a responsive list has a credible reason to own an address. A sender pushing 100,000 messages a month in four irregular bursts, with nine or ten quiet days between them, does not. The monthly total looks larger. The daily signal is weaker.
Why monthly totals mislead
Mailbox providers read daily patterns, not monthly reports. A new dedicated address needs a steady drip of accepted, engaged, complaint-free mail before it becomes predictable. Idle gaps cool that signal. A sharp spike after a quiet stretch reads like a compromised host or a fresh bulk sender, not a planned campaign.
Compliance thresholds do not change that judgment. Google treats roughly 5,000 messages to personal Gmail accounts within 24 hours as its bulk-sender line, with SPF and DKIM authentication, a DMARC record, one-click unsubscribe and a spam rate kept under 0.3%, ideally under 0.1%, as set out in Google's bulk sender guidelines. Microsoft's Outlook.com high-volume sender requirements, in force since May 2025, add another set of authentication and complaint expectations. Clearing those requirements keeps you in the game. It does not prove that a dedicated address fits your sending pattern.
The benchmark comparison in the previous section makes the same point from another angle: dedicated IPs had a lower spam rate, while shared IPs performed better on Gmail Primary placement. Keep those measurements separate. Fewer complaints help, but placement still depends on engagement, list quality, content and domain reputation.
Metric | Dedicated IPs (7,166) | Shared IPs (8,008) |
Spam rate | 11.99% | 13.22% |
Gap in spam rate | 1.23 percentage points lower | baseline |
Gmail Primary placement | 19.24% | 21.09% |
Gmail Promotions placement | 68.77% | 65.69% |
Read those columns as separate measurements, not as a chain of cause and effect. The dedicated group had the lower spam rate and reached Gmail's Primary tab less often. That result kills the lazy assumption that buying an address automatically buys better placement.
The bad dedicated outcome has a recognizable shape. Early traffic gets throttled. 4xx deferrals arrive while reputation is thin, the ESP dashboard shows retries, and someone pushes the rest of the list to catch up. That decision turns a slow ramp into a complaint spike. The address then carries the damage, and recovery takes longer than the original delay.
Score four things before asking the ESP for an address: daily consistency of volume, the share of recipients on Gmail and Microsoft, complaint history by stream, and whether anyone can monitor and warm the address continuously. Two weak answers are enough to keep the shared pool. It is already absorbing variability you have not learned to control.
Which Sending Situation Fits Each IP
IP choice is not an account-level toggle. Transactional mail, lifecycle campaigns, promotions, outbound prospecting and support replies have wildly different complaint and engagement patterns, and a dedicated address does not separate them by itself.
Put cold outbound on the same address as password resets and you've concentrated damage rather than isolated it. One bad prospecting week then shows up as login emails timing out, and the support queue fills with "I can't get into my account" tickets that have nothing to do with email strategy. A shared pool has the opposite problem: it hides the cause, because another tenant may be involved and the provider, not you, files the mitigation request.
The verdict by sending situation
Situation | Better default | Why | Do not |
Irregular newsletter under a few thousand messages a day | Healthy shared pool | Pool history covers the gaps | Buy a dedicated IP to look more serious |
Steady 3,000 to 5,000 or more marketing messages every day | Dedicated marketing IP | Enough signal to own the reputation | Mix promotions with cold outbound |
Password resets, receipts, invoices | Separate subdomain, dedicated only if daily volume is real | Stops campaign complaints from blocking login mail | Leave transactional on the promo IP |
Cold outbound or purchased-list risk | Isolated stream, often its own domain and IP | Complaints stay off customer mail | Share it with lifecycle or receipts |
Several clients or brands on one ESP account | Separate domain per client, shared or dedicated by that client's volume | Stops cross-client blame | One dedicated IP for every client |
Sudden deferrals on a long-used shared pool | Diagnose the pool before moving | The neighbor may be the cause | Flip everything to a cold dedicated IP the same day |

What separation buys you is narrow and valuable: a complaint on the promotional stream that never touches a login email. It doesn't improve copy, fix a stale list or raise engagement.
A migration sequence that holds up
Don't treat this as a warmup tutorial, because warming is its own project with its own calendar. Treat it as a change-control exercise.
- Baseline for one week. Record spam rate, deferrals and complaints by stream before anything moves, so you can tell improvement from noise.
- Move the healthiest slice first. Send your most engaged, most recently active segment on the new address and keep everything else where it is.
- Write rollback criteria in advance. If deferrals climb past baseline or complaints double on a stream, move that stream back. No debate required in the moment.
- Review weekly against the baseline, not against opens. Opens are a weak proxy for placement.
Cold outbound deserves its own treatment, and cold email deliverability covers the domain separation and complaint tolerance it needs.
The cost of skipping separation is specific. One bad campaign blocks receipts and login mail at the same time, support tickets climb, and the domain reputation the business owns takes the blame for what was really an IP and stream problem.
What We Check First on an IP
Don't change IP type until the current failure is named. Teams that skip that step pay for a dedicated address and migrate a reputation problem onto it intact, or keep buying infrastructure while a broken list keeps generating complaints. Here's the order a consultant works in.
- Separate acceptance from inbox. The ESP's delivered figure can read 99% while Google Postmaster Tools shows a spam rate climbing and placement sliding. Delivered means a server said yes, not that a person saw it.
