Most ecommerce email marketers think about inbox placement as a binary — either your email arrives or it doesn’t.
The reality has three destinations, and for ecommerce the interesting contest is not Inbox versus Spam. It’s Promotions versus Spam. Promotions is where customers actively go looking for deals; spam is where campaigns go to die.
We analysed 671,329 monitored messages from roughly 19,400 ecommerce sending domains in Q1 2026 and tracked which of Gmail’s three destinations each one reached. The headline finding isn’t about any single lever. It’s about shape: at the brand level, Gmail deliverability is close to binary, and the comfortable middle most senders assume they’re sitting in barely exists.
Gmail Placement — Q1 2026, senders with 100+ monitored messages
What This Study Measures — And What It Doesn’t
This matters more than usual here, so it goes first.
InboxEagle runs inbox placement monitoring. We observe messages arriving at monitored Gmail mailboxes and record which destination Gmail chose: Primary, Promotions, or Spam. Every number in this article is derived from that observation.
We do not measure spam complaints. Complaint rate — the share of recipients who press “Report spam” — is reported to senders through Google Postmaster Tools, aggregated per domain. It is not visible in placement monitoring, and nothing in this dataset stands in for it.
Two further limits worth stating plainly:
This is a monitored panel, not a census. These are domains under active deliverability monitoring, which skews the sample toward brands that already suspect they have a problem. Panel-wide spam placement of around 18–21% is far above what a representative sample of all ecommerce senders would show. Read this as a distribution, not as an industry benchmark.
Small senders distort brand-level percentages. Of ~19,400 domains in the raw export, 8,241 had fewer than 10 monitored messages in the quarter. A domain with one monitored message that happened to be filtered shows “100% spam placement” — which is noise, not a signal. In the unfiltered data, 1,807 domains sat at exactly 100% spam placement and 1,545 of those had fewer than 10 messages.
So the analysis below uses a floor of 100 monitored messages per sender: 1,354 senders and 318,521 messages. Where the unfiltered numbers differ materially, we say so.
Deliverability Is Bimodal at the Brand Level
Here is the distribution that the aggregate numbers hide. Each sender is placed by the share of its own mail that Gmail filtered to spam.
| Sender’s spam placement | Senders | Share of senders |
|---|---|---|
| Under 1% | 507 | 37.4% |
| 1–5% | 194 | 14.3% |
| 5–25% | 284 | 21.0% |
| 25–50% | 141 | 10.4% |
| 50–75% | 138 | 10.2% |
| Over 75% | 90 | 6.6% |
Read the ends against the middle. 37.4% of senders are essentially clean — under one message in a hundred filtered. 27.2% are in serious trouble — more than a quarter of their mail filtered, including 6.6% who are functionally unable to reach Gmail at all. Between them, the 1–5% band holds just 14.3%.
That shape is the actual finding. Gmail’s treatment of a sending domain tends to settle into one of two stable states — broadly trusted or broadly filtered — with comparatively few senders holding a middling position for a whole quarter. Deliverability behaves less like a dial you drift along and more like a state you occupy.
For a sender, the practical implication is uncomfortable: there probably isn’t a gentle warning zone. The population that should be your early-warning cohort is the smallest group in the data.
The Numbers, By Sender Bucket
| Sender’s spam placement | Senders | Messages | Primary | Promotions | Spam |
|---|---|---|---|---|---|
| 0–1% | 507 | 123,151 | 24.13% | 75.75% | 0.11% |
| 1–3% | 102 | 23,924 | 27.19% | 71.04% | 1.77% |
| 3–5% | 92 | 23,813 | 22.64% | 73.46% | 3.90% |
| 5%+ | 653 | 147,633 | 18.93% | 42.49% | 38.58% |
A warning about how to read this table. Senders are grouped by their spam placement, so the Spam column largely restates the grouping — of course the 5%+ bucket has high spam placement; that is what put those senders in the bucket. That column is definitional, not a discovery, and any version of this analysis that presents it as a finding is arguing in a circle.
The genuinely informative columns are Primary and Promotions, because nothing about the grouping determines how the surviving mail divides between them.
And the top bucket is unbounded: it runs from 5% to 100% spam placement and contains 653 senders, many of them near-total filtering. The 38.58% average is a property of that open-ended range, not a cliff edge at the 5% boundary.
What Happens to the Mail That Gets Through
This is the part the bucketing doesn’t predetermine. Of the messages that escaped the spam folder, how did Gmail split them between Primary and Promotions?
| Sender’s spam placement | Primary share of delivered mail | Promotions share |
|---|---|---|
| 0–1% | 24.16% | 75.84% |
| 1–3% | 27.68% | 72.32% |
| 3–5% | 23.56% | 76.44% |
| 5%+ | 30.82% | 69.18% |
The Primary share drifts upward — from 24.16% to 30.82% — as spam placement rises. That is mildly counterintuitive, and it is worth being careful about what it means.
