SMTP is only the transport protocol. The product is the infrastructure behind it: who operates the routes, how reputation is shared, what evidence is available, and what happens when performance changes.
SMTP is the protocol used to move email between servers. Calling a product an SMTP box does not tell you whether the underlying infrastructure is shared, dedicated, static, rotating, or how it is operated.
Some providers use the label to distinguish custom sending infrastructure from a conventional Microsoft 365 or Google Workspace mailbox. Providers use the label differently, so the important questions begin after that distinction.
Who operates the servers and IP addresses? Which customers share reputation? How is bad traffic isolated? What can the buyer inspect? What happens when an IP starts receiving deferrals or appears on a blocklist? How does the provider measure long-term performance?
A cheap mailbox price does not answer any of those questions.
With an SMTP-box product, you are buying access to a provider-controlled sending system. Its long-term value depends on pool management, tenant isolation, abuse controls, diagnostic access, reputation response, and continuity.
Shared IPs can make sense for lower volume or less predictable sending patterns, where one customer may not generate enough steady traffic to establish and maintain a dedicated IP. AWS recommends shared IPs for this use case, while Postmark generally reserves dedicated IPs for customers with substantial monthly volume. Those documents describe managed delivery platforms; they do not prove that a particular SMTP-box product will perform well for cold outbound. AWS Postmark
Large delivery platforms publish different examples of pool controls. Microsoft places messages it classifies as spam in a separate high-risk pool. Twilio says it dynamically groups shared-pool customers using engagement-quality signals. AWS describes managed dedicated pools that scale and warm new IPs by receiving network. Each example shows that pool operations matter; none proves that another provider uses the same controls. AWS Microsoft Twilio
Ask who shares the route, how tenants are separated, what triggers quarantine or throttling, and what evidence the customer receives when performance changes.
Every sender on a shared IP contributes to the history receivers see for that IP. Google says the activity of any sender using a shared IP affects the reputation of all senders on it. AWS says shared reputation is not isolated. Twilio warns that another customer on a shared IP can contribute to throttling or blocking. Google AWS Twilio
Shared-pool providers therefore need admission controls, complaint handling, abuse detection, and tenant isolation.
The buyer should know whether cold outreach shares infrastructure with other cold outreach, how customers are grouped, and what happens when one tenant creates trouble. Ask for concrete information about admission, complaint handling, tenant isolation, and incident response rather than relying on the phrase "managed pool" alone.
New IPs have little or no sending history. AWS says mail from a new IP may be blocked or placed in junk until the address is introduced gradually. Google tells senders to maintain consistent traffic, avoid bursts, and increase volume slowly. SendGrid also warns that an IP warm-up does not guarantee a good reputation. AWS Google SendGrid
A dedicated IP gives one sender more control over IP reputation. It also makes that sender responsible for building and maintaining it. Low or erratic volume may not provide enough representative traffic to justify the model.
Compare shared and dedicated infrastructure on control, tenant isolation, volume consistency, monitoring, and the quality of the operator managing it.
Some SMTP-box products make IP rotation central to the offer. Maildoso, for example, says its customers cannot pin mailboxes to a dedicated IP or a hand-picked subset of its pool. Its current product page says rotation makes messages less visible to spam filters, while its help center says rotating across a large pool spreads volume and risk. Maildoso product page, accessed September 3, 2026 Maildoso help center, accessed September 3, 2026
Rotation changes which IP carries a message. The public product page does not disclose the methodology behind its claim that this makes messages less visible to spam filters.
Google recommends using the same IP for the same category of mail when possible and says it evaluates both domain and IP reputation. Yahoo says its decisions can consider IP, domain, URL, sender, ASN, DKIM, DMARC, and recipient complaints. AWS explains that a newly introduced IP begins without an established history. Google Yahoo AWS
Changing the sending IP does not change the authenticated domain, content, links, list quality, or recipient behavior associated with the campaign. Google and Yahoo both document reputation signals beyond the sending IP. Buyers should ask for the methodology and longitudinal evidence behind the visibility claim.
A sending IP can develop poor receiver telemetry, receive repeated deferrals, or appear on a blocklist. Google says messages from blocklisted IPs are more likely to be marked as spam. A listing is evidence to investigate, not a universal verdict. Twilio says not every blocklist affects every account, while AWS says different DNS-based blocklists can have different effects on delivery and recommends confirming that the rejection actually cites the list. Google Twilio AWS
Maildoso's help article says that if a tool shows one of its sending IPs on a blacklist, "it's almost always a non-issue," because rotation means the tool may be flagging a previously used IP. It also says Maildoso monitors the pool, rotates listed or poor-reputation IPs out, and does not require customers to manage sending IPs themselves. Maildoso, accessed September 3, 2026 Those are Maildoso's operating claims. For a specific incident, the buyer should still ask which route carried the message and whether deferrals, rejections, or placement changed at the affected receiver.
An operator should first determine whether the IP was active, which receivers were affected, what the full SMTP responses said, which domains and tenants used the route, and whether complaints, bounces, or volume changed. The provider may reduce traffic, isolate a tenant, remediate the listing, retire the IP, or move a controlled portion of traffic. The action should follow the evidence.
Large providers can add, lease, and retire IP space. The important constraint for email is that a newly activated sending IP does not arrive with positive history. Google, AWS, and SendGrid all recommend controlled, consistent introduction of traffic from new IPs. Google AWS SendGrid
A provider can rotate routes, but it should be able to explain how new routes are introduced, monitored, isolated, and removed without creating unstable sending patterns. A larger pool therefore needs a documented process for introducing and monitoring new routes.
An advertised delivery percentage may sound like an inbox-placement result when it actually measures SMTP acceptance.
AWS defines delivery as successful handoff to the recipient's mail server. Twilio explains that a final 250 OK after message content can still be followed by junk placement or other downstream handling. Google says it cannot guarantee that messages sent by email providers will pass Gmail's spam filters. AWS Twilio Google
Before relying on a percentage, ask:
Without the methodology, the percentage cannot answer a capacity-planning question.
Before moving campaign capacity onto SMTP boxes, get direct answers to these questions:
Apply the same questions to Microsoft- and Google-based services.
The Maildoso product page and the two help articles cited above describe the architecture and operating claims. When accessed on September 3, 2026, those pages did not provide a dated cohort analysis of long-term inbox placement or incident recovery.
A mailbox price by itself does not show the quality of IP operations, abuse controls, monitoring, replacement procedures, or incident support.
Mailrun inboxes run on Microsoft Azure infrastructure and are managed by Mailrun. They are not Microsoft 365 or Exchange Online accounts. Buyers should apply the same questions to us: how the route is operated, what evidence is available during an incident, how authentication is handled, and who is accountable when performance changes.
Size domains and density to your sending target and see the capacity that holds.
Book a Call