---
title: "My Cold Emails Are Going to Spam - Should I Switch Inbox Provider?"
description: "A 20-minute test for whether your inbox provider is actually the cause of cold email landing in spam. The four things to rule out first, the four signals that do mean switch, a decision table, and how to switch without compounding the damage."
canonical: "https://maildeck.co/blog/cold-emails-going-to-spam-should-i-switch-provider"
last-updated: "2026-10-03"
---

> A 20-minute test for whether your inbox provider is actually the cause of cold email landing in spam. The four things to rule out first, the four signals that do mean switch, a decision table, and how to switch without compounding the damage.

Guide

# My Cold Emails Are Going to Spam - Should I Switch Inbox Provider?

Oct 3, 2026 10 min read

Contents

Short AnswerRule Out the Four Things That Look Like a Provider ProblemFour Signals That Do Mean You Should SwitchThe Evidence to Gather Before You Call ItThe Decision TableIf You Do Switch, Switch to the Right ThingSwitching Without Compounding the DamageFrequently Asked QuestionsMethodology

## Short Answer

Usually not. In the large majority of cases we see, cold email that lands in spam is not caused by the inbox provider, and switching providers moves the same problem to new infrastructure — at the cost of a fresh warmup window and a new set of cold domains.

Infrastructure determines roughly 60% of the deliverability outcome and domain health plus copy the other 40%, so the provider is worth examining. But before you switch, three causes account for most of what people blame on their provider:

- Broken DNS authentication. 67% of the 1,000+ cold email domains we have audited carried at least one critical error — 23% had multiple SPF records on one domain, which invalidates SPF entirely, 19% had no DMARC, 14% had SPF ending in +all.

- Sending over the daily limit for the inbox type. The limit is a property of the inbox, not of your ambition, and exceeding it triggers spam classification at the provider's end.

- List quality. A bounce rate above 7% or a spam complaint rate above 0.3% damages the domain faster than any provider can compensate for.

This page is the 20-minute test that tells you which of those you have — and the four signals that do mean your provider is the problem.

## First, Rule Out the Four Things That Look Like a Provider Problem

### 1. Check authentication before anything else

Run your sending domain through [MXToolbox](https://mxtoolbox.com/) or EasyDMARC. You are looking for exactly one SPF record resolving in under 10 lookups, DKIM signing every outgoing message, and a DMARC record present. If any of those fails, stop here — this is your problem, it is free to fix, and no provider change improves it.

| Record | What must be true | Where it is specified | Failure we see most |
|---|---|---|---|
| SPF | Exactly one record on the domain, resolving within the 10-DNS-lookup limit | [RFC 7208](https://datatracker.ietf.org/doc/html/rfc7208) (the lookup limit is [§4.6.4](https://datatracker.ietf.org/doc/html/rfc7208#section-4.6.4)) | Two SPF records on one domain, which invalidates SPF entirely — 23% of audited domains |
| DKIM | Every outgoing message signed, with the public key published and resolving | [RFC 6376](https://datatracker.ietf.org/doc/html/rfc6376) | Selector published but the provider never enabled signing |
| DMARC | A record present, aligned with the domain you send from | [RFC 7489](https://datatracker.ietf.org/doc/html/rfc7489) | No DMARC record at all — 19% of audited domains |

These are not opinions about best practice; they are the conditions the specifications set, and a receiving server applies them before it ever looks at your copy. 67% of the 1,000+ cold email domains we have audited carried at least one critical error here, which is why this check comes before every other question. Our [SPF, DKIM and DMARC setup guide](https://maildeck.co/blog/spf-dkim-dmarc-cold-email-guide) has the exact records and the full error analysis.

### 2. Check your sends per inbox per day against the limit for your inbox type

| Inbox type | Safe cold sends / inbox / day |
|---|---|
| Google Workspace | 18–22 |
| Microsoft 365 Outlook — Premium licence | 8–10 |
| Microsoft 365 Outlook — Normal licence | 4–5 |
| Private SMTP | 10–15 |

Divide your daily volume by your inbox count. If the answer is above the row for your inbox type, you have found the cause. Cut to the limit and allow 2–4 weeks of reduced volume for the reputation to recover; this is slower than buying new inboxes and it is the only thing that actually works. Also check the interval between sends from one inbox — we hold a minimum of 61 minutes.

