Cold Email Deliverability: What Decides Inbox Placement
Cold email deliverability depends on authentication, complaint rates, list quality and sending habits. Learn the thresholds that keep emails in the inbox.
A user-reported spam rate of 0.3% can make a bulk sender ineligible for Gmail delivery mitigations, according to Google's sender requirements FAQ. That's why cold email deliverability isn't a question of getting a server to accept your message. You need to verify the sender, measure where accepted messages land, and respond when recipients signal that the outreach is unwanted.
For B2B teams selling across markets, the practical challenge is connecting those checks. A sequencing dashboard can report successful delivery while recipients never see the message. A technically authenticated campaign can still generate complaints. No DNS setting, warm-up tool, or sending schedule guarantees the inbox.
Table of Contents
Six Factors That Decide Cold Email Deliverability
Factor | What it controls | Benchmark or check |
|---|---|---|
Domain authentication | Whether mailbox providers can verify the sender | SPF or DKIM at minimum; DMARC published for bulk senders |
Sender reputation | Trust built from past mail performance | Spam complaints below 0.1%, never 0.3% or higher |
List quality | Whether contacts are real, current and relevant | Verified, deduplicated contacts; bounces reviewed before every send |
Sending practices | Volume, pace and cadence per sending domain | Gradual ramp; no sudden volume spikes |
Opt-out handling | Whether recipients can leave easily | One-click unsubscribe; requests honored within 48 hours |
Monitoring | Whether problems are caught early | Complaint, bounce and alignment checks per domain, reviewed daily |
These benchmarks combine provider requirements with operating checks. Authentication and unsubscribe obligations are compliance controls; verification, gradual ramping, and daily reviews are practical safeguards. They aren't equivalent guarantees, and passing the checklist doesn't establish that a campaign reached its intended audience.
Authentication establishes identity, not permission or relevance. Providers can verify who sent a message and still decide not to put it in the inbox. Sender history, recipient responses, and campaign behavior remain separate considerations. Our guide to domain name reputation helps distinguish that accumulated history from the records used to authenticate a particular message.
Treat the factors as connected controls
List quality and opt-out handling affect what recipients experience. If your team keeps contacting people who have left their roles, don't fit the offer, or have already declined, adjusting the sending schedule won't resolve the underlying problem. Likewise, excellent targeting won't repair a broken authentication setup.
For tooling, Instantly is one option among several for warm-up and sending. Treat it as support for part of the workflow, not as evidence that every message reaches an inbox. Someone still needs to own targeting, suppression, authentication checks, and campaign decisions.
The Social Search has sent 500K+ cold emails. That credential establishes operational exposure, not a universal placement promise. The useful discipline is to audit these factors in order, starting with authentication, because mailbox providers need to verify the identity before your team can meaningfully assess its performance.
Practical rule: Fix the earliest broken control before optimizing a later one. Better copy can't compensate for a sender identity that fails verification.
Three Layers Behind Every Cold Email Result
The bounce-rate assumption is simple and misleading: if an email didn't bounce, it reached the prospect. In a 2025 dataset of 7.53 million B2B cold emails, a 1.71% bounce rate left 98.29% labeled deliverable, while independent global benchmark reporting put inbox placement at 83.5%, spam placement at 6.7%, and missing mail at 9.8%, as summarized in Belkins' email deliverability analysis.
Those are different measurements, not directly comparable campaign results. The first describes messages that weren't rejected at the SMTP stage. The placement benchmark describes where messages ended up. Subtracting one benchmark from the other wouldn't reveal your campaign's hidden losses.
A courier accepting a parcel at the door isn't the same as the parcel reaching the right desk. Even reaching that desk doesn't guarantee the recipient opens it or acts.

