The criteria that actually make up domain reputation

No provider publishes a reputation formula, and there's no single number to query. What exists instead is a set of specific, individually checkable inputs that providers weigh in forming a judgement about a sending domain. Knowing the inputs is the closest thing to understanding "reputation" that's actually available.

Authentication correctness

SPF, DKIM, and DMARC alignment are close to a prerequisite rather than one input among equals — mail that can't be reliably attributed to the domain is handled differently from mail that can, regardless of how good every other signal looks. This is checkable directly: an SPF lookup, a DKIM selector check, and a DMARC record with alignment reports all confirm this independently of anything a vendor dashboard reports.

Bounce rate

The proportion of sends that hard-bounce is one of the most directly measurable inputs, because it's visible in your own sending platform's data without needing any third-party tool. A hard bounce rate climbing above roughly 2% is a widely cited point where providers start treating a sender with more caution, though exact thresholds aren't published.

Complaint rate

The proportion of recipients marking mail as spam, visible through feedback loops where your sending platform or ESP supports them. Google's own published threshold — a spam rate kept under 0.3%, with 0.1% a safer target — is one of the few concrete, provider-confirmed numbers available in this whole domain, since Google has stated it directly for Gmail.

Blocklist status

Whether the domain or its sending IP appears on any blocklist that feeds into provider filtering decisions. This is checkable through blocklist lookup tools, though their relevance varies — a listing on an obscure, rarely-consulted list matters far less than one on a list that actively feeds major providers' filtering, such as Validity's Heatwave for domains associated with synthetic engagement.

Sending consistency

A pattern of sending that looks steady and explainable — a gradual ramp during warm-up, consistent volume day to day once established — rather than erratic spikes and silence. This is harder to check directly but shows up indirectly in your own send volume history over time.

Content signals

Patterns within the message content itself — identical templates sent repeatedly, suspicious link structures, content that trips spam-keyword classifiers — contribute independently of sender-level signals. This is the input most senders have the most direct control over and the one most often under-checked, since it requires actually reading mail the way a filter would rather than only checking delivery metrics.

Engagement

Opens, clicks, and especially replies, weighted roughly in that order of strength, as discussed in how each signal is interpreted. Genuine engagement is the hardest of all these inputs to manufacture convincingly, which is exactly why it carries real weight.

Why no single score captures all of this

Each provider weighs these inputs differently and doesn't disclose the weighting, which is why a single third-party "reputation score" should be read as one vendor's estimate of these inputs, not an authoritative readout of what any specific provider actually thinks. The more useful approach is checking the individual inputs directly where you can, and treating any aggregate score as a summary rather than a source of truth.

Why this list is worth keeping somewhere you can check it

Most senders check one or two of these inputs when something already looks wrong, rather than checking all of them on a regular cadence. Keeping a simple record — even a basic spreadsheet updated weekly with bounce rate, complaint rate, and blocklist status per domain — turns this from a reactive list into something that catches a problem while it's still small, which is the actual value of knowing what the inputs are in the first place.

A note on signals you can't check directly

Some inputs providers may weigh — IP neighbourhood reputation, for instance, or signals tied to hosting infrastructure rather than the domain itself — aren't fully checkable by a sender using public tools. Where a signal isn't directly checkable, the more productive move is controlling the inputs that are, rather than spending time trying to infer something that stays opaque regardless of effort.

MailPilot checks the individually verifiable inputs — SPF, DKIM, and DMARC alignment, bounce and complaint rate, and real per-provider placement through a seed network — rather than reducing domain reputation to a single opaque number.