Email marketing KPIs: 10 metrics to track, with formulas
Ten email marketing KPIs, each with a formula, how to read it, and where it misleads. Then the lifecycle KPIs SaaS teams use to tie email to activation, upgrades, and revenue.
Every number in this section comes from one example send: Acme’s monthly product update, which announced a feature on its $100 plan, sent to 10,000 subscribers. For a percentage, divide, then multiply by 100.
Our weekly dashboard would hold three: conversion rate, spam complaint rate, and unsubscribe rate. Conversion is the result. The other two are early warnings: complaints get your mail filtered, and unsubscribes count the people who’ve had enough but were polite about it. Click-through, open, click-to-open, bounce, and inbox placement are diagnostics you open when one of those three moves. List growth is a monthly read, ROI a quarterly one.
The share of delivered emails someone opened. Formula: unique opens ÷ emails delivered.
Acme: 4,428 unique opens ÷ 9,840 delivered = 45%.
Verdict: a smoke alarm, not a score. Apple Mail Privacy Protection loads the tracking pixel on arrival, so open rate counts opens nobody made. Loops’ guide to your first onboarding emails says to pause and check if a campaign falls below 30%, or a welcome workflow below 40%, while you build sender reputation.
The share of delivered emails that got at least one click. Formula: unique clicks ÷ emails delivered.
Acme: 412 unique clicks ÷ 9,840 delivered = 4.2%.
Verdict: the best diagnostic you have: when conversion moves, check this first. Compare it only with sends of the same kind. Security filters that follow links add a few clicks nobody made.
Of the people who opened, the share who clicked. Formula: unique clicks ÷ unique opens.
Acme: 412 unique clicks ÷ 4,428 unique opens = 9.3%.
Verdict: fine for comparing two versions of one email, weak on its own. Apple Mail’s automatic opens sit in the denominator, so it inherits every problem open rate has.
The share of recipients who did what the email asked within a set window. Formula: recipients who convert within the window ÷ emails delivered.
Acme: 24 upgrades to the $100 plan within 7 days ÷ 9,840 delivered = 0.24%.
Verdict: on the dashboard, because it’s the result. Count it in billing or your product, with the same window every time. In Loops, a campaign goal counts contacts who reach a target within a window. It shows what followed the email, not what it caused.
Emails the receiving server rejected, as a share of emails sent. Formula: hard and soft bounces ÷ emails sent.
Acme: 160 bounces (40 hard, 120 soft) ÷ 10,000 sent = 1.6%, leaving a 98.4% delivery rate.
Verdict: a guardrail you read when it jumps, reasons first. Hard bounces are addresses that don’t exist, so remove them. Soft ones, like a full inbox, often clear on their own. Delivered only means the server accepted the email.
The share of delivered emails that led someone to unsubscribe. Formula: unsubscribes ÷ emails delivered.
Acme: 25 unsubscribes ÷ 9,840 delivered = 0.25%.
Verdict: on the dashboard. Compare it with past sends: a jump points at the audience, the frequency, or the content. A low rate doesn’t prove people like the email, since some report spam instead.
How often recipients mark your email as spam. Formula: spam reports ÷ emails delivered.
Acme: 6 spam reports ÷ 9,840 delivered = 0.06%.
Verdict: on the dashboard, and the one KPI with a limit set by someone else. Google requires senders to keep the spam rate in Postmaster Tools below 0.3% and recommends under 0.1%. Not every mailbox provider reports single complaints back, so check Gmail’s rate there too.
How fast the number of people you can email grows. Formula: new subscribers minus unsubscribes and removed contacts, ÷ subscribers at the start of the month.
Acme: (585 new minus 25 unsubscribes minus 40 hard bounces removed) ÷ 10,000 at the start = 5.2%, so 10,520 at month end.
Verdict: a monthly read. Count only people who opted in: a bought list grows the number and hurts deliverability.
The share of delivered emails that land in the inbox, not spam. Formula: emails that reach the inbox ÷ emails delivered, measured on a seed list of test addresses.
Acme: 46 of 50 seed addresses got it in the inbox, 46 ÷ 50 = 92%.
