The usual answer is that plain text always wins on trust, but client share data shows almost all opens happen inside one rendering engine, which changes what the test is actually measuring.
The common mistake is treating “plain text beats designed email” or the reverse as a property of the format itself, when it is almost always a property of which rendering engine opened the message. Apple devices alone account for 62.26% of opens, so before any test result gets reported, the real question is how that one engine handled the plain text version against the designed version, because everything else is a rounding error by comparison.
Client share figures make the inbox look less fragmented than marketers assume. Apple devices show up as Apple at 62.26% of opens, Learn more clients (the catch-all category Litmus uses for everything it cannot otherwise classify) sit at 27.03%, Outlook for macOS holds 5.83%, Android and iOS combined hold 2.59%, and the Samsung Mail app on Android accounts for 0.02%. Add those together and almost 90% of opens are explained. That concentration means a plain text versus designed test run on a typical list is really a test of two formats rendered mostly by one engine family, with everything else trailing far behind.
The numbers above rest on Litmus’s published client share breakdown as loaded on October 15, 2026, which groups opens by the rendering engine or app that displayed them rather than by device brand alone.
| Client group | Share of opens |
|---|---|
| Apple devices | 62.26% |
| Learn more (catch-all) | 27.03% |
| Outlook for macOS | 5.83% |
| Android and iOS | 2.59% |
| Samsung Mail app | 0.02% |
When a merchant reports that plain text lifted clicks, the first question should be what fraction of that list’s opens came from each row in this table. A list skewed heavily toward the 62.26% Apple share will behave differently from one with a larger slice of the 5.83% Outlook for macOS share, because that client has its own history of mangling HTML styling. A result that does not break down by client group is a single blended number standing in for several different experiences, and the blend can hide a win in one group and a loss in another.
Quarterly online retail sales moved by 0.4 percent from the first quarter of 2026 in one measurement and by 0.2 percent in another. Both are small figures, well within the range of normal quarterly variation that has nothing to do with email formatting. When a merchant sees a sales shift after switching from designed emails to plain text or back, a shift of that same rough size should raise a question rather than settle one, because the topline retail numbers show movement of that magnitude happening independently of any single channel decision. Crediting an email format change for a swing this small requires isolating what the email channel alone contributed, not reading it off a storewide sales figure.
All opens in the Samsung Mail app on Android measure 0.02%, which is near enough to zero that optimizing a template specifically for that app will not move the needle for a typical list. This is a useful check on how testing time gets spent. A merchant debugging why a designed email rendered oddly in a rarely used client is working on a problem that affects a fraction of a percent of opens, while the larger question of how the email performed across the 62.26% Apple share, the 27.03% Learn more share, and the 5.83% Outlook for macOS share goes unanswered. Prioritizing fixes by the size of the affected client group, using the table above, keeps effort pointed at where opens actually happen.
A merchant running this test should segment results by the client categories in the table above before drawing any conclusion about format, and should treat any revenue change smaller than the 0.4 percent and 0.2 percent swings already seen in quarterly retail data as inconclusive until the email channel’s contribution is isolated separately.
The client share numbers show why that question is hard to answer cleanly. Apple devices show up as Apple at 62.26% of opens, and since almost 90% of opens happen across a short list of dominant clients, so most inboxes render both plain text and designed HTML through the same small set of engines. A test result from one list says more about that list's Apple concentration than about format superiority in general. Anyone citing a plain text win should also report what share of their list opened on Apple versus Outlook for macOS at 5.83% or mobile clients at 2.59%, because the mix changes what the result means.
Outlook for macOS sits at 5.83% of opens, which is small next to the Apple share but still larger than Android and iOS combined at 2.59%. A designed email that breaks in that one client can quietly depress results for a measurable slice of opens even while the headline open rate looks fine. Segmenting results by client, not just by format, is the only way to see that kind of damage.
Not from the numbers available here. Quarterly online retail sales moved by 0.4 percent and separately by 0.2 percent from the first quarter of 2026, and both figures are small enough that normal quarter to quarter noise could explain them without any email change involved. Attributing a shift that size to a format switch requires isolating the email channel's contribution, which these topline retail figures do not do.
They barely register next to the Apple figure. All opens in the Samsung Mail app on Android sit at 0.02%, which is close enough to zero that design decisions aimed at that app specifically will not move results for most lists. Time spent fixing rendering there is better spent elsewhere given how small that number is.
Learn more clients, the catch-all Litmus category, make up 27.03% of opens, which is large enough that a test ignoring it is ignoring more than a quarter of the audience. Combined with the 62.26% Apple share, the 5.83% Outlook for macOS share, and the 2.59% mobile share, these categories account for almost 90% of all opens. A credible test reports results broken out across at least these groups rather than a single blended number.