Email marketing metrics: opens, clicks, and conversions

Choose what your report should measure

Delivery tells you about the receiving server’s response. Recorded opens and clicks describe tracking events. Conversions describe the business action you have defined and measured. Each helps answer a different question.

Before sending, agree on one primary outcome: for example, completed bookings from the campaign within seven days. Then define the event source, attribution rule, and deduplication method. Keep that rule consistent in the final report.

For a campaign selling tickets, the useful progression might be sent messages, delivered messages, unique ticket-link clickers, and confirmed purchasers. Report revenue only when you can join the attributed purchases to reliable order values.

Understand each email metric

• Sent: messages counted as sent under your provider’s definition.
• Delivered: messages with a recorded delivery outcome. This does not establish inbox placement.
• Unique recorded openers: recipients with at least one tracked open during the reporting window.
• Unique clickers: recipients with at least one tracked click, subject to the provider’s counting and filtering rules.
• Total clicks: all recorded click events, including repeat events.
• Conversions: completed actions, counted using the business event and deduplication rule you specified.

Delivery, open, click, bounce, and complaint events are separate event types in systems such as Amazon SES. Check your own provider’s definitions before copying a rate from a dashboard into a client report.

How to calculate email rates

These definitions are a proposed reporting convention. A provider may use a different denominator or event rule. Label the formula you actually use. If the denominator is zero or missing, report the rate as unavailable instead of forcing a percentage.

Example campaign calculations

This example is fictional and is not a SendMailOS customer result. Suppose a campaign has 10,000 sent messages, 9,800 delivered, 2,940 unique recorded openers, 392 unique clickers, and 588 total clicks. Within the agreed attribution window, 49 distinct recipients complete a purchase attributed to this campaign.

• Delivery rate: 9,800 ÷ 10,000 × 100 = 98%.
• Recorded open rate: 2,940 ÷ 9,800 × 100 = 30%.
• CTR: 392 ÷ 9,800 × 100 = 4%.
• CTOR: 392 ÷ 2,940 × 100 = 13.33%, rounded.
• Recipient conversion rate: 49 ÷ 9,800 × 100 = 0.5%.

The 588 total clicks are not 588 people. One recipient can click more than once or use several links. Likewise, one purchaser can place several orders. Keep recipient counts, order counts, and event counts in distinct fields.

Limits of open and click tracking

Apple’s Mail Privacy Protection can download remote content in the background regardless of whether the recipient engages with the message. A tracking image loading therefore does not prove that a person read the email. Apple explains this behavior in its Mail Privacy Protection documentation.

Automated security tools can also generate opening or clicking activity. Mailchimp’s bot-activity guidance describes why filtering matters. Check which filters your reporting system applies; do not assume every click event represents a person.

Use engagement as supporting evidence. Connect it with sessions, replies, and confirmed business events where those measurements are available. CTOR inherits the uncertainty of its open-count denominator, so it should not be the only basis for choosing a campaign.

What to include in your campaign report

Include the campaign identifier, audience definition, send time and timezone, reporting window, count definitions, and primary outcome. For each destination, report unique clickers separately from total clicks. Recipients who click two destinations can appear in both destination counts, so those counts may not add up to campaign-wide unique clickers.

When comparing campaigns, use the same elapsed reporting window, note meaningful audience differences, and disclose tracking interruptions. An unavailable metric is not zero. If delivery looks healthy but purchases are low, investigate the offer and destination journey before assuming the email template caused it.

For your next SendMailOS campaign, write the reporting definitions before launch. If your application receives delivery events, use the developer overview to plan how those events connect to your own records.

Further reading

• Apple: Mail Privacy Protection
• Amazon SES: Email sending events
• Mailchimp: Bot activity and filtering

Sources reviewed on October 23, 2026.