It is not a benefit. Absolute volume reaching any inbox destination collapses in the 5%+ group; a larger slice of a much smaller pie is still less mail. The likeliest explanation is compositional rather than causal: heavily filtered domains send a different mix of mail, with a higher proportion of transactional and one-to-one messages that Gmail routes to Primary, while their bulk promotional sends are the ones being filtered out of the sample. This dataset can’t separate those two effects, so treat the pattern as an observation that needs a follow-up study, not a mechanism.
What the table does establish is that Promotions placement is remarkably stable — between 69% and 76% of delivered mail across every bucket. For ecommerce senders, Promotions is the normal destination. Chasing the Primary inbox is largely chasing a category Gmail has decided your mail doesn’t belong to.
Spam Placement Is Not Spam Complaint Rate
The two get conflated constantly, so it’s worth separating carefully.
Spam complaint rate is the share of delivered messages that recipients actively report. Google publishes it per domain in Postmaster Tools. Google’s sender guidelines ask bulk senders to keep it below 0.10% and to never reach 0.30%. As of Google’s November 2025 enforcement ramp, non-compliant traffic gets real SMTP rejection codes rather than silent filtering.
Spam placement rate is the share of messages the filter files into spam, whether or not anyone reports anything. That’s what this study measures.
The scales are not comparable. A 1% complaint rate would be more than three times Google’s hard ceiling and a genuine emergency. A 1% placement rate, by contrast, puts a sender in the healthiest group in this dataset. Applying Google’s 0.10%/0.30% thresholds to placement numbers makes clean senders look reckless and reckless senders look merely untidy.
The metrics are related. Complaints are one input to the reputation models that drive filtering, so sustained complaints will eventually show up as placement damage. But they are measured differently, sourced differently, and sit on different scales, and one cannot be substituted for the other.
If you want the complaint-side thresholds and what triggers enforcement, that’s covered in our Gmail spam rate threshold guide.
What This Means For Your Sending
The distribution says the goal is to stay in the clean state rather than to manage a slow decline, because the data suggests the slow decline is not where most senders actually live.
Watch complaints and placement as separate instruments. Postmaster Tools gives you the complaint signal against Google’s 0.10% line. Placement monitoring tells you where mail is actually landing. Neither one substitutes for the other, and this article is a case study in what goes wrong when you treat them as interchangeable. Our spam rate monitoring guide covers the Gmail and Yahoo thresholds.
Make unsubscribing easier than reporting. One-click unsubscribe (RFC 8058) is a Google requirement for bulk senders, and a subscriber who can’t find the exit is a subscriber who reaches for “Report spam” instead. The unsubscribe costs you a list entry; the complaint costs you reputation.
Segment before you send. The 27.2% of senders with over a quarter of mail filtered are, overwhelmingly, senders mailing their full list regardless of engagement. Suppress or re-engage long-dormant subscribers rather than including them in campaigns.
Treat authentication as the floor, not the strategy. SPF, DKIM and DMARC alignment is a prerequisite. It won’t lift you into Promotions on its own, but its absence caps everything else.
The Bottom Line
Across 1,354 monitored ecommerce senders and 318,521 messages, Gmail delivered 59.81% to Promotions, 21.84% to Primary and filtered 18.35% to spam.
But the aggregate is the least interesting number here. The distribution behind it is close to bimodal — 37.4% of senders nearly untouched by filtering, 27.2% heavily filtered, and a thin band in between. Promotions placement among delivered mail holds steady at roughly 70–76% regardless of which group a sender is in.
The question worth asking about your own programme isn’t “how far along the slide am I?” On this evidence, that framing doesn’t match how senders are actually distributed. The question is which of the two states you’re in — and placement monitoring, not complaint rate alone, is what tells you.
Want to know where your campaigns are actually landing? Run a free spam test with InboxEagle — no credit card required.
For the data on how authentication failures affect spam placement, see our DMARC failure study — 2.2M emails analysed.
Methodology: Placement data from InboxEagle’s inbox placement monitoring, Q1 2026 (January–March), 671,329 monitored messages across approximately 19,400 ecommerce sending domains. Primary analysis restricted to the 1,354 domains with at least 100 monitored messages (318,521 messages) to avoid small-sample distortion. Destinations recorded as observed by monitored Gmail mailboxes. This is a monitored panel and over-represents senders with known deliverability issues; it is not a representative sample of ecommerce email. Complaint-rate thresholds cited are Google’s published guidance, not measurements from this dataset.
Note: Content created with the help of AI, human edited and fact-checked. External statistics cited with links to original sources.