Match each question to its evidence
Layer | Question | Evidence to inspect |
|---|---|---|
Technical acceptance | Did the receiving server accept the message? | SMTP responses, bounce logs, rejection reasons |
Inbox placement | Did the message reach an inbox rather than spam or another filtered destination? | Seeded inbox tests and provider-level telemetry |
Engagement and conversion | Did the recipient respond, qualify, or progress commercially? | Classified replies, meetings, and CRM outcomes |
A positive reply is evidence of engagement, but silence doesn't identify the failure. The recipient might not have seen the message, might not need the offer, or might have chosen not to answer. Don't diagnose a placement problem from reply rate alone.
Run a controlled placement check
Use representative seed accounts. Include Gmail, Yahoo, and Microsoft accounts relevant to your audience.
Send the campaign message through the actual sending inbox and workflow, rather than a separate manual test.
Record whether each copy lands in the inbox, spam, or a categorized tab such as Promotions. Keep those destinations separate.
Compare the same campaign's evidence. Inspect bounce logs and reply data alongside the placement results, segmented by receiving provider.
Seed accounts are diagnostic instruments, not a representative sample of every buyer's inbox. Use them to identify patterns and test a suspected fault. Combine their results with complaint signals and Google Postmaster data where available before deciding whether the issue is technical, provider-specific, or commercial.
Gmail and Yahoo Rules and Complaint Thresholds
Google introduced mandatory sender requirements for mail sent to Gmail accounts on February 1, 2024, with additional obligations for senders sending approximately 5,000 or more messages daily to Gmail recipients, according to Google's email sender guidelines.
The distinction matters: basic sender requirements aren't reserved for large marketing operations. Bulk senders face additional authentication, alignment, and unsubscribe obligations. Labeling a campaign sales prospecting rather than marketing doesn't remove applicable provider requirements.
What providers expect from senders
For Gmail, all senders need authentication, while bulk senders need both SPF and DKIM, published DMARC, and alignment between the visible sender identity and an authenticated domain. Promotional bulk messages also need a one-click unsubscribe mechanism.
Under Yahoo's sender best practices, all senders need SPF or DKIM; bulk senders need both, a valid DMARC policy, aligned sender identity, a functioning List-Unsubscribe header, a visible body unsubscribe option, and unsubscribe requests honored within two days. Yahoo also requires spam complaints to remain below 0.3%.
A footer asking recipients to reply with an opt-out request isn't the same implementation as a functioning unsubscribe header. Test the header and the visible link separately. Then confirm that each route updates the suppression system used by every sending workflow.
Complaints are a daily operating signal
Google's bulk-sender FAQ recommends keeping user-reported spam below 0.1%, identifies 0.3% as the critical boundary, and describes mitigation eligibility returning after seven consecutive days below 0.3%.
Don't infer complaint health from opens or positive replies. A campaign can produce useful conversations while another group of recipients reports it as spam. Review complaints by sending domain, mailbox provider, campaign, and sending identity so an aggregate average doesn't conceal a damaged segment.
Yahoo calculates complaint rate using messages delivered to the inbox rather than total messages sent. That denominator matters when interpreting reports: a rate you calculate from your sequencing platform's sent count may not match the provider's assessment. Keep provider-reported measurements distinct from internally calculated campaign ratios.
A recipient who finds your message irrelevant doesn't care whether SPF passed. Authentication verifies the sender; it doesn't validate the targeting decision.
Trace complaints back to their cause
Start with an agreed ideal customer profile, including the accounts and buyer roles worth contacting. Add hiring, funding, technology, or growth signals only when they explain why the offer is relevant now. Whitewhale is a recommended option for flagging accounts worth reviewing, but an intent signal still needs human interpretation.
Verify contacts before sending, deduplicate across lists, and remove bounced or unsubscribed addresses. Suppress opt-outs across every sending domain, not just the sequence in which the request arrived. Otherwise, a recipient can decline once and receive another version of the same outreach from your team.
Use an email delivery report to connect symptoms with decisions. If complaints concentrate in a particular audience segment, review fit and frequency before rewriting the subject line. If opt-outs continue receiving messages, repair suppression before launching another campaign.
