The industry has spent five years quietly agreeing that opens are unreliable and then continuing to report them. The 2026 trend data says the shift is finally happening. Here is what to measure instead.
Litmus surveys email marketers every year, and the 2026 edition — published in December 2025 — put its first trend as a shift away from unreliable metrics like open rates, and toward privacy-proofing and long-term consent strategy. That is a polite way of saying the industry has accepted that a number it has reported for two decades no longer measures anything real.
Most programs still lead with opens on the dashboard. That is worth fixing, and not for philosophical reasons. Measuring the wrong metric changes what you do.
Open tracking worked because a tiny invisible image in the message called home when it rendered. Mail privacy features changed the sequence: the image is now frequently fetched by the provider rather than the person, at a time of the provider's choosing, whether or not anyone opened the message. Every one of those prefetches is recorded as an open.
The result is a metric that inflates and, worse, inflates unevenly. The distortion concentrates in whatever share of your list uses those providers, so segment comparisons become meaningless before the headline number becomes obviously wrong. Two audiences with identical real behaviour can show very different open rates purely because of which mail clients they use.
Before any metric goes on a report, ask how it would change if every recipient used a different mail client. If the answer is "a lot", it is not a business metric — it is a measurement of the mail client.
Do not delete the data and do not build on it. Demote it. Opens remain vaguely useful as a directional signal at very large volume, and they are still sound when a click follows — an open that led to a click was definitely a real open. What you cannot do is compare this month to last month, or this segment to that one, and draw a conclusion.
Practically, that means walking into the next review meeting with a different lead slide. Revenue per thousand recipients, by journey. Click rate by template. Complaint rate by segment. If someone asks about opens, the honest answer is that it is now an estimate rather than a measurement.
We stopped reporting opens to a client in 2024. Nothing about their program got worse. Several things got easier, because the conversation moved to money.
The reason this matters beyond one metric is that it is a rehearsal for the next one. Personalisation signals, cross-site tracking, and device-level data are all on the same trajectory: more privacy controls, less reliable telemetry, more pressure to measure things you own.
Programs that build their measurement on their own systems — the order database, the CRM, the click, the renewal — keep working through each of these transitions. Programs that built on observations of recipient behaviour have now done this exercise twice, and will do it again.
Written by the team at LMR — a senior lifecycle marketing studio working in Iterable, Braze, Klaviyo and the tools you already run.
Start a project ↗