The Email Marketing Metrics Worth Watching

Email reporting can create a false sense of precision. A dashboard may show dozens of charts, yet still fail to answer the questions that matter: Are our emails reaching people who expect them? Are they useful enough to prompt action? And do they produce a worthwhile commercial outcome?

The answer is not to watch every available number. It is to use a small, connected set of metrics that covers the journey from list quality and delivery through to engagement, conversion and retention. Crucially, treat each metric as evidence, not a verdict. A high click rate can coexist with poor profitability; a low unsubscribe rate can reflect a disengaged list rather than strong loyalty; and open rates are no longer reliable enough to be a primary success measure.

This guide explains the email marketing metrics worth watching, how to calculate them and what to do when they move in the wrong direction.

Start with a measurement framework, not a dashboard

Before assessing an individual send, decide what that email was meant to achieve. A welcome email, a product launch, a replenishment reminder and a monthly newsletter should not be judged by identical standards.

Objective	Primary metrics	Useful supporting metrics
Protect sending health	Hard-bounce rate, spam-complaint rate, authentication pass rate	Temporary failures, domain reputation, inactive-contact trend
Win attention	Unique click rate, click-to-open rate	Open rate, read-time signals where available, unsubscribe rate
Generate sales or leads	Conversion rate, revenue per delivered email, qualified leads	Average order value, landing-page conversion rate, assisted revenue
Build a sustainable audience	Net list growth, source quality, preference changes	Unsubscribes, complaints, engagement by signup cohort
Improve lifecycle automation	Conversion by automation step, time to conversion	Exit rate, frequency exposure, revenue per entrant

Use one clear primary metric per campaign. For example, the purpose of a sale email may be revenue per delivered email, while a content newsletter may be qualified visits to an article hub. Deliverability and complaint metrics remain guardrails for every send, regardless of its commercial objective.

1. Deliverability metrics: measure whether you are earning the right to be seen

Delivered usually means the receiving mail server accepted the message. It does not prove that the email reached the inbox, appeared prominently or was read by a person. A message can be accepted and then filtered to spam or a secondary tab. Treat delivery as a technical hand-off metric, not an inbox-placement metric. (Postmark guidance)

Delivery rate

Formula: delivered emails ÷ attempted emails × 100

A sudden fall can indicate a sending or list problem, but do not overreact to a single aggregate number. Break results down by mailbox provider, sending domain, acquisition source and campaign type. A 98% delivery rate could conceal a serious issue if one important corporate domain is rejecting a large proportion of messages.

Hard-bounce rate and temporary-failure rate

Hard bounces are permanent failures, such as an address that does not exist. Soft bounces, delays or temporary failures can result from a full mailbox, a temporarily unavailable receiving server or rate limiting. Review the actual SMTP response where possible: it is more useful than a generic bounce label when diagnosing an issue. Suppress confirmed hard bounces promptly so they are not mailed again. (Postmark guidance)

Formula: hard bounces ÷ attempted emails × 100

Watch both the overall rate and the change in rate. If bounces rise after importing an older list, changing a signup integration or adding a new data source, investigate that source first. Email validation can be a helpful pre-send control, but it is not a substitute for permission, clear expectations and ongoing list hygiene.

Spam-complaint rate

Formula: spam complaints ÷ delivered emails × 100

This is one of the most important early-warning metrics because it measures a recipient’s explicit negative reaction. Google advises senders to keep the spam rate reported in Postmaster Tools below 0.10% and to avoid reaching 0.30% or higher. Its data is Gmail-specific and not necessarily real-time, but it is highly useful for spotting reputation risk among Gmail recipients. (Google’s sender guidance)

Do not wait for the total account-level complaint rate to become alarming. Look at complaints by campaign, audience segment, signup source, frequency band and automation step. A welcome email can generate complaints if people did not understand what they had signed up for; a discount campaign can do the same if it reaches people who only expected transactional updates.

