How to land in Gmail Primary tab (and out of Promotions)

I have stopped counting how many operators have shown me a "deliverability problem" that was really a Gmail tab problem. Their messages are inboxed. The headers are clean. SPF, DKIM, and DMARC all align. They just keep landing under the green Promotions ribbon, and their open rates have fallen off a cliff over the last six months. So they switch ESPs, pay for warmup pools, hire consultants, and nothing moves. The problem was never their infrastructure. It was how their HTML looks to Gmail's tab classifier.

The fix for most Promotions placements is on the page you already control. Here is what Gmail scores, how to test it, and a real send we moved to Primary in one round of edits.

Promotions is not Spam, but it is not free either

First a calibration. The Promotions tab is inbox-delivered mail. Gmail has accepted the message, authenticated the sender, and decided the user wants to see it. Spam filtering already passed. The classifier you are fighting is a separate, lighter-touch model that sorts inboxed mail into Primary, Social, Promotions, Updates, and Forums. Falling into Promotions does not damage your sender reputation and does not increase your bounce rate.

What it does is change reader behaviour. Litmus and Return Path have both put the open-rate gap between Primary and Promotions in the 18 to 35 percent range, depending on category. For a daily editorial newsletter, that gap is brutal. For a weekly product launch announcement to engaged subscribers, Promotions is fine and sometimes preferable. The right question is not "am I in Promotions" but "is Promotions hurting this specific send".

If your last six sends averaged 38 percent opens and you suddenly dropped to 26 percent, check tab placement before you hunt for content issues. Half the time that is the whole story.

The 12 signals Gmail scores for tab placement

Google does not publish the classifier. But the pattern across thousands of audited sends and what their own Postmaster Tools reveals lines up to a reasonably stable list. These are the twelve signals we track at Newsletrix when we run a tab audit.

1. Link density. Links per 100 words of body text. Above roughly 1.5 you start looking promotional. Editorial newsletters that hit 4 or 5 links per 100 words are practically begging for Promotions even when the content is essays. Cutting two redundant CTAs out of a 1,200-word issue is often enough by itself.

2. Image-to-text ratio. Total image bytes versus rendered text length. A hero image plus three product tiles in a 400-word email reads as a catalogue. The same images in a 1,500-word essay are fine. Gmail is not anti-image, it is anti-image-dominant.

3. Promotional vocabulary. Words like sale, discount, free, percent-off, limited time, exclusive, and offer in the subject and first paragraph weight heavily. The classifier is not naive; "free" inside a sentence about a free trial scores differently than "FREE!" in a subject line. But the lexical signal is real when stacked with other tells.

4. List-Unsubscribe and List-Unsubscribe-Post headers. Required for bulk mail under RFC 8058 and Gmail's 2024 sender rules, and you should have them. They do mark the mail as bulk, which leans toward Promotions, but the alternative is going to Spam outright. Keep the headers, fight the placement elsewhere.

5. ESP fingerprint and shared IP reputation. Mail sent through Mailchimp's shared IPs, Klaviyo, HubSpot, and similar marketing-first ESPs carries a return-path domain that Gmail has seen a billion promotional sends from. Beehiiv, ConvertKit, and Substack tend to land in Primary more often for editorial content out of the box, partly because their average customer is editorial. Mailgun and Postmark on dedicated IPs sit in between.

6. HTML structure. Table-based campaign templates with nested tables, inline styles, and tracking-pixel scaffolding read promotional. Semantic HTML with paragraphs and minimal styling reads like correspondence. The literal HTML shape, before any content, is part of the score.

7. Tracking pixel and tracker domain count. One open-tracking pixel is invisible to Gmail. Five trackers from five third-party domains plus link-wrapping through a click-tracking domain different from the sender domain reads like commercial infrastructure.

8. Recipient engagement and drag-to-Primary history. The strongest single signal. If a recipient has opened, replied to, or dragged your past mail to Primary, Gmail tilts that recipient's future placements toward Primary almost regardless of content. This is why brand-new subscribers and dormant ones see Promotions while your power users see Primary, from the same send.

9. SPF, DKIM, DMARC alignment. These do not directly move you to Primary, but failures push you toward Promotions or Spam. Set them up once correctly. Our SPF, DKIM, and DMARC walkthrough covers the alignment rules.

10. Send cadence consistency. A sender who emails the same list every Tuesday at 9am for six months earns Primary placement faster than the same sender batching sporadic bursts. Cadence is a behavioural fingerprint.

11. Footer length and unsubscribe placement. Long legal footers with address blocks, social icons, preference centers, and forwarded-to-a-friend widgets read promotional. They are also legally required in some jurisdictions, so the play is to keep them short and compliant. Our footer compliance checklist walks the minimum.

12. Subject line promotional triggers. All caps, exclamation points, currency symbols, percent-off numbers, and emoji in the subject all push the score. None alone is a death sentence; three stacked usually are.

How to test your current tab placement

Before you change anything, measure. Otherwise you are debugging blind and any improvement you see could be regression to the mean.

Build a seed list. Eight to twelve Gmail accounts is enough. Mix engagement profiles: a couple that have opened every issue, a couple that have not opened in 90 days, a couple of brand-new accounts you just created with no history. Send your next issue to the seed list and the rest of your real list at the same time. Then open each Gmail account manually and note which tab the message landed in.