Verdict: a diagnostic for when opens and clicks drop and the content hasn’t changed. Platforms see deliveries, not folders, so this comes from seed tests and Google Postmaster Tools. Seeds are a sample: treat the number as a hint.
What email returns for what it costs. Formula: revenue attributed to email minus email costs, ÷ email costs.
Acme: 24 upgrades × $50 more a month = $1,200. ($1,200 minus $400 in software and writing time) ÷ $400 = 200%.
Verdict: a quarterly number for budget talks, only as good as its attribution rule, so pick one and keep it. For subscriptions, revenue per send, below, compares single emails more fairly.
Loops reports delivery and engagement for every campaign and workflow email. Payments, upgrades, and cancellations come from your billing system. The table splits them and includes the five lifecycle KPIs covered below.
| KPI | Formula | Where to find it in Loops |
|---|---|---|
| Outcome KPIs | ||
| Activation rate | Users who activate within 7 days ÷ users who got the onboarding email | Your app sends an event or sets a property such as activatedAt. Workflow Metrics show sends and clicks. |
| Conversion to paid | Accounts paying within 14 days ÷ accounts that got the email | A campaign goal with paid as the target. For workflows, count payments in billing. |
| Upgrade rate | Paying accounts that upgrade within 30 days ÷ paying accounts that got the email | An event at a plan limit starts the upgrade workflow. Upgrades come from billing. |
| Save rate | At-risk customers still paying after 30 days ÷ at-risk customers who got the email | A churnRisk property or a Stripe subscription event starts the workflow. Payments come from billing. |
| Revenue per send | New monthly revenue from recipients within 14 days ÷ emails sent | Sends from campaign and workflow Metrics, revenue from billing, joined on userId. |
| Guardrails | ||
| Spam complaint rate | Spam reports ÷ emails delivered | Spam reports in Metrics. Gmail’s own rate in Postmaster Tools. |
| Bounce rate | Hard and soft bounces ÷ emails sent | Hard and soft bounces in Metrics for every email, transactional included. |
| Unsubscribe rate | Unsubscribes ÷ emails delivered | Unsubscribes in campaign and workflow Metrics. |
| Diagnostics | ||
| Open rate | Unique opens ÷ emails delivered | Campaign and workflow Metrics. A trend only, since Apple Mail can load images on its own. |
| Click rate | Unique clicks ÷ emails delivered | Campaign and workflow Metrics. Not tracked on transactional email. |
Published averages mix industries, list sizes, and opens inflated by Apple Mail, so they make a weak target. Your own history is a better one.
To set a baseline, pick one kind of email, such as a weekly newsletter or one onboarding step. Take the median of each KPI over its last 10 sends, then look into any send that lands far from it. Compare campaigns with campaigns and workflow emails with the same step, since a welcome email and a product update don’t behave alike.
Two lines are fixed: Google’s spam rate limits and Loops’ open rate checks while you build sender reputation, both above. For lifecycle KPIs, the sections below cite public surveys where one exists.
The ten KPIs above measure the email. For SaaS, the payoff is what people do next: new users activate, trials pay, customers upgrade and stay. Five KPIs track those moves: activation rate, conversion to paid, expansion, retention, and revenue per send.
Each rate has its own denominator and window. Activation counts everyone who signed up, while upgrades count only accounts that already pay. Fix both before you send, so this month compares with last month.
Activation rate is the share of new users who reach the first moment of real value in your product, your aha moment, within a set window. Onboarding emails exist to move this number.
In Loops, send an event or set activatedAt when a user activates, and an Audience filter on All following nodes ends the onboarding sequence.
Conversion to paid is the share of trial or free users who start paying within a window after an email. Count the first successful payment in your billing system, not a click on the pricing page.
In Loops, a campaign goal with paid equals true as the target counts this for a campaign, but goals don’t attach to workflows, so count trial workflow conversions in billing.
Expansion is new revenue from customers who already pay: a higher plan, more seats, or more usage. Upgrade emails work best when your product shows a reason, like an account near its plan limit.
In Loops, send an event when an account nears a plan limit to start an upgrade workflow, and count upgrades in billing.