Setting Up SPF, DKIM, DMARC and PTR Records
Record | Purpose | Requirement |
|---|---|---|
SPF | Lists servers allowed to send for the domain | Publish for every sending domain as an operating standard |
DKIM | Signs messages with a domain key | Unique key per sending domain; required for bulk senders |
DMARC | Defines handling for failing mail and enables reports | Bulk senders need at least a monitoring policy; enable aggregate reporting |
PTR, reverse DNS | Connects the sending IP to a hostname | Valid forward and reverse records |
TLS | Encrypts mail in transit | Required for Gmail senders; a transport setting, not a DNS record |
Google's authentication and infrastructure requirements require at least SPF or DKIM for senders, both for bulk senders, valid forward and reverse DNS, and TLS; bulk domains must publish DMARC with at least p=none. Use the table as a deployment standard, while recognizing that some rows describe stronger operating practice than the minimum for smaller senders.
Check alignment rather than record existence
Say your visible From address uses your company's domain, but a sequencing vendor authenticates SPF using its own unrelated envelope domain. SPF can pass without aligning to the identity the recipient sees. If there's no aligned DKIM signature either, DMARC can fail.
That's why checking whether a DNS record exists isn't enough. Inspect the authentication results on an actual received message, including the domains used for SPF and DKIM. A green status indicator in a vendor dashboard doesn't replace that evidence.
Deploy in a controlled sequence
Publish one SPF record per sending domain. Include authorized sending services so your allowed sources are explicit. Don't create competing SPF records when adding another vendor.
Generate a unique DKIM key for each domain. Separate signing identities make troubleshooting and key management more controlled.
Enable DMARC monitoring and aggregate reporting. Use the reports to identify legitimate services and unexpected senders before enforcing a stricter policy.
Confirm that PTR and forward DNS correspond correctly. Your mailbox or infrastructure provider may control these records, so verify the configuration through that provider.
Send real test messages from each legitimate workflow. Inspect SPF, DKIM, and DMARC results, not just a DNS lookup.
Read the aggregate reports. Investigate known senders that fail alignment and unfamiliar sources using your domain.
Tighten the DMARC policy only after CRM, sequencing, and transactional sources are accounted for. Otherwise, you risk disrupting legitimate mail along with unauthorized traffic.
Keep a sender inventory alongside the configuration: service owner, sending domain, authentication method, and last successful test. Recheck after switching vendors or changing DNS, because an initially correct deployment doesn't establish that later changes are safe.
For another practical reference, Steeve Elgui's email deliverability guide can support your implementation review. Pair that with a documented cold email infrastructure plan so the next operator knows which services must remain authorized.
Sending Practices That Keep Emails Out of Spam
Use a staged ramp as a controlled test, not a countdown to unrestricted sending. Start with a small group of verified contacts whose relevance you can explain. If these are cold prospects, don't label them engaged merely because they fit your ideal customer profile.
The progression should follow observable campaign health rather than a fixed calendar. Warm new domains and inboxes gradually, but don't treat automated warm-up activity as proof of real-recipient placement.

Increase only when the evidence supports it
Begin with the test cohort, inspect acceptance and placement, then increase volume in measured steps while watching complaints and bounces. Hold a stable plateau long enough to distinguish repeatable performance from an isolated good send. Add new domains only after the existing operation is stable, not as a way to escape unresolved reputation problems.
If the next list produces more bounces, investigate its source and verification status. If complaints concentrate after later follow-ups, inspect the sequence's frequency and usefulness. These are different problems and shouldn't trigger the same automatic response.
Sending discipline: Increase volume because the current process is healthy, not because the sales target requires more activity.
Make every follow-up earn its place
Set a finite follow-up limit before launch. Stop the sequence when a recipient replies or unsubscribes, and route replies to a person promptly. Test that stop logic across connected tools so a CRM delay doesn't leave someone receiving automated follow-ups after answering.
