How we measure email
Every Meisa Research figure comes from real sending data, not surveys. This page explains where the data comes from, how each open is classified and the rules that keep senders and recipients anonymous.
- Updated 22 October 2026
- 71,602 delivered emails
- Selected sample
1. Data
What the studies are built on
Meisa is an email platform that software products use for onboarding sequences, product notifications, newsletters, win-back campaigns and sign-in emails. Meisa Research uses the send and event logs of those products: when an email was sent, whether it was delivered, bounced or drew a complaint, and each open and click with its time and user agent.
The current window is 3 March to 14 October 2026. Of 83,039 attempted sends, 71,602 were delivered with open tracking switched on, to more than 12,000 recipients, producing 59,149 open events. Emails sent with tracking off, bounced emails and internal test sends are left out.
2. Classifying opens
Four sources for every open
| Source | How we spot it | Counts as human? |
|---|---|---|
| Security scanner | The user agent matches a known scanner or link-protection signature | No |
| Apple Mail Privacy Protection | The bare user agent sent by Apple's image relay | No |
| Instant pre-fetch | Any other open within 10 seconds of delivery (sign-in and welcome emails exempt) | No |
| Person | Everything else, including Gmail's image proxy, Outlook, and mail apps on phones and desktops | Yes |
An email counts as opened by a human if at least one of its opens came from a person. The human open rate is those emails divided by delivered emails. Sensitivity checks with no speed rule and with a stricter 30-second rule are reported in the True Open Rate Report.
The bots vs humans study also re-runs Meisa's live scoring rules, which judge every open and click on its own: known scanner signatures, opens in the first two minutes, and link sweeps count as bots, and later activity counts as the person even when a bot opened first.
3. Grouping
Email types and mailbox providers
Email type comes from how the email was sent: an onboarding sequence, a broadcast, a win-back or recovery campaign, a behavioral trigger, a sign-in or verification email, or a direct product notification through the API. Onboarding stages are ordered per recipient by send time.
Mailbox provider comes only from the domain of the recipient address, grouped into Gmail, Outlook.com and Hotmail, Yahoo and AOL, iCloud, and everything else. No other domain is ever reported.
Hours since signup is the gap between the time a contact signed up and the time an email was sent to them, used in the onboarding timing study. Unsubscribes are counted against the last email the contact received in the seven days before.
4. Privacy
The minimum group size rule
We publish aggregates only. A figure appears only when its group has at least 300 delivered emails. Each study is a selected sample: its methodology says how many emails we selected and the selection criteria. No recipient addresses, sender names or product names appear in any study.
Every study states its base, and we refresh the figures every quarter as more data comes in.
FAQ
Common questions
Where does Meisa Research data come from?
From anonymized send and event logs of software products that send their email through Meisa. For the current studies we selected 71,602 delivered emails, 3 March to 14 October 2026. Selection criteria: delivered, sent with open tracking on, and in the study window.
Do you publish data about individual senders or recipients?
No. Only aggregates are published. Every group has at least 300 delivered emails, each study states how many emails it selected and why, and no recipient address or sender name appears anywhere.
Can I use these numbers?
Yes. Quote, chart and republish any figure with a link to the page it came from (CC BY 4.0). Each study has a ready-made citation.
More from Meisa Research
True Open Rate Report 2026
Raw vs bot-filtered open rates for SaaS email, by email type and mailbox provider, with Apple Mail Privacy Protection and security-scanner opens removed.
ReadBots vs humans: how email opens get judged
When scanners strike, how long until the real reader arrives, and why throwing away every bot-touched email loses real readers.
ReadEmail opens by mailbox provider
Raw vs human open and click rates for Gmail, Outlook.com, Yahoo, iCloud and business domains.
ReadThe first 24 hours after signup
How open and click rates change by hours since signup, and why the hot window is one day, not three.
ReadTriggered emails vs broadcasts
Open, click and unsubscribe rates for emails sent on user behavior against list-wide broadcasts, plus launch and deal sends.
ReadOnboarding email benchmarks 2026
Human open, click, unsubscribe and complaint rates for the welcome email, the day-2 nudge, later onboarding emails and win-back.
ReadOnboarding email engagement and activation
How often new users activated after clicking, only opening or never opening their onboarding emails.
ReadShort vs spread-out onboarding sequences
Every onboarding email inside 3.5 days against 2 to 3 emails over 6 to 10 days: opens, clicks, unsubscribes and reach.
ReadSpintax and email open rates
Broadcasts where every recipient got their own variation, against the same sender's broadcasts without spintax.
ReadPhased broadcasts, wave by wave
What happens when a broadcast goes out in waves, most engaged contacts first: opens, clicks and unsubscribes per wave.
ReadTrue open rate calculator
Turn your raw open rate and Apple Mail share into an estimated human open rate in seconds.
Read