Open rates are a vanity metric, so track these instead

I was looking at our campaign dashboard last September, staring at a 52% open rate, and something felt off. I spent about 15 minutes convincing myself it was legit before I actually looked into it. It wasn't.

Your open rate is lying to you.

It's been lying since September 2021, when Apple shipped Mail Privacy Protection with iOS 15. What MPP does; and I didn't really understand this until I spent a frustrating weekend reading Apple's developer documentation; is pretty simple. When an email lands in Apple Mail, Apple's servers pre-fetch the tracking pixel whether or not the human on the other end ever reads it. The pixel fires. Your ESP records an "open." Everyone celebrates.

Nobody opened anything.

(I wish I were exaggerating. I really do.)

The scale of the problem

So here's where it gets ugly. Apple Mail accounts for somewhere between 48% and 55% of all email opens globally. Roughly 64% of your subscribers are running an MPP-capable version right now; that second number still surprises me every time I look it up, and I've looked it up probably a dozen times hoping it would shrink.

The effect on your data is brutal. I had this one test campaign; same list, same content, same everything. Pre-MPP it tracked at about 28% open rate. After MPP rolled out to our test group, it showed 52%. Click rates were identical. The clicks didn't change because clicks require an actual human tapping an actual link. The "opens" just got inflated by 15 to 20 percentage points of phantom activity. When I first saw that gap in one of our accounts, I spent 47 minutes convinced the data pipeline was broken. It wasn't. The data was accurate; it was just measuring robot behavior instead of human behavior.

Gmail isn't clean either. Gmail's image proxy caches images server-side, so only the first open gets tracked accurately. Subsequent opens from the same subscriber? Invisible. Cached image loads can also fire as "instant opens," recorded before any human could've read a single word. (I could be wrong about the exact caching mechanics here; there's conflicting information depending on which deliverability consultant you ask. But the practical result is the same.)

So your open rate is a mix of Apple bots, Gmail caching artifacts, some real humans sprinkled in, and probably a few other things nobody's fully documented yet. You can't tell which is which. I've tried. I built three different filtering approaches in MailNeo's analytics before settling on the one I have now, and even that one is an approximation.

What to track instead

I should have started paying attention to this sooner. Click tracking is unaffected by both Apple MPP and Gmail's proxy. A click means a real person made a real decision; tapped, scrolled, chose. That's your new foundation. Everything else builds on top of it.

Click-through rate (CTR)

The average sits at 2.09%. If you're hitting 2 to 5%, you're in solid shape; above 5% means your content and targeting are genuinely working. We had one automation flow at MailNeo hit 8.3% CTR last quarter and I took a screenshot of it because I couldn't believe it (it was a back-in-stock notification for a product with a waitlist of 200+ people; so, a bit of an outlier).

CTR has already overtaken open rate as the primary metric for email marketers; about 34% of teams now report on CTR first versus 31% who still lead with opens. I think that crossover happened sometime in late 2024, maybe early 2025. Fast, either way.

Track CTR per campaign, per automation flow, per segment. The segment-level view is where you actually learn something. If your VIP segment clicks at 8% and your general list clicks at 1.2%, that tells you more about content relevance than open rates ever will. This is the report I pull first every Monday morning; it's changed how I write subject lines at MailNeo. I'm probably obsessive about it.

Click-to-open rate (CTOR)

Average sits between 5.3% and 6.81%. CTOR measures clicks divided by opens, so yes, it still uses open data. I know; that seems like a contradiction given everything I just said. It's useful as a relative metric, though, when comparing subject line performance against content performance within the same audience. Think of it as directional, not absolute.

High CTOR but low CTR? Your subject lines aren't pulling people in. Decent CTR but low CTOR? People open but your email content isn't compelling enough to drive a click. I wouldn't report CTOR to leadership as a KPI; it's more of an internal diagnostic tool, the kind of thing you look at on a Wednesday afternoon when you're trying to figure out why a campaign underperformed. Not something you put in a board deck.

Conversion rate

This is the one that pays your salary.

Average email conversion rate: 0.08%. The top 10% of senders reach 0.44%. That gap between average and top performers; 5.5x; genuinely shocked me the first time I ran the numbers. I was in our office, it was late, and I just sat there staring at the spreadsheet for a while. Click-to-conversion also jumped 53% year over year, meaning the people who do click are converting at higher rates than before. Your job is to get the right people to click; the conversion math takes care of itself after that.

If you're not tracking conversion per email, you're guessing about what works.

I've been guilty of this too. It took me an embarrassingly long time to set up proper UTM tracking on every single link; I think it was March 2024 before I finally had it fully implemented across all my automated flows. I'd been putting it off since the previous November. Should have done it sooner.

Revenue per email

Automated emails pull about $2.87 per send. Campaign emails: $0.18 per send. Overall channel ROI lands somewhere around $36 to $42 returned per $1 spent; I've seen estimates as high as $45 but I'm not sure those account for the full cost of the team running the program.

Revenue per email is the metric that kills internal debates. When someone asks "should we invest more in email?" you show them this number and the conversation ends. Track it separately for automations versus campaigns; the difference between those two numbers is staggering, and it should shape where you invest your time.

Actually, that's not quite right. It should shape where you invest your time. Whether it does is another question; I know plenty of teams that see the data and still spend 80% of their energy on campaign blasts because that's what they're used to doing. Habit is a powerful thing.

Unsubscribe rate

Median: 0.22%, up from 0.08% in 2024.

That's nearly a 3x increase. I still find that hard to believe, but I've seen the data from Mailchimp, Klaviyo, and our own MailNeo numbers now; they all tell the same story.

Rising unsubscribe rates across the industry aren't a reason to panic. But they are a signal that subscribers have less patience than they used to. Irrelevant emails get punished faster. If your unsubscribe rate on a specific campaign spikes above 0.5%, something went wrong; bad targeting, misleading subject line, too-high frequency, or some combination of all three.

Watch unsubscribes at the campaign level, not the monthly aggregate. The aggregate smooths out the spikes that actually tell you something. I learned this the hard way after a product launch email in March 2025 that looked fine in our monthly rollup but had a 0.7% unsub rate buried in it. I didn't notice for nine days. By then we'd already sent a follow-up to the same segment. It did not go well.

Building a report that means something

Drop the open rate from the headline number in your next email report. When stakeholders ask about it; and they will, because everyone's been trained to look at open rates for years; explain the Apple MPP situation in two sentences and move on.

I'd lead with these five numbers:

1. CTR by segment

2. Conversion rate (total and per-email)

3. Revenue per email (automated vs. campaign)

4. Unsubscribe rate per campaign

5. List growth rate (net new subscribers minus unsubscribes and bounces)

Five numbers. Real human behavior, not bot activity.

I built MailNeo's reporting dashboard around this exact hierarchy. Open rates are still visible because some teams need them for historical comparisons and A/B testing subject lines, but they're not the headline. They're a footnote. I considered removing them from the dashboard entirely but kept them for now. Might revisit that decision.

Do I think open rates will ever become useful again? Honestly, probably not. Not without a fundamental change to how email clients handle tracking pixels; and Apple, Google, and Yahoo are all moving in the direction of more privacy, not less. I read somewhere that Yahoo Mail is testing similar pre-fetching behavior; I haven't been able to confirm it independently but if it's true, open rate data gets even worse.

If you're still optimizing for open rates in 2026, you're optimizing for Apple's servers. Not your customers. Stop celebrating fake numbers; start measuring what your subscribers actually do.

Worth it? Yes. Annoying to set up? Also yes.