Metrics defined
Every analytics report shows numbers and percentages. This article defines what each means so you can interpret your reports accurately.
Every analytics report shows numbers and percentages. This article defines what each means so you can interpret your reports accurately.
Send-level metrics
Sends — total number of emails sent for an issue (or across a date range).
Delivered — emails that successfully reached the recipient's mail server. Doesn't necessarily mean inbox (could be spam).
Bounces — emails that failed delivery.
- Hard bounce — permanent failure (invalid email, deleted account, domain doesn't exist). Suppressed automatically.
- Soft bounce — temporary failure (full inbox, server issue). rasa.io retries.
Bounce Rate — bounces ÷ sends, expressed as a percentage. Healthy is under 2%; 5%+ is a warning sign.
Open metrics
Opens — total times the email was opened. One person opening twice counts as two opens.
Unique Opens — number of unique recipients who opened. Better measure of audience reach than total opens.
Open Rate — unique opens ÷ delivered, as a percentage.
Important note about opens: Open tracking has gotten less reliable since Apple's Mail Privacy Protection (which inflates iOS opens). Treat open rates as directional, not absolute. Click rates are more trustworthy.
Click metrics
Clicks — total clicks on any link in the email.
Unique Clicks — number of unique recipients who clicked at least one link.
Click Rate (CTR) — unique clicks ÷ delivered, as a percentage.
Click-to-Open Rate (CTOR) — unique clicks ÷ unique opens. Measures how often openers actually engaged with content. Useful for evaluating content quality independent of subject line performance.
Suspect clicks
Suspect Click — a click that's likely from a bot, scanner, or corporate security tool rather than a real human reader.
Most reports have a Without Suspect Clicks toggle so you can choose whether to include or exclude them. See Suspect clicks for more.
Article and content metrics
Article Clicks — clicks on links to specific articles.
Top Articles — articles ranked by total clicks.
Articles Read — estimated articles a subscriber engaged with (based on click patterns).
Avg. Articles — average articles per subscriber per send. Higher numbers usually indicate deeper engagement.
Subscriber metrics
Active Subscribers — subscribers who have opened or clicked recently (default range: 30 days).
Subscribed — date a contact was added or opted in.
Last Activity — most recent open or click for a specific contact.
Last Open / Last Click — date of last specific action.
List and audience metrics
Total Contacts — everyone in your database.
Total Subscribers — contacts opted in to at least one subscription type.
Total Recipients — for a specific send, the count of contacts who received it (excluding suppressions, exclusions, etc.).
Unsubscribe metrics
Unsubscribes — contacts who clicked the unsubscribe link or otherwise opted out during a send.
Unsubscribe Rate — unsubscribes ÷ delivered, as a percentage. Healthy is under 0.5%; 1%+ is a warning.
Spam Complaints — recipients who marked the email as spam. Auto-unsubscribes them and is a strong negative signal to inbox providers. Aim for 0; 0.1% is concerning.
Engagement segments
Highly Engaged — subscribers who open and click frequently.
Moderately Engaged — subscribers who open occasionally.
Inactive — subscribers who haven't opened or clicked in a long time (typically 90+ days).
For more on managing inactive subscribers, see List hygiene and cleaning.
Comparing metrics
Different metrics tell different stories:
- Open rate dropping but click rate stable — subject line issue or open tracking issue
- Open rate stable but click rate dropping — content quality issue, or maybe deliverability suspicion
- Bounce rate climbing — list hygiene problem or new bad import
- Spam complaints rising — content or frequency mismatch with expectations
- Unsubscribes rising — content drift or frequency too high
What's next
- Analytics overview — the full map of reports
- Suspect clicks — the bot filtering metric
- Industry benchmarks — what good looks like