Personalization should explain relevance, not just demonstrate that you know a name or job title. A hiring signal might justify a message about onboarding capacity; a generic congratulations line doesn't establish a reason to buy. Avoid blasting identical copy across many inboxes, but don't assume cosmetic variation makes unwanted outreach useful.
Use EmailScout's spam filter avoidance guide as a supplementary campaign review resource, not a placement guarantee. Before importing the next audience, follow a documented process for cleaning an email list.
Avoid these preventable operating mistakes:
Sudden volume spikes: Don't change the sending load before reviewing the current cohort's results.
Identical mass distribution: Reassess account relevance instead of multiplying copies across inboxes.
Unverified lists: Resolve questionable records before they enter a sequence.
Ignored bounce reports: Investigate rejection reasons rather than repeatedly retrying the same audience.
Set Pause Rules and Audit Your Setup Before Sending
Metric | Healthy | Warning | Action |
|---|---|---|---|
Spam complaint rate | Below 0.1% | 0.1% to below 0.3% | At 0.3% or above, pause the sequence and investigate |
Bounce rate | Stable and low | Rising after a list import | Pause and re-verify the list |
DMARC alignment | Legitimate sources pass | Failures from a known sender | Fix configuration before the next send |
Unsubscribe processing | Within 48 hours | Delayed beyond 48 hours | Check the unsubscribe flow and suppression sync |
Reply rate | Stable against baseline | Drop with flat volume | Run seeded inbox tests before changing copy |
The complaint and unsubscribe figures reflect the provider standards described above. The pause actions are operating rules, not claims that a provider automatically blocks every campaign at the same moment. Stop the affected sequence when complaints reach the critical boundary, and investigate the warning band before continuing to increase volume.
Don't invent a universal bounce or reply threshold to fill the gaps. Compare each campaign with its own relevant baseline, including audience source, receiving provider, and sending identity. A sudden change deserves investigation even when the account-wide average looks normal.
Run the pre-send audit
Inspect a real message. Confirm SPF, DKIM, and DMARC pass using the production sending workflow. Save the headers so another operator can reproduce the check.
Confirm PTR and forward DNS match through the infrastructure provider. Record who owns the configuration and who can repair it.
Test both unsubscribe routes. Check the header mechanism and visible body link, then verify that the address enters suppression.
Verify and deduplicate the audience. Remove bounces and suppressed contacts across all domains, including lists already scheduled in other campaigns.
Compare daily sending volume per domain with the approved ramp. Check scheduled follow-ups as well as new contacts, since both contribute to the workload.
Configure segmented complaint alerts. Assign an owner for each domain, provider, and campaign, with authority to pause sending rather than merely report a problem.
A repeatable email deliverability test should produce evidence, not just a pass label. Retain the test message, destination, folder placement, authentication result, and decision taken. That record helps distinguish a new configuration failure from a change in audience quality.
Repair the foundation before resuming
Fix authentication first, then the list, then volume. Authentication faults undermine the validity of later placement tests; an unreliable list makes a sending adjustment difficult to evaluate. Once those foundations are sound, change the pace cautiously and observe the result before making another adjustment.
Assign suppression ownership separately from campaign ownership if several teams send mail. Someone must be accountable for removing an opted-out contact everywhere, including workflows outside the original campaign. A sequence owner who can pause only their own inbox can't resolve a company-wide suppression failure.
Teams can build this internally or hand off implementation. Our email infrastructure service covers domains, inboxes, SPF/DKIM/DMARC setup, and warm-up from the ground up; our outbound work combines signal-led email with LinkedIn outreach, built with Whitewhale and HeyReach. These are implementation options, not promises of inbox placement or commercial results.
The Social Search builds client-owned email infrastructure and connected outbound systems, with documented handover or ongoing operations. Visit The Social Search to discuss your sending setup, targeting, and monitoring responsibilities before scaling the next campaign.