Authentication and reputation signals

Authentication is not a campaign-performance metric, but it is foundational. SPF, DKIM and DMARC help mailbox providers verify who is allowed to send for a domain. For higher-volume senders to Gmail, Google requires SPF, DKIM and DMARC, aligned authentication, and marketing mail that supports one-click unsubscribe as well as a visible unsubscribe link. (Google’s sender guidance)

Monitor authentication pass rates and mailbox-provider diagnostics alongside campaign results. Gmail Postmaster Tools reports spam rate, authentication, delivery errors and compliance data for mail sent to personal Gmail accounts. Its legacy domain and IP reputation dashboards are being retired, so do not build a reporting process that depends on those views alone. (Google’s sender guidance)

What to do if deliverability worsens:

• Pause any risky list imports, reactivation campaigns or abrupt volume increases.
• Check SPF, DKIM and DMARC alignment for the domain in the visible From address.
• Separate marketing and transactional traffic where your sending setup allows it.
• Inspect hard bounces, temporary failures and complaints by recipient domain.
• Reduce frequency for low-engagement audiences before attempting to send more.
• Resume volume gradually when changing domains, routes or sending patterns.

2. Engagement metrics: useful signals, imperfect measurements

Open rate: retain it, but demote it

Formula: unique opens ÷ delivered emails × 100

An open is normally inferred when a hidden tracking image downloads. That creates blind spots: an email read with images disabled may not register, while privacy tools and security systems may create an open without a person deliberately reading the message. Apple Mail Privacy Protection can preload and cache images through a proxy, artificially inflating opens. Use open rate as a broad directional trend for comparable sends, not as proof that a subject line won or that a contact is genuinely active. (Postmark guidance)

For the same reason, avoid defining your most engaged audience solely as “people who opened in the last 90 days”. Combine opens with clicks, site activity, purchases, form submissions and other meaningful actions. A person who has clicked, bought or logged in is generally a stronger engagement signal than someone whose email app may have downloaded a pixel.

Unique click rate

Formula: unique recipients who clicked at least once ÷ delivered emails × 100

Click rate measures how much of the delivered audience took an observable action from the email. It is generally more useful than opens because it requires a journey beyond the inbox. However, it is not flawless: security scanners and corporate firewalls can occasionally follow links automatically. Look for unexpected spikes, repeated clicks immediately after delivery or clicks from unusual locations as clues that automated activity may be involved. (Postmark guidance)

Click-to-open rate (CTOR)

Formula: unique clickers ÷ unique openers × 100

CTOR attempts to answer a narrower question: among people who appeared to open, how persuasive was the email content and call to action? It can be helpful when comparing similar campaigns, such as two versions of a product launch. But because the denominator is open data, it inherits the weaknesses of open tracking. Never use CTOR alone to select a winner.

Link-level performance and click heatmaps

Aggregate clicks tell you whether the email worked; link-level data can suggest why. Review which content blocks, products, calls to action and navigation links received attention. A click heatmap can reveal, for instance, that recipients repeatedly choose a product image rather than the intended button, or that several links split attention away from the main action.

Interpret this in context. Header and footer links often attract routine clicks, while a product link may drive fewer clicks but much more revenue. Tag links consistently, and distinguish primary calls to action from secondary navigation in your reporting.

3. Commercial metrics: connect email activity to outcomes

For most organisations, the metrics that deserve the most attention occur after the click. An email cannot be considered successful simply because it earned engagement; it must contribute to the outcome it was designed for.

Conversion rate

Choose the denominator that answers your actual question, then name it clearly:

• Post-click conversion rate: conversions ÷ unique clickers. Useful for judging message-to-landing-page fit.
• Delivered-email conversion rate: conversions ÷ delivered emails. Useful for comparing the efficiency of campaigns with different reach.
• Session conversion rate: conversions ÷ tracked email sessions. Useful in web analytics, but affected by consent settings, browser behaviour and cross-device journeys.

For an ecommerce campaign, a conversion may be a completed order. For B2B, it may be a qualified demo request rather than every content download. Agree the definition before reporting results; otherwise, teams can optimise a low-value action simply because it is easy to count.

