Apple Mail alone accounts for 64.7% of all tracked email opens — more than Gmail, Outlook, and Yahoo Mail combined — and it never renders AMP for Email. This report combines Litmus's May 2026 market-share data with a caniemail.com-verified rendering matrix to show exactly which interactive-email techniques reach which share of a real inbox list.
Methodology
This is a compiled, sourced synthesis, not a new proprietary lab study — every number below traces to one of two public sources, cited in full at the bottom of this page:
- Share of opens by client comes from Litmus's email client market share data, based on roughly 1 billion tracked opens as of May 2026.
- What each client can actually render comes from caniemail.com cross-referenced against MailInApp's own hands-on testing of its kinetic HTML/CSS techniques, last verified July 2026 (see Email client support for the full matrix this table is drawn from).
Both figures are re-verified on a quarterly cadence. If a client's rendering behavior or market share has shifted meaningfully since the dates above, that will show up here on the next review rather than silently going stale.
Share of opens by client
| Client | Share of opens | | --- | --- | | Apple Mail | 64.7% | | Gmail | 24.1% | | Outlook (all versions) | 6.9% | | Yahoo Mail | 2.6% | | Everything else | 1.7% |
One limitation worth stating plainly: Litmus's "Outlook (all versions)" bucket doesn't separate classic Outlook (Windows, Word-rendering-engine) from the newer Chromium-based New Outlook/Outlook.com — and those two render very differently (see the next table). The public market-share data isn't broken down finely enough to say what fraction of that 6.9% is which.
What each client can actually render
| Capability | Gmail | Apple Mail | Outlook classic (Windows) | New Outlook / Outlook.com | Yahoo Mail | Samsung Mail |
| --- | --- | --- | --- | --- | --- | --- |
| Kinetic CSS (:checked interactivity) | ✅ | ✅ | ❌ | ⚠️ partial | ✅ | ⚠️ partial |
| In-email form submission | ✅ web/Android | ⚠️ version-dependent | ❌ | ⚠️ unreliable | ✅ | ⚠️ |
| CSS animation | ✅ | ✅ | ❌ | ✅ | ✅ | ✅ |
| AMP for Email | ✅ registration required | ❌ | ❌ | ❌ | ✅ | ❌ |
| Plain links | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
Effective reach by technique
Cross-referencing the two tables above — adding up share-of-opens for the clients that fully support each technique — gives an honest reach estimate per approach:
- Plain links: 100%. Every client in the first table renders a plain
<a href>. It's the one technique with no asterisk. - AMP for Email: ~27% (Gmail 24.1% + Yahoo Mail 2.6% = 26.7%, rounded). Registration-gated on top of that — a sending domain has to pass Google's review before any recipient sees an AMP message at all.
- Kinetic CSS: at least 91.4% (Gmail 24.1% + Apple Mail 64.7% + Yahoo Mail 2.6% — the three clients marked fully ✅ in the capability table above). The real number is higher once New Outlook/Outlook.com and Samsung Mail's partial support is counted, but neither is broken out separately in the public market-share data, so this report states only the floor it can actually prove.
The gap between 27% and 91.4%+ is the entire argument for building interactivity on kinetic HTML/CSS instead of AMP for a general-purpose recipient list — see Why not AMP for Email? for the full trade-off, including the cases where AMP's higher ceiling in its narrow lane is still worth it.
Outlook's slow transition
Microsoft has been migrating classic Outlook users to the Chromium-based New Outlook through 2026–2027. Until that migration completes, both engines are real, current targets — a technique aimed only at "New Outlook" today ignores a meaningful and still-shrinking slice of the classic-Outlook base, and vice versa. Conditional comments (<!--[if mso]-->) are only honored by the classic Word rendering engine, which makes them a reliable way to target it exclusively without also catching New Outlook.
Sources
- Litmus — Email Client Market Share — ~1 billion tracked opens, May 2026.
- caniemail.com — per-client CSS/HTML rendering support data, cross-referenced with MailInApp's own hands-on testing, last verified July 2026.