ISP-complaint rate. Above 0.1% Gmail starts marking sender domain.
At a glance
The share of delivered emails that recipients flagged “this is spam” through their mail client. The single most damaging deliverability metric, because the receiving ISPs (Gmail, Outlook, Yahoo) feed it directly into their inbound-filtering decisions. Above 0.10% Dotdigital’s compliance team will pause sending; above 0.30% Gmail starts routing your sends straight to the spam folder regardless of content quality. The healthy band is under 0.05% for established lists; under 0.10% during a re-engagement push.
| What it counts | numComplaints / numDelivered per send, rolled up across the period. Numerator is the count of feedback-loop (FBL) complaints reported by ISPs back to Dotdigital’s MTA. |
| Receiving ISPs covered | Gmail (Postmaster Tools FBL), Yahoo, AOL, Outlook (Smart Network Data Services), Comcast, Cox, and the major UK and EU ISPs. Apple iCloud Mail does not publish a feedback loop, so iCloud “Mark as Junk” actions do not appear here. This is a real reporting gap and means the true complaint rate is structurally a touch higher than reported. |
| Definition of “complaint” | A user-initiated “Mark as Spam” or “Report Junk” click. Auto-filtering by the mail client (Gmail’s spam tab, Outlook’s Focused inbox routing) does NOT count, only deliberate user action. |
| Reporting lag | 6 to 24 hours typical. Some ISPs batch FBL reports daily, so today’s complaint number for “today” is provisional. |
| Complaint vs unsubscribe | A complaint is a hostile signal (“this email shouldn’t have reached me”); an unsubscribe is a polite one (“I don’t want this anymore”). Dotdigital’s deliverability team treats 1 complaint as roughly equivalent to 25 unsubscribes for IP-reputation purposes. |
| Compliance threshold | Dotdigital’s internal threshold is 0.10% over a rolling 7-day window. Crossing triggers an automated notification and may pause sending. Above 0.30% Gmail will spam-folder the sending domain at the inbox-placement layer. |
| What drives complaints | (1) sending to a list the recipient doesn’t remember signing up for, (2) too-frequent sending to the same contact, (3) a misleading subject line, (4) hard-to-find unsubscribe link (CAN-SPAM and PECR violation in UK and EU). |
| Auto-suppression | Every complaint auto-suppresses the contact (they will not receive future sends from this account). This is required by ISP feedback-loop terms; you can see the action on Suppressed Contacts. |
| GDPR scope | UK and EU recipients are particularly quick to mark as spam if consent isn’t fresh; PECR enforcement is real, complaints to the ICO are rising YoY. |
| Time window | 30D vsP |
| Alert trigger | >0.1%, drives a gauge_inverse with good <0.1%, warn 0.1 to 0.3%, bad >0.3% |
| Roles | owner, marketing, engineering |
Calculation
Calculated automatically from your Dotdigital data. See the At a glance summary above for what the metric tracks and the worked example below for a typical reading.Worked example
A UK B2C subscription-box brand running on Shopify with Dotdigital. Active list 86,000 contacts. The 30-day window covers 03 Apr 26 to 02 May 26.| Metric | Value |
|---|---|
| Total delivered sends | 412,000 |
| Spam complaints | 188 |
| Spam complaint rate (this card) | 0.046% |
| Same period unsubscribes | 1,480 (0.36%) |
- 0.046% is healthy, well inside the under-0.05% band for established UK B2C senders. The brand has been on Dotdigital for 3 years with consistent re-permissioning every 18 months.
- Complaints came overwhelmingly from one campaign. A 24 Apr 26 win-back send to “Inactive 6m+” (12,400 contacts) generated 92 of the 188 complaints. That one campaign’s complaint rate was 0.74%, more than 7x the warning threshold. The brand should not have sent it; the segment needed a re-permissioning email first.
- Unsubscribes were 8x complaints at 1,480 vs 188. This is the healthy ratio. When complaints approach or exceed unsubscribes, the list is hostile (spam complaints rising, polite opt-outs falling) and sending should pause for a deliverability reset. The 8:1 ratio means most disengaged contacts are choosing the polite exit, which is what well-placed unsubscribe links produce.