Revenue per delivered email (RPD)

Formula: attributed revenue ÷ delivered emails

RPD is one of the most practical comparison metrics for commercial email because it combines reach and conversion. Suppose Campaign A produces £8,000 from 80,000 delivered emails (£0.10 RPD), while a targeted replenishment message produces £2,000 from 5,000 (£0.40 RPD). The larger campaign may still be worthwhile, but the comparison tells you which message is more economically efficient.

Record the attribution rule next to every revenue figure. Is revenue credited after an email click only, an open or click, or any purchase within a fixed window after delivery? Different rules legitimately produce different totals. A robust report can include both direct revenue (for example, a purchase following a tracked email click) and assisted revenue, but should never present them as interchangeable.

Average order value, margin and incremental impact

Revenue can overstate value if a campaign relies on steep discounting or merely shifts an order that would have happened anyway. Where possible, add:

• Average order value: attributed revenue ÷ attributed orders.
• Gross margin after discount: particularly important for promotional mail.
• Repeat purchase or retention: useful for post-purchase and lifecycle programmes.
• Incremental lift: the difference versus a comparable group that was not sent the email.

A properly designed holdout group is the strongest way to estimate incrementality, though it is not always practical for every campaign. Use it periodically for large, recurring automations and major promotions rather than trying to create a complex experiment for every newsletter.

Reliable campaign attribution

Use consistent UTM parameters on every destination link so web analytics can identify email traffic. Google recommends including utm_source, utm_medium and utm_campaign; use a controlled naming convention rather than allowing each marketer to improvise. For example: utm_source=email&utm_medium=crm&utm_campaign=2026-09-autumn-launch&utm_content=hero-cta. (Google’s sender guidance)

In Email Foundry, campaign analytics, click heatmaps and revenue attribution can make this analysis easier when ecommerce events and website tracking are connected. The important work still happens before the send: define events, apply a naming convention and decide the attribution window consistently.

4. Audience-health metrics: a smaller, more willing list can perform better

Net list growth

Formula: (new subscribers − unsubscribes − hard bounces − complaint suppressions) ÷ starting mailable audience × 100

List size alone is a poor health metric. A database can grow while its reachable, interested audience shrinks. Track where new subscribers come from—checkout, content download, event registration, referral or imported CRM data—and compare the subsequent quality of each cohort: clicks, purchases, complaints, unsubscribes and bounce rates.

Unsubscribe rate and preference changes

Formula: unsubscribes ÷ delivered emails × 100

An unsubscribe is not automatically a failure. It is often preferable to a complaint, and it can be useful feedback about frequency, relevance or expectations. In the UK, marketing emails are subject to PECR rules, and people must be able to opt out; honouring an unsubscribe is therefore both a trust and compliance requirement. (ICO guidance)

Watch unsubscribe rate alongside complaint rate. If both rise, investigate relevance and frequency urgently. If unsubscribes rise but complaints remain low, your exit path may be working as intended. A preference centre—allowing people to choose topics or frequency—can retain subscribers who do not want every message.

Engagement decay and reactivation

Build engagement cohorts based on substantive behaviour, such as recent clicks, purchases or website activity. Then monitor what proportion of each cohort remains active after 30, 90 and 180 days. This gives a much clearer picture of list ageing than an all-time engagement average.

Before re-mailing a long-inactive group, weigh the potential revenue against deliverability risk. A focused re-permission or preference-update campaign is usually safer than dropping a large dormant audience into a major promotion. If there is no positive response, suppressing them may protect the people who remain engaged.

5. Automation metrics: measure the journey, not just the final total

Automations often outperform one-off campaigns because they are triggered by a relevant event, but they can also cause unnoticed fatigue. Report at three levels:

1. Entry: how many eligible people enter the journey, and how quickly after the trigger?
2. Step: delivery, clicks, unsubscribes, complaints and conversion for each email.
3. Journey: conversion rate, revenue per entrant, time to conversion and exit or suppression reasons.

For example, if a three-email abandoned-basket sequence produces most conversions after the first reminder but most complaints after the third, shortening the sequence or refining the final audience may improve the programme overall. Visual automations, dynamic segments and ecommerce-event tracking are particularly useful here because they allow marketers to inspect the decision rules behind the numbers, rather than treating the journey as a black box.

