Gmail Deleted Your Reputation Score: A Postmaster Tools v2 Playbook

The first thing most senders did when Gmail placement dropped was open Postmaster Tools and look at the colour of the Domain Reputation chart. High, Medium, Low, Bad. It was crude and it lagged reality by days, but it gave you somewhere to start. That chart is now gone. Google retired the v1 dashboard and its reputation panels, the v1 API stopped answering, and in August 2026 the last of the reputation data disappeared for senders who still had it. You are not getting it back.

What replaced it is arguably more useful and definitely more work. Postmaster Tools v2 gives you a Compliance Status page and a Deliverability Analysis verdict written in plain sentences instead of colours. Neither tells you why something broke. That part is now entirely yours.

This is a working process for diagnosing a Gmail problem without a reputation score.

What you need before you start

You cannot investigate what you are not measuring. Before a problem hits, have these four things in place:

• Postmaster Tools v2 access, with your sending domain verified via DNS TXT record. Verify every subdomain you send from, not just the root.
• Gmail’s Feedback Loop configured, which means adding a Feedback-ID header to your outgoing mail so complaint data is grouped by campaign or stream. Most platforms do this automatically; check yours.
• Event-level data from your sending platform (accepted, deferred, rejected, bounced, opened, clicked, unsubscribed) filterable by recipient domain. If you cannot isolate Gmail from everything else, you cannot tell a Gmail problem from a general one.
• A record of what you changed and when. Sending volume, audience source, DNS edits, template changes, platform migrations. Most deliverability investigations end with “we changed something three weeks ago.”

Step 1: Confirm you actually have a Gmail problem

Start by establishing what changed, when it changed, and who it affected. Three questions, in that order.

Segment your last 90 days of sending by recipient domain. If accepted rates and engagement fell across Gmail, Outlook, and Yahoo at the same time, the cause is on your side and Gmail is a symptom. If Gmail alone declined, keep going. If Gmail held steady while another provider dropped, you have been investigating the wrong mailbox provider for the last hour.

Also check where the mail is landing. Marketing email arriving in the Promotions tab is not a deliverability failure. That is where Gmail files marketing email, and chasing it into the Primary tab is a losing project.

Step 2: Read Compliance Status before anything else

Compliance Status is the closest thing v2 has to a pass/fail. It checks whether your mail meets Gmail’s published sender requirements: SPF and DKIM authentication, DMARC alignment, TLS encryption, valid forward and reverse DNS, one-click unsubscribe headers, working unsubscribe processing, and a reported spam rate under 0.1%.

Fix anything failing here first. A broken DKIM signature on one sending subdomain explains a lot more than any reputation score ever did, and it is a concrete thing you can repair this afternoon. Our SPF, DKIM and DMARC guide covers the records themselves.

Compliance does not guarantee inbox placement. It rules out the foundational technical failures so you can stop suspecting them.

Step 3: Treat the Deliverability Analysis verdict as a clue, not a diagnosis

The Deliverability Analysis panel returns one of a small set of verdicts. Each one narrows the search. None of them closes the case.

Gmail’s verdict	What it establishes	Where you look next
Not enough outgoing email	Google lacks the volume to judge	Your own event data; do not inflate sends for the dashboard
Many messages have delivery errors	Elevated rejections or deferrals	SMTP response codes, volume changes, one stream or all
Sender does not meet requirements	A named requirement is failing	Compliance Status for the specific check
Spam rate above 0.1%	Inboxed mail is generating complaints	Feedback-ID data by campaign, audience and acquisition source
Users do not want your messages	Strong negative recipient signals	Permission, frequency, recently added segments
Users don’t take action	Neither positive nor negative signals	Whether mail is inboxed and ignored, or already filtered
Users signal they want your messages	Strong positive signals	Compare this traffic against your weaker segments

“Users don’t take action” is the verdict that traps people. It can mean recipients saw your mail and ignored it. It can equally mean the mail is already in spam, where nobody generates engagement signals in either direction. Those two situations call for opposite responses, and Gmail will not tell you which one you are in. Your own inbox placement testing will.

Step 4: Interpret spam rate carefully

Gmail’s reported spam rate counts complaints from mail that reached the inbox. That single detail changes how you read it.

A rate of 0.00% attached to collapsing engagement is not good news. It usually means most of your mail is already being filtered, so very few people have the opportunity to report it. Low complaints plus low delivery is a worse position than 0.08% complaints with healthy placement.

Watch the trend across two weeks rather than reacting to a single day. Gmail’s daily figures move around on low volume, and one bad Tuesday is not a signal.

Step 5: Use your own event data as the primary evidence

Your sending platform sees things Gmail’s dashboard never will: the exact SMTP response text on a deferral, the hour a rejection pattern started, which template was sending at the time.

