Every outbound deck I’ve seen in the last two years still has an “open rate” line. Every one. And every one of them is quietly lying about how well the sequence is actually working.
Why open rates are theatre
Since Apple Mail Privacy Protection shipped in 2021, roughly half of your inbox opens are prefetches — Apple’s servers pulling the tracking pixel on behalf of the user, whether they ever look at the email or not. Outlook does variants of this. Corporate mail scanners do it. Gmail’s link-preview crawlers do it.