6. Avoid the most common reporting mistakes

• Comparing unlike campaigns. A welcome email sent immediately after signup should not be benchmarked against a broad monthly newsletter.
• Using industry averages as targets. List source, offer, audience, send frequency and measurement method differ too much. Compare against your own historical baseline and similar campaign cohorts first.
• Calling delivered “inbox placement”. Server acceptance is not inbox visibility.
• Making opens the main KPI. Privacy features and image loading make them inherently uncertain.
• Counting every click as human intent. Filter known bot activity where your tools support it, and investigate anomalous patterns.
• Claiming revenue without stating attribution. A figure with no attribution model or window cannot be compared reliably.
• Declaring an A/B-test winner too early. Set the audience, test variable, success metric and evaluation period before sending. Change one meaningful variable at a time where possible.

A practical 30-day action plan

1. Write a metric dictionary. Document definitions and denominators for delivered, unique click, conversion, attributed revenue, unsubscribe and complaint rate. Make this the shared reporting standard.
2. Set up the deliverability baseline. Check SPF, DKIM and DMARC; verify that bounces, unsubscribes and complaints suppress contacts correctly; and review Gmail Postmaster Tools if Gmail volume is sufficient.
3. Audit the last 10 sends. Group them by campaign type, audience and mailbox provider. Identify the median—not merely the average—for delivery, click rate, unsubscribe rate, complaint rate and RPD.
4. Standardise tracking. Create a UTM naming convention, ensure purchase or lead events are captured, and agree a documented attribution window.
5. Build three operational segments. Create engaged, recently inactive and long-inactive groups using clicks, purchases and site activity in addition to opens.
6. Choose one campaign to improve. Select a high-volume recurring send. Test a single meaningful change, such as offer framing, primary call to action, audience rule or send timing. Judge it on the predefined commercial and deliverability guardrail metrics.
7. Review monthly at programme level. Ask: are we gaining willing subscribers, protecting sender reputation, increasing valuable actions and avoiding unnecessary frequency? Use the answer to change the programme, not merely to decorate a report.

The most valuable email metrics do not simply make performance look measurable. They help you make better decisions: whom to mail, what to send, when to stop, and where email is genuinely contributing to customer value.

Frequently asked questions

What is the most important email marketing metric?

There is no single universal metric. Spam complaints and hard bounces are essential health guardrails, while revenue per delivered email, qualified leads or conversions should usually be the main success measure for commercial campaigns.

Is open rate still a useful email metric?

Yes, but only as a directional secondary signal. Image blocking, Apple Mail Privacy Protection and automated activity mean it should not be the main KPI or the sole basis for engagement segments.

How do you calculate email click-through rate?

Divide the number of unique recipients who clicked at least one link by the number of delivered emails, then multiply by 100. State clearly whether your platform uses delivered or sent messages as the denominator.

What is a good spam-complaint rate?

Keep it as low as possible. Google advises keeping the Gmail spam rate below 0.10% and avoiding 0.30% or higher. Investigate spikes by campaign, audience and signup source rather than relying on an overall average.

What is revenue per delivered email?

It is attributed revenue divided by delivered emails. It is useful for comparing campaigns of different sizes, provided you apply the same attribution rules and time window to each campaign.

Should an unsubscribe increase always be treated as bad?

No. An unsubscribe can be healthier than a spam complaint and may indicate that your opt-out process is clear. Review it alongside complaints, frequency, audience source and preference choices to understand the cause.

Which metrics should be used for email automations?

Measure entry volume, delivery and engagement at each step, conversion and revenue per journey entrant, time to conversion, unsubscribes and complaints. This shows where a journey is helping or creating friction.

Sources and further reading

• Email sender guidelines — Google
• Postmaster Tools dashboards — Google
• Open Tracking and Apple Mail (Mail Privacy Protection) — Postmark
• What are bounces and spam complaints? — Postmark
• URL builders: Collect campaign data with custom URLs — Google Analytics
• Guide to PECR — Information Commissioner's Office