Pull these for Gmail recipients specifically and compare against a stable control period:

• Accepted, deferred and rejected counts, with response codes
• Bounce classification, split hard versus soft
• Opens, clicks, replies and unsubscribes
• Volume and send frequency per recipient
• Performance broken out by campaign, segment, sending domain and IP

The point is not to stare at every chart until one confesses. You are looking for the moment the line bent and what else moved at the same time.

Postmark is the strongest platform in this category for diagnosis, mostly because of how much detail it keeps. Message streams separate transactional from broadcast traffic so a marketing problem cannot drag down password resets, full message content is retained for 45 days by default, and bounce data is classified rather than dumped as raw SMTP text. The Basic plan runs $15/month for 10,000 emails as of September 2026, and DMARC monitoring is a $14/month per-domain add-on through DMARC Digests.

The limitation is scope. Postmark bans cold outreach outright, enforces content policies more strictly than most competitors, and is not built to run a general marketing programme. It is a diagnostic instrument, not a marketing suite.

Step 6: Rebuild the signals Gmail took away

The reputation chart aggregated several things into one colour. To replace it, you need those things separately, and no single tool covers all of them.

Inbox placement testing. Seed lists tell you where controlled messages land at each provider. They are a sample, not a census, and seed accounts do not behave like real recipients, so treat a clean seed result as weak evidence rather than proof.

Blocklist monitoring. Useful for catching a listing quickly. A listing on a minor blocklist often has no measurable effect on Gmail placement, so confirm impact in your own data before you panic.

DMARC aggregate reports. These show every source sending as your domain and whether it authenticates. They explain authentication and infrastructure, not whether anyone wants the mail.

Mailgun packages inbox placement and reputation monitoring into Optimize, sold separately from its Send plans, a detail worth knowing before you budget. Send pricing runs from a free tier at 100 emails/day up to $90/month for 100,000 emails on Scale, with dedicated IP pools and send-time optimisation appearing at that level. The catch is that the deliverability tooling most people associate with the brand is not in those numbers. Additional dedicated IPs are $59 per IP per month on top.

If you are choosing a platform primarily on inbox performance, our best deliverability tools comparison ranks the field, the Mailgun vs Postmark breakdown weighs the two head to head, and the Postmark pricing page covers what the volume tiers actually cost as you grow.

Step 7: Change one variable, then wait

Once you have a hypothesis, test it properly. Change one meaningful thing. Hold a comparable cohort steady. Give it at least a week: Gmail’s assessment of a sender moves slowly, and it moved slowly back when there was a chart to watch too.

Common single-variable changes worth trying, roughly in order of how often they work:

1. Suppress the audience segment added immediately before the decline
2. Cut send frequency to Gmail addresses by half
3. Move one stream to a separate authenticated subdomain
4. Reduce volume to the level of your last stable month and hold it there
5. Run a confirmed re-opt-in to your dormant segment, then suppress non-responders

Keep a rollback plan for each. If you change four things at once and placement recovers, you have learned nothing you can reuse.

Common mistakes

Replacing one score with another. The instinct after losing a metric is to find a substitute number. Every candidate (seed test scores, third-party reputation grades, blocklist counts) measures a fragment. Combining them into a hypothesis works. Treating any one as the new mood ring does not.

Blaming Gmail first. Mailbox providers do ship bugs and adjust thresholds. But external explanations should come after you check your own changes. If the decline is limited to Gmail, started on a specific date, and no other sender you know reported the same thing, look at what you did.

Suppressing everyone who looks inactive. “No engagement” from Gmail can mean the mail is already in spam. Mass-suppressing those addresses removes the evidence and the audience at the same time. Investigate placement before you sunset a segment.

Routing around an unsubscribe with a second domain. Recipients do not see your infrastructure diagram. They told your brand to stop. Stream separation protects important mail from problems elsewhere; it is not a way to keep mailing someone who said no.

Manufacturing volume for the dashboard. If Gmail reports insufficient data, sending more to trigger a verdict is exactly the behaviour that creates the problem you are trying to diagnose.

What success looks like

You will know the process is working when your answer to “how is our Gmail deliverability?” stops being a colour and starts being a sentence with a date in it.

Concretely: Compliance Status shows no failing checks. Spam rate sits under 0.1% and your accepted rate to Gmail is holding above 95%. Engagement among recipients acquired in the last 90 days looks materially better than your dormant segment, which is what a healthy acquisition source produces. And when something does move, you can name the change that caused it within a day rather than waiting for a chart to catch up.

That last one is the actual upgrade here. The reputation score told you about a relationship that had already gone wrong. Your own data tells you while it is happening.

Sources

1. Gmail Reputation Data Is Dead. Now What? — Mailgun — accessed 2026-09-01
2. Gmail Feedback Loop — Google Support — accessed 2026-09-01
3. Postmark Pricing — Official — accessed 2026-09-01
4. Mailgun Pricing — Official — accessed 2026-09-01

Further reading: Mailgun’s write-up of the reputation removal, Google’s Feedback Loop documentation, and the Postmaster Tools dashboard itself.