### 3. Check bounce and complaint rates

Above 7% bounces or 0.3% spam complaints, your list is burning your domains and the provider is a bystander. The 0.3% figure is not ours — it is the threshold in Google's own [email sender guidelines](https://support.google.com/a/answer/81126), and Google says senders should stay well below it. You do not have to infer your complaint rate either: [Postmaster Tools](https://postmaster.google.com/) reports it directly for Gmail traffic, and [SNDS](https://sendersupport.olc.protection.outlook.com/snds/) reports complaint and trap data for the IPs you send to Microsoft from. These are also the thresholds at which a domain should be rotated out; see [when and how to rotate sending domains](https://maildeck.co/blog/cold-email-domain-rotation).

### 4. Check whether the domain was ever warmed

A new domain sending at full volume on day one looks exactly like a spam operation, because that is what spam operations do. Premium Outlook warms in 3–5 days; pre-warmed inventory skips the window entirely. If you launched cold on cold domains, the fix is time, not a vendor.

If all four come back clean and placement is still below 85%, the provider genuinely is in scope. Read on.

## Four Signals That Do Mean You Should Switch

### 1. Your placement drops when nothing on your side changed

This is the signature of a shared IP pool. On shared infrastructure your reputation includes the behaviour of every other sender on those IPs, so a stranger's bad week becomes your bad week with no corresponding change in your own list, copy or volume. Some providers rotate flagged IPs automatically, but rotation happens after the damage. If you cannot see who you share with, you cannot manage this risk — only move away from it.

### 2. Your provider cannot tell you which IP or pool you send from

A provider that cannot answer this does not have an answer you would like. Official Microsoft and Google IP pools carry inherited reputation that receiving servers already trust; a dedicated IP is yours alone but starts at zero reputation; a shared third-party pool is the one to be nervous about. Not knowing which of the three you are on is itself the finding.

### 3. Your provider publishes no per-inbox daily limit

Every inbox type has a ceiling above which spam classification rates climb. A provider that sells you inboxes without stating one is selling you the rope. If their documentation has no number in it, you have no way to know whether you are inside the envelope.

### 4. Burned domains are not replaced, or replacement is slow

At 100,000+ sends a month, 10–20% of domains burn monthly and a domain's working life under load is roughly 45 days to two months. That is normal. What is not normal is replacement that costs extra or takes longer than a few days — without warmed reserves, a replacement cycle is 7–10 days of lost sending. Keep 20–25% of your active domain count warming as reserves regardless of provider.

## The Evidence to Gather Before You Call It

Nearly every argument about whether to switch happens without data, which is why it stays an argument. Four sources settle it, all free, and together they separate your own mistake from your provider's.

| Source | What it tells you | What it proves about the provider |
|---|---|---|
| [Google Postmaster Tools](https://postmaster.google.com/) | Domain and IP reputation, spam rate and authentication pass rates for your Gmail traffic | A low spam rate with poor placement points away from your list and towards the sending infrastructure |
| [Microsoft SNDS](https://sendersupport.olc.protection.outlook.com/snds/) | Complaint rates, trap hits and filter result for the IPs you send from | If the IPs are red and your own behaviour did not change, you are carrying someone else's reputation |
| Seed or placement tests ([mail-tester](https://www.mail-tester.com/) and similar) | Where a given message lands per receiver, and the authentication and content flags on it | Isolates content and authentication from reputation — a clean score with bad placement is a reputation problem |
| Your provider's written answers | Which pools you send from, the per-inbox daily limit, the replacement policy | Silence on all three is itself the finding, and the strongest single argument for moving |

Gather all four before you make the call. The common and expensive mistake is switching on the strength of a placement drop alone: if the cause was your own volume or your own list, it moves with you, and you have added a migration to the original problem. Our [30-minute infrastructure audit](https://maildeck.co/blog/cold-email-infrastructure-audit) is the same sweep in checklist form.

## The Decision Table

| What you are seeing | Most likely cause | Fix or switch |
|---|---|---|
| Spam from day one on new domains | No warmup, or DNS misconfigured | Fix — verify DNS, then warm |
| Placement fell after you raised volume | Over the per-inbox daily limit | Fix — cut to the limit, wait 2–4 weeks |
| High bounces, replies complaining | List quality | Fix — clean the list |
| One domain burned, others fine | Normal domain burn | Fix — rotate, keep 20–25% reserves |
| Placement fell with no change on your side | Shared IP pool | Switch — to official pools or dedicated IPs |
| Everything on one infrastructure type degraded at once | Concentration risk | Switch partly — diversify, do not replace |
| Provider cannot name your IP model or daily limits | Provider opacity | Switch |
| Burned domains not replaced within days | Provider terms | Switch |

## If You Do Switch, Switch to the Right Thing

Match the destination to the failure, rather than to the price:

- Failed on a shared pool → official provider IP pools. Microsoft 365 Outlook and Google Workspace send through Microsoft's and Google's own infrastructure, which every major receiving server already trusts. Outlook runs $0.30–0.40 per inbox at 100 inboxes per tenant; Google $2.99 at 18–22 sends per inbox per day.

