Most email teams still talk about sender reputation as if it’s a single number. It isn’t and it never really was. Each of the mailbox providers (Gmail, Outlook, Yahoo, Apple Mail etc) has their own view of your domain and your sending IPs and none of them will ever tell you how they arrive at that view.

That distinction is no longer academic. In September 2025 Google announced it was retiring the domain and IP reputation dashboards in Postmaster Tools (the closest thing most senders had to an official Gmail score). The old interface was retired at the end of October 2025. If the biggest mailbox provider is no longer giving you a score then the real question is what are they actually looking at?

• Sender reputation isn’t a single score. Each mailbox provider has its own way of calculating it and none of them will publish their full algorithm.
• Each provider looks at two reputations simultaneously: your domain and your sending IP. Since the 2024 bulk sender rules the domain has taken on more of the burden.
• The inputs are all well known (complaints, bounces, spam traps, engagement, auth, volume consistency and unsubscribe) even if the weighting of them isn’t.
• A good score on one tool tells you absolutely nothing about another provider. You need to look at each major provider in isolation and then build your sending program around what each one is capable of seeing.

Your reputation score tells an incomplete story

Open up a reputation dashboard, and you’re pretty much always going to get a verdict in the form of a Sender Score out of 100, a Microsoft SNDS color, and a status in your ESP. The thing you then do is you check that number, you read it as how “the inbox” sees you, and you go from there to optimize.

For a long time that shortcut was perfectly valid. When IP reputation was the main thing driving filtering and the vast majority of traffic was coming from a very small number of dedicated IPs, an IP score gave you something tangible to go on.

The issue is coverage. Validity’s Sender Score is an IP score from 0 to 100 based on data supplied by over 80 mailbox and security providers. Validity don’t publish the full list of providers so you can’t be sure that Gmail, Outlook or Yahoo are in that list or that any of them actually use the score in their filtering. Microsoft SNDS only covers Microsoft’s consumer mailboxes. As for Gmail’s own view, that’s actually become less readable than it was. Postmaster Tools v2 does provide the spam rate, feedback loop and delivery error data plus a compliance status dashboard, but no domain or IP grade.

So every dashboard is perfectly accurate about its own slice and completely silent on the rest. In our deliverability work we see time and again examples of programs that have a very good Sender Score and at the same time are seeing a steady increase in spam folder placement at Gmail. Both of those are true. They’re just different receivers.

What sender reputation is actually made of?

Every mailbox provider runs their own proprietary model. They don’t share one and they don’t tell you how the various signals are weighted. What they will tell you is the raw material.

Two reputations, judged together

Each message is assessed against two things:

• Domain reputation is attached to your authenticated From and DKIM domains and rides with them across ESPs, IPs and infrastructure changes.
• IP reputation is attached to the sending server.

Yahoo put it down in their Sender Best Practices: “Each IP and DKIM domain has a reputation which can impact the delivery of your email.”

The balance has swung to the domain. Since February 2024 Gmail and Yahoo have required bulk senders to authenticate via SPF and DKIM, publish DMARC and have the From domain aligned. Microsoft followed for high volume senders to Outlook.com, Hotmail and Live in May 2025. Once identity is anchored to an aligned domain that domain is the one that the providers can then track. That’s why we think a move to a fresh IP rarely, if ever, fixes a reputation issue any more.

The signals providers draw on

The list is consistent across providers, even if the weighting isn’t:

• spam complaints from recipients
• hard bounces and sends to unknown users
• spam trap hits
• engagement, positive and negative: opens, clicks, replies, deletes without reading (opens are a weaker signal than they were, since Apple Mail Privacy Protection preloads images and records opens that never happened)
• authentication results for SPF, DKIM and DMARC
• volume consistency, especially sudden spikes
• unsubscribe handling, including one-click unsubscribe

The one threshold that’s published

Gmail gives the clearest reference point. Its sender guidelines ask bulk senders, defined as sending close to 5,000 or more messages a day to personal Gmail accounts, to keep the user-reported spam rate below 0.1% and never let it reach 0.3%. Since June 2024, senders above 0.3% have been ineligible for mitigation until the rate stays below that line for seven consecutive days. Yahoo sets the same 0.3% ceiling.

One detail changes how you read that number. Both Gmail’s Postmaster Tools v2 and Yahoo calculate spam rate on mail delivered to the inbox. Mail already filtered to spam can’t be reported as spam, so a falling complaint rate can mean recipients are happier, or that fewer of your messages are reaching them.

And as each provider calculates their own score you can be in great shape on Gmail and in real trouble on Outlook in the same week. Monitoring across the providers is far more important than just looking at any one score.

Where sender reputation is really decided

Monitoring tells you what each provider sees. What they see is determined earlier on in the process by how your program collects addresses, how it selects its audience and how it manages its volume. Those are lifecycle decisions and they are typically the responsibility of a different team to the one that’s looking at the dashboards. In our experience that gap is responsible for more reputation issues than any one bad send.

Five practices do most of the work:

• Start clean at sign up. Confirmed opt in and address validation mean that typos, traps and invalid addresses are all eliminated from the list before they can turn into bounces.
• Send by engagement not by list size. Mail out your most recently engaged subscribers first and only broaden the list once you’ve got a good level of engagement (i.e. low complaint and bounce rates). Run a re-engagement campaign for the lapsing subscribers and then sunset the ones that don’t respond.
• Separate your mail streams. Yahoo recommend keeping your bulk marketing mail separate from your transactional/alert mail by IP or by DKIM domain. Using different subdomains helps to contain a complaint spike in one stream (though the parent domain will still have the shared history).
• Ramp up prior to peaks. Warm up your new domains and IPs gradually and build volume over the weeks leading up to a seasonal peak so it doesn’t come as a sudden spike.
• Monitor via the provider. Check Google Postmaster Tools (for Gmail, SNDS and JMRP for Outlook.com) and Yahoo’s Complaint Feedback Loop. Then break your own bounce and complaint data down by recipient domain as that’s the only view that gives you a view of every provider in one go.

Nothing is quick with this approach but it does work. Adina Eden, a jewellery brand, had a very bad domain and IP in Gmail Postmaster Tools when they rebuilt their email programme with Mavlers and the majority of their campaigns to Gmail (their biggest audience) were landing in spam. The recovery was achieved almost exactly as per the above steps. The Gmail sends were put on hold whilst positive engagement was built up at the other providers. Unengaged subscribers were suppressed, SPF/DKIM/DMARC were all aligned to the sending domain and volume was ramped back in stages with Gmail being re-introduced to the most engaged of the opt-in subscribers first.

Within two months domain and IP reputation had moved from bad to good and then to very good and inbox placement in Gmail had come back. Once the mail was in the inbox again the lifecycle flows (welcome, cart abandonment, winback) kept engagement at a steady level and email’s contribution to revenue increased by 30%. The order of things is important here in that the flows only worked once reputation had been restored and reputation only stayed where it was because of the flows keeping us in touch with people who wanted to hear from us.

Summing it up

Sender reputation was never a single score someone could hand you. It was always a set of private judgments, one at each mailbox provider, built from how recipients respond to your mail. Google’s decision to stop showing domain and IP reputation in Postmaster Tools simply makes that easier to see.

That shifts where attention belongs. A dashboard can tell you something has gone wrong at one provider, but it can’t fix it. What fixes it, and keeps it fixed, sits upstream: who gets added to the list, who gets mailed first, how fast volume grows, and how quickly unengaged subscribers are let go.

The senders who come through this best won’t be the ones with the best-looking score. They’ll be the ones whose recipients keep opening, keep clicking, and rarely reach for the spam button, at every provider at once.