- Read the deferral text. A 421 or 450 naming rate, unusual traffic or "try again later" is throttling and usually temporary. A 550 naming a blocklist or a policy is a refusal and needs a different fix. Paste the code into your notes before you do anything else.
- Confirm whose address sent the mail. Read the Received and Authentication-Results headers and check whether the ESP rotated pool IPs between campaigns. On a shared pool, the sending address may not be the one you checked last month.
- Check that address against blocklists and read the spam-rate and domain-reputation views in Postmaster Tools, keeping Microsoft 365 deferrals separate from Gmail problems. The mailX blacklist checker, built by the same company as MailAdept, returns current listing status without an account.
- Split the last 14 days by stream. Promotions, lifecycle, transactional and outbound. A spike confined to one stream usually means the IP label isn't the root cause at all.
- Verify SPF and DKIM alignment and that a DMARC record exists, which takes minutes rather than a project, and is covered in email authentication.
- Only then decide. Shared, dedicated, or a stream split with one stream moved and the rest left alone.

One number belongs alongside those checks: hard bounces should stay under the common 2% guidance, and a list above that line will damage a brand new dedicated address faster than any neighbor ever could. Mailbox providers revise their threshold language, so confirm current figures in their own sender documentation rather than in a blog post.
Deliverability is a retention problem as much as an infrastructure one, and this explanation of why deliverability matters for retention revenue covers the revenue side of what a blocked stream really costs.
Questions Before You Change IP
Is a dedicated IP worth the extra fee
The fee is rarely the deciding factor. Providers typically bill a modest monthly add-on per address, but the bigger expense is monitoring: blocklist checks, complaint review, deferral logs and warmup discipline. Paying for an address nobody watches is worse than staying on a pool that someone else maintains.
How long before a dedicated IP behaves well
Reputation builds from consistent daily sending and complaint-free engagement over weeks, not days. An idle week sets the address back. A long-established pool carries history that a new address has to earn from zero, so expect a ramp period and don't judge placement in the first fortnight.
Can we move back to a shared IP
Yes, and it's often the right call when volume drops or the campaign schedule becomes irregular. Ask the ESP how your domain rejoins the pool, whether the sending subdomain changes, and what transition period applies. Rejoining is usually simpler than leaving, but it isn't instant.
Our volume is seasonal, what then
Keep the dedicated address only if the quiet months still send a steady floor of transactional mail. Otherwise stay shared, or move the seasonal marketing stream to the pool and keep transactional separate. Parking a silent address for four months and expecting it to be trusted at peak is a plan that fails.
Common Mistakes When Choosing an IP
A PTR mismatch on the new address. Reverse DNS that doesn't match the sending domain makes a clean dedicated IP look untrustworthy, and the symptom is deferrals that survive every content fix. Set it before the first send, not after the first complaint.
Judging the switch by open rate. Moving from Primary to Promotions looks like an IP failure when the opens were inflated by a tracking pixel in the first place. Watch complaint rate, deferrals and measured spam placement instead.
Assuming a dedicated address removes monitoring. It increases the monitoring duty. A new address has no history to fall back on, and the spam-rate and feedback-loop views matter more than they did on the pool, not less.
Leaving the address with no named owner. When nobody is responsible for reading Postmaster Tools and the deferral log, the week a neighbor or a list goes bad is the week nobody notices. Ownership is the cheapest control you can buy.
Buying a dedicated IP while sharing a domain with promotional mail. Splitting addresses doesn't split domain reputation. If promotions still send from the same domain, complaints still land on the identity that carries your password resets.
A shared pool with a named owner beats an unmonitored dedicated IP every time. Infrastructure without attention is just a more expensive way to be surprised.
When to Bring in an Expert
The decision has outgrown an internal guess when the same deferral returns after a list cleaning, when Gmail and Microsoft disagree for weeks, or when a migration or a new dedicated address is already on the calendar. It has also outgrown a guess when nobody owns the pool-versus-dedicated call, or when promotional mail and password resets keep knocking each other offline.
A free deliverability audit looks at authentication, DNS configuration, domain structure, routing, provider setup and sender reputation risk at a single point in time. Ongoing help is a different shape: a dedicated deliverability expert joins the team, calls with experts are included in every plan, and the subscription covers fixes first, then monitoring with 12 or more daily checks per domain and a weekly expert review. Authentication problems are often quick to correct, while reputation recovery after a bad IP move takes longer and depends on sending history, which is why a one-time DNS change is the wrong fit for this problem. MailAdept does not guarantee inbox placement, and any provider that does should be treated with suspicion.
Shared versus dedicated is a tenancy and accountability choice, not a quality tier. Daily volume and stream separation decide it, and monitoring is part of whichever option you pick.
The right answer depends on your cadence, your stream structure and who's watching the address, not on which option sounds more premium. If deferrals started after an infrastructure change and no one on the team owns the sending domain, MailAdept assigns a dedicated deliverability expert to audit the domains, streams and provider setup, fix what's hurting placement, and then keep every sending domain under daily checks with a weekly expert review. Start with the free deliverability audit and get a straight answer on whether your volume can carry a dedicated IP or should stay on a healthy pool.