- Failed because volume outgrew the inbox type → either more inboxes at the same limit, or a type with a higher ceiling. Private SMTP carries 10–15 per inbox per day with a dedicated IP at $0.50 per inbox, with a 100-inbox minimum.

- Everything failed at once → this is concentration, not a vendor fault. Spread across types in roughly a 50% Outlook / 30% SMTP / 20% Google ratio; our [diversification guide](https://maildeck.co/blog/cold-email-infrastructure-diversification) explains why one provider type is never enough, and the Diversified Stack bundles start at $99/month for 40,000 sends.

- Audience mismatch → if you sell into enterprises running Microsoft and you send only from Google, or the reverse, the fix is matching infrastructure to audience, not a better vendor. See [Microsoft 365 vs Google Workspace](https://maildeck.co/blog/microsoft-365-vs-google-workspace-cold-email).

## Switching Without Compounding the Damage

A switch done badly costs more than the problem it was meant to solve, because warmup time is the one input you cannot buy back after the fact. The order matters more than the speed:

| Step | What it means | Time to allow | What happens if you skip it |
|---|---|---|---|
| 1. Fix DNS and volume first | Correct authentication and bring sends per inbox to the limit for your inbox type, on the infrastructure you already have | Same day for DNS; 2–4 weeks for volume to recover | You carry the real cause onto new inboxes and reproduce it in about three weeks, with the warmup wasted |
| 2. Run in parallel | Move a slice of volume to the new infrastructure and compare placement against the old | 2–3 weeks | You commit the whole book on an untested assumption and lose your only comparison |
| 3. Warm, or buy pre-warmed | Ramp new domains, or take pre-warmed inventory that sends from day one | 5–7 days on Outlook, 2–3 weeks on Google, 3–4 weeks on SMTP; zero if pre-warmed | New domains at full volume — the original mistake, repeated on inventory you just paid for |
| 4. Build reserves | Keep 20–25% of your active domain count warming before you need it | Continuous | The next burn costs you weeks of volume instead of a swap |
| 5. Measure placement, not opens | Seed accounts across Gmail and Outlook, plus a [mail-tester](https://www.mail-tester.com/) score of 9+ | Ongoing, weekly | You cannot tell whether the switch worked, which is how the same decision gets made twice |

Step 1 is the one teams skip, and it is the one that decides whether the switch was worth doing. Industry placement for cold email runs 75–85%; strong operations hold 90–98%. If you are still below 85% after a clean switch, the provider was never the problem — go back to the evidence and the four checks above it.

## Frequently Asked Questions

### My cold emails are going to spam — should I switch inbox provider?

Usually not. Infrastructure accounts for roughly 60% of the deliverability outcome, but the three most common causes of cold email landing in spam are fixable on the infrastructure you already have: broken DNS authentication, which 67% of the 1,000+ domains we have audited carried in some form; sending above the safe daily limit for your inbox type; and poor list quality showing up as bounces above 7% or complaints above 0.3%. Check those three before switching, because a switch carries the same problem onto new domains plus a fresh warmup window.

### How do I tell whether my inbox provider is the cause?

Four signals point at the provider rather than at you. Your inbox placement drops when nothing on your side changed, which is the signature of a shared IP pool. Your provider cannot tell you which IP or pool you send from. Your provider publishes no per-inbox daily sending limit. Or burned domains are not replaced, or replacement takes longer than a few days. If none of those apply and your DNS, volume and list all check out, the provider is probably not your problem.

### Will switching cold email providers fix my deliverability?

Only if the cause was the provider. Switching moves you to new domains that have no warmup and no reputation, so a switch that does not fix the underlying cause makes the next four to six weeks worse, not better. If you do switch, fix DNS and sending volume first on your existing infrastructure, run the new setup in parallel on a slice of volume for 2-3 weeks, and compare measured inbox placement before moving the whole programme.

### How many cold emails per inbox per day is safe?

It depends on the inbox type: Google Workspace 18-22, Microsoft 365 Outlook on a Premium licence 8-10, Microsoft 365 Outlook on a Normal licence 4-5, and private SMTP 10-15. Keep at least 61 minutes between sends from a single inbox. These limits come from MailDeck platform data across 833K+ managed inboxes. Exceeding them triggers spam classification at the provider level, and recovery takes 2-4 weeks of reduced volume - which is slower than simply staying inside the limit.