- iCloud Mail blind spot. Roughly 18% of the brand’s UK list uses Apple Mail with
@icloud.comor@me.comaddresses. Apple does not publish FBL data, so the true rate is probably 0.055 to 0.060%; still healthy. For US-heavy lists this gap is smaller (Gmail dominates US consumer email). - The Boxing Day 2025 send (run with a previous CSV import that hadn’t been re-permissioned) hit 0.31% complaint rate on a single day. The brand received the Dotdigital compliance email within 4 hours, paused the campaign mid-send, and ran a re-permissioning sweep before resuming. That experience is why the current 0.046% is sustained, the brand learned to pre-validate every win-back segment.
Sibling cards merchants should reference together
| Card | Why pair it with Spam Complaint Rate |
|---|---|
| Dotdigital Bounce Rate | The other deliverability red-line. A spike in both at once means the list itself is the problem, usually a stale CSV upload. |
| Dotdigital Unsubscribe Rate | The polite-exit signal. Healthy ratio: unsubscribes are 5 to 10x complaints. When they invert, the list is hostile. |
| Sender Reputation alert | The downstream effect. A sustained complaint rate above 0.10% degrades sender reputation with a 24 to 72h lag, which is what Sender Reputation tracks. |
| Dotdigital Suppressed Contacts | Every complaint is auto-suppressed. The cumulative count tells you whether complaint volume is structural or one-off. |
| Dotdigital Send Cadence | The most common cause of rising complaints is sending too frequently to the same contacts. This card surfaces over-sending. |
| Dotdigital Open Rate | A campaign with low opens AND high complaints is the classic “wrong audience” pattern, the segment didn’t expect to hear from you. |
| Klaviyo Spam Complaint Rate | The closest cross-platform peer, same definition same threshold. |
| Mailchimp Spam Complaint Rate | Same definition, slightly stricter Mailchimp threshold (0.08% vs Dotdigital’s 0.10%). |
Reconciling against the vendor’s own dashboard
Where to look in Dotdigital: r1-app.dotdigital.com → Insights → Email Health → Complaints for the rolling complaint view; Account Settings → Email Sending Reputation for the IP-pool view that drives compliance decisions; per-campaign forensic detail at Campaigns → Reports → Detailed Statistics → Complaints. The Postmaster-side view (Gmail’s perspective on your sending domain) is at postmaster.google.com; add your sending domain and you’ll see Gmail’s view of your complaint rate, which can differ from Dotdigital’s because Gmail uses a different denominator (delivered to inbox vs delivered to spam folder). Why our number may legitimately differ from Dotdigital’s dashboard:| Reason | Direction of divergence |
|---|---|
| FBL reporting lag. ISPs batch complaints daily; some by 24h. Vortex IQ’s “today” reads will lag Dotdigital’s “today” by a similar window. | Vortex IQ slightly lower for the most recent day |
| iCloud reporting gap. Apple does not publish a feedback loop. Both Vortex IQ and Dotdigital miss iCloud complaints; the figure understates the true rate by ~10 to 25% on iCloud-heavy UK lists. | Both equally understated; consistent gap |
| Time-zone. Dotdigital reports run on account locale; Vortex IQ on UTC. Boundary days will differ. | ±0.005 to 0.02 percentage points at the boundary |
| Region pod. Multi-pod accounts only count the connected pod. | Vortex IQ lower if other pods exist |
Postmaster vs FBL. Gmail Postmaster Tools uses a slightly different denominator (inbox-placed only); Dotdigital uses raw FBL count divided by numDelivered. The two will not match exactly. | Gmail Postmaster typically lower |
| Card | Expected relationship | What causes legitimate divergence |
|---|---|---|
klaviyo.klv_spam_rate | Same definition, comparable rate | Klaviyo’s threshold is also 0.10%; ratios should match within 0.02 percentage points on the same list. |
mailchimp.mc_spam_rate | Same shape, stricter Mailchimp threshold (0.08%) | Mailchimp throttles 0.02 percentage points earlier. |
brevo_sendinblue.bs_spam_rate | Same definition | Brevo also UK and EU shaped, similar GDPR profile. |
Known limitations / merchant FAQs