For a deeper read, view the raw headers in Gmail (three-dot menu, "Show original") and search for X-Gmail-Labels. The label tells you exactly which tab Gmail assigned. This is the only first-party answer there is. Inbox monitor services like GlockApps and Mail Tester also work, but their seed accounts have no engagement history with you, which biases their results toward worst-case.

Track seed-list results over three or four sends before you trust them. One issue can swing on noise. A pattern across four sends is signal.

A real before-and-after

Last month we audited a B2B newsletter that had drifted from 42 percent open rate to 28 percent over a quarter. The operator had already tried two subject-line rewrites and was about to switch from Mailchimp to Beehiiv. The content was excellent. Tab placement was the problem.

The original send had 6 inline images, 14 outbound links, a 1,200-word essay, a footer with five social icons and a preference center widget, and the subject line "Inside: how Patagonia rebuilt their loyalty program (+ a free template)". Seed-list test: 9 out of 10 Gmail accounts placed it in Promotions.

We made four changes to the same content. Cut images from 6 to 2. Cut links from 14 to 6 by consolidating five "read more" links to one source and dropping two social-share buttons. Removed "free" from the subject. Trimmed the footer to the legally required address plus a single unsubscribe link. We did not change ESPs, IPs, or send time.

Next send: 8 out of 10 seed accounts landed in Primary. Open rate on the full list moved from 28.4 percent to 37.9 percent. CTR held steady, which mattered, because we wanted to confirm we had not cut engagement-driving links. The ones we removed were redundant. Three weeks later the operator had dropped the ESP migration entirely.

The 7-point pre-send checklist

Run this before every send. It takes two minutes once you have the muscle memory.

1. Body has fewer than 1.5 links per 100 words.

2. Image bytes are less than 40 percent of total message size, and there are no more than 3 images.

3. Subject line has no all-caps words, no currency symbols, and no emoji.

4. The first 90 characters of body copy contain no promotional vocabulary.

5. Footer is under 150 words: address, unsubscribe, one or two social links maximum.

6. SPF, DKIM, and DMARC all pass alignment for the sending domain (check the raw headers of a test send).

7. List-Unsubscribe and List-Unsubscribe-Post headers are present and one-click compliant.

Tactics that do not work

A few things people try that we have watched fail.

Asking your whole list to drag you to Primary. At small list sizes this can move the needle for the few who comply. At any real scale you are asking thousands of people to take a manual action almost none of them will, and the campaign to ask is itself a promotional email that lands in Promotions. Earn the drag-to-Primary one reader at a time by making the content worth opening.

Stripping every image. A plain-text email is hard for the classifier to call promotional, but you lose brand recognition, you lose layout, and you cannot reliably track opens without a tracking pixel. The fix is fewer images, not zero.

Switching ESPs without changing content. If your HTML structure, link density, and footer all scream campaign, the new ESP is shipping the same signals through a different relay. We have seen senders move from Mailchimp to ConvertKit and stay in Promotions because they brought the template with them. To compare ESPs on tab placement, our ESP detection guide and the Mailcharts comparison walk through the differences.

Frequently asked questions

Why is my newsletter going to Gmail's Promotions tab?

Gmail classifies a message as Promotions when its content and headers look commercial: many links, high image-to-text ratio, promotional vocabulary in the subject and body, a List-Unsubscribe header, a known marketing ESP fingerprint, and a sender the recipient has not engaged with personally. None of these alone forces Promotions, but they stack. Most newsletters land in Promotions because their HTML reads like a sales email even when the content is editorial.

Does Gmail's Promotions tab hurt open rates?

Yes, though less than going to Spam. Independent measurements from inbox monitoring vendors put the open-rate hit at roughly 18 to 35 percent compared to Primary placement for the same audience and send. The tab itself does not block delivery, but readers check Primary daily and Promotions in batches or not at all. For editorial newsletters where opens drive revenue, this is the difference between a healthy issue and a quiet one.

Can I force Gmail to deliver to the Primary tab?

No. Gmail's tab classifier is unilateral and there is no header, setting, or paid option that overrides it. What you can do is change the signals it scores: cut link density, drop image weight, remove promotional vocabulary, send from a clean authenticated domain, and earn drag-to-Primary actions from real readers. Over a few sends Gmail learns that your messages behave like correspondence rather than marketing and reclassifies accordingly.

How do I test which tab Gmail is using for my newsletter?

Build a seed list of 8 to 12 Gmail accounts with varied engagement histories: some that have opened your past sends, some that have ignored them, some fresh accounts with no history. Send your next issue to the seed list and check each inbox manually. Look at the raw headers for X-Gmail-Labels, which lists the tab Gmail assigned. Free inbox monitors give a starting view but they cannot replicate engagement history, which is the strongest single tab signal.

Does plain-text email always land in the Primary tab?

Usually but not always. Plain-text removes most of the visual signals Gmail uses to flag promotions, so it tilts the classifier hard toward Primary. But a plain-text send from a marketing ESP IP, with a List-Unsubscribe header and a "SALE" in the subject, will still get flagged. Plain-text is a strong lever, not a guarantee. It also kills your ability to track opens reliably, so weigh the tradeoff.

Related reading