### What inbox placement rate should I expect for cold email?

Industry average inbox placement for cold email is roughly 75-85%. Strong operations run 90-98%. If you are below 85%, something is fixable. Measure it with seed accounts across Gmail and Outlook rather than inferring it from open rates, which are distorted by image-proxy prefetching, and check a Mail-Tester score of 9 or above before launch.

### Is a shared IP always bad for cold email?

No, and the distinction matters. Microsoft 365 and Google Workspace inboxes send through official Microsoft and Google IP pools, which are shared across those providers' customers but carry inherited reputation that receiving servers already trust. The risk case is a shared pool operated by a small SMTP vendor, where a handful of aggressive senders can move your reputation and you have no visibility into who they are. A dedicated IP avoids both, but starts at zero reputation and has to earn it.

### How often do cold email domains burn, and is that the provider's fault?

At 100,000+ sends per month, 10-20% of domains burn monthly and a sending domain's working life under active load is roughly 45 days to two months. That is normal and not evidence of a bad provider. What is worth switching over is how it is handled: whether replacement is free, and whether it happens in days. Without warmed reserves a replacement cycle costs 7-10 days of lost sending, so keep 20-25% of your active domain count warming regardless of who you buy from.

### How do I check whether my provider's IP addresses are the problem?

Use the two free reputation dashboards the receivers themselves publish. Google Postmaster Tools reports domain and IP reputation, spam rate and authentication pass rates for your Gmail traffic. Microsoft SNDS reports complaint rates, spam trap hits and filter results for the IPs you send to Outlook and Hotmail from. If your own spam rate is low and your authentication passes but those dashboards show poor IP reputation, the sending infrastructure is in scope rather than your list or your copy. That combination, with nothing changed on your side, is the strongest evidence for switching. If you cannot work out which IPs or pools you send from at all, that answer is itself the finding.

### What spam complaint rate is too high for cold email?

Google's email sender guidelines set the threshold at 0.3% and advise senders to stay well below it, so treat 0.3% as a hard ceiling rather than a target. Bounces should stay under 7%. Above either figure your list is burning your domains and changing provider will not help, because the complaint and bounce behaviour travels with the list. Both numbers are visible to you: Postmaster Tools reports the Gmail spam rate directly, and your sequencer reports bounces. Fix the list first, then re-measure placement before deciding anything about the provider.

## Methodology

Figures come from MailDeck platform data: 833K+ managed inboxes, 3,000+ domains, 1,631+ clients, 50M+ emails delivered monthly, with safe sends per inbox per day being the limits we enforce across all three infrastructure types. The DNS error breakdown comes from an audit of 1,000+ client cold email domains. Burn rates, domain lifespan and recovery windows are aggregate platform figures; individual results vary with list quality, copy and target audience. Industry inbox placement benchmarks are external and approximate.

Last updated: October 2026

Written by [Nikita Stoletov](https://maildeck.co/authors/nikita-stoletov)

CTO | Outbound Infrastructure & Deliverability Engineering

Built the infrastructure delivering 50M+ emails monthly across 3,000+ domains with 98% inbox placement at MailDeck.

[LinkedIn](https://www.linkedin.com/in/nikita-stoletov-30b817292)

[View Profile](https://maildeck.co/authors/nikita-stoletov)

## Ready to Scale Your Email Infrastructure?

Join top outbound teams using MailDeck for enterprise-grade deliverability.

[Get Started](https://app.maildeck.co)

## Continue Reading

[Guide](https://maildeck.co/blog/cold-email-going-to-spam-fix)

### [Why Your Cold Emails Go to Spam (and How to Fix It) [2026 Data]](https://maildeck.co/blog/cold-email-going-to-spam-fix)

[The full fix guide once you know the cause is on your side rather than your provider's.](https://maildeck.co/blog/cold-email-going-to-spam-fix)

[Guide](https://maildeck.co/blog/cold-email-deliverability-checklist)

### [Cold Email Deliverability Checklist: 15-Point Audit Template](https://maildeck.co/blog/cold-email-deliverability-checklist)

[Run this before every campaign launch, with a pass threshold for each of the fifteen checks.](https://maildeck.co/blog/cold-email-deliverability-checklist)

[Guide](https://maildeck.co/blog/cold-email-infrastructure-audit)

### [How to Audit Your Cold Email Infrastructure in 30 Minutes](https://maildeck.co/blog/cold-email-infrastructure-audit)

[A structured pass over DNS, inbox types, limits and domain health to find what is actually broken.](https://maildeck.co/blog/cold-email-infrastructure-audit)