Why is 0.10% the threshold? It looks like a tiny number. At Dotdigital scale (and most ESPs) 0.10% means 1 complaint per 1,000 sends. Across a list sending 500,000 emails per month that’s 500 complaints, which is roughly the volume that Gmail’s spam-classifier needs to start de-prioritising the sender domain. The threshold isn’t arbitrary; it’s the level at which the ISP’s machine-learning classifier flips its decision on inbox placement. At 0.30% Gmail will not just de-prioritise, it will actively spam-folder you regardless of content quality. One bad campaign sent my rate to 0.20%. Will it recover? Yes, gradually. The complaint rate is a 30-day rolling number on this card. A single bad campaign with 0.50% complaint rate will average out as more healthy sends accumulate. The repair takes 2 to 4 weeks of normal sending to bring the rolling rate back under 0.10%. Don’t try to “send through it”, the right move is to pause promotional sending for 1 week, run only transactional and high-engagement-segment sends, then resume gradually. Sender Reputation tracks the recovery curve. Why is my rate higher on Yahoo and Outlook than on Gmail? Yahoo and Outlook users tend to be older, less email-literate, and more likely to “Mark as Spam” instead of unsubscribe. Gmail users tend to use the unsubscribe link. The pattern is structural and shows up across every ESP. Mitigation: make sure your unsubscribe link is large, obvious, and at the top of the email (not just the footer), and consider implementing list-unsubscribe headers (which Dotdigital adds automatically for accounts on shared IP pools). My rate is 0.04% on Vortex IQ but 0.06% on Dotdigital’s own dashboard. Why? Almost certainly time-zone. Vortex IQ runs UTC by default; Dotdigital runs account locale. The 30-day windows differ by up to 24 hours at each end. A high-complaint campaign that fell on the boundary day will be in one window and not the other. The trend will reconcile after a few days. What’s “list-unsubscribe” and does Dotdigital handle it for me? List-unsubscribe is an email header (List-Unsubscribe-Post, List-Unsubscribe) that lets the mail client display a one-click “Unsubscribe” button at the top of the message. Gmail and Apple Mail show this prominently. Dotdigital adds these headers automatically for all sends on shared IP pools; for dedicated-IP customers it’s an account setting. The benefit is large: contacts who would have hit “Mark as Spam” instead hit one-click unsubscribe, which is a polite signal that doesn’t damage sender reputation.
Is iCloud Mail’s “Mark as Junk” actually invisible?
Yes, and this is widely under-appreciated. Apple’s iCloud (covering @icloud.com, @me.com, @mac.com, and Apple’s relay-anonymous addresses) does not publish a feedback loop. So when a UK list has 18 to 25% iCloud users, the true complaint rate is structurally a touch higher than what any ESP can report. The gap is consistent over time, so trends are still useful, but absolute numbers under-state.
Do unsubscribes count toward my complaint rate?
No. Unsubscribes are a separate metric (Dotdigital Unsubscribe Rate). They reduce your sendable list size but do not damage IP reputation. The ratio of unsubscribes to complaints is itself a useful signal: healthy is 5 to 10x more unsubscribes than complaints; at parity or worse, the list is hostile.
My account just received an automated compliance email from Dotdigital. What now?
Don’t panic, this is the standard early-warning. The email will list: which campaigns spiked complaints, which ISPs are flagging, suggested remediation. Standard playbook: (1) pause all non-transactional sending for 24 to 72h, (2) run a list-cleaning sweep on segments older than 6 months, (3) re-permission inactive contacts via a dedicated re-permissioning email, (4) resume promotional sending at half the prior cadence and ramp back up over 2 weeks.
Should I move to a dedicated IP to fix this?
Probably not. A dedicated IP is useful for senders above 500k/month who want full control of reputation. For accounts under 500k/month, a dedicated IP often hurts deliverability because there isn’t enough volume to maintain warm reputation; the shared pool’s volume is what keeps you in the inbox. If your shared-pool reputation is bad, fix the list, don’t move IPs.