Retention emails reach customers who might leave: usage dropped, a payment failed, or they started to cancel. The KPI is how many are still paying later. List churn from unsubscribes is a different measure, covered in email churn rate.
In Loops, keep a churnRisk property current or start from a Stripe subscription event, as in churn prevention emails and dunning emails.
Revenue per send puts different emails on one scale, so a small upgrade campaign can sit next to a large product update. Count monthly recurring revenue, leave out refunds, and credit each payment to one email.
In Loops, sends are in each email’s Metrics, and the campaign.email.sent and loop.email.sent webhooks tie each send to your userId.
Lifecycle KPIs only count while email reaches the inbox, so check spam complaint, bounce, and unsubscribe rates, KPIs 5 to 7 above, beside every one of them. Look into any rise before it reaches more sends.
Loops shows spam reports, bounces, and unsubscribes in the Metrics for every campaign and workflow email. Transactional emails show sends, bounces, and spam complaints, but don’t track opens or clicks.
Choose the unit, the window, and the attribution rule before you send, then keep them for every report. Count people or accounts, not both. For team products, count accounts, so five teammates don’t become five upgrades.
Tag links with the same UTM pattern every time and join visits to outcomes on a stable user or account ID. Keep personal data out of URLs. When several emails come before one upgrade, credit one by a fixed rule or label them assisted conversions. Don’t add them up.
Attribution shows what happened after an email, not what the email caused. To measure that, hold a random group out of the emails and compare outcomes. The email A/B testing guide covers holdouts.
Connect Claude, ChatGPT, Claude Code, or Codex to the Loops MCP server, then paste a prompt:
The agent reads campaign metrics and workflow email metrics. They’re all-time totals with no date filter, so read them as a lifetime snapshot and join them with your product and billing data for outcomes.
What are good metrics for email marketing?
Start with the ones that show whether email arrives and works: bounce, spam complaint, and unsubscribe rate for delivery and list health, click-through and conversion rate for results, and open rate as a trend only. SaaS teams add activation, conversion to paid, and retention. Judge each against your own past sends.
What are the 5 key performance indicators in email marketing?
Click-through rate, conversion rate, bounce rate, unsubscribe rate, and spam complaint rate. Together they show whether email arrives, gets acted on, and keeps your list healthy. Open rate is left out on purpose, since Apple Mail Privacy Protection inflates it.
What is the difference between click-through rate and click-to-open rate?
Click-through rate divides unique clicks by emails delivered, so it measures the whole send. Click-to-open rate divides unique clicks by unique opens, so it measures the content among people who opened. Automatic opens from Apple Mail make click-to-open rate less reliable.
What are the most important email marketing KPIs for SaaS?
Activation rate, conversion to paid, expansion, retention, and revenue per send. They measure what people do after an email. Track spam complaints, bounces, and unsubscribes beside them as guardrails.
Is open rate still a useful KPI?
Only as a trend and a deliverability check. Apple Mail Privacy Protection loads images as soon as an email arrives, which records opens nobody made. Judge emails by clicks and product outcomes.
What is a good activation rate for SaaS?
A 2022 survey of more than 500 products by Lenny Rachitsky and Yuriy Timen found SaaS products averaged 36% activation, with a median of 30%. That counts all new users, so measure an onboarding email against users who didn’t get it.
What is a good trial-to-paid conversion rate?
A 2023 survey of more than 1,000 products by Lenny Rachitsky and Kyle Poyar put good free-trial conversion at 8% to 12% and great at 15% to 25%. For freemium self-serve products, 3% to 5% is good and 6% to 8% is great.
How do I attribute paid upgrades to lifecycle emails?
Count the upgrade in your billing system, use one window and one attribution rule for every email, and credit each upgrade once. That shows what happened after an email. To see what the email caused, compare with a random group held out of it.
How do I track email conversions in Loops?
Attach a goal to a campaign before you send it. The goal counts contacts who reach a target, such as a paid plan, within a window you set. For workflows, send activation and payment events from your app and count conversions in your own data.
What spam complaint rate is too high?
Google requires a spam rate below 0.3% in Postmaster Tools and recommends staying under 0.1%. Look into any rise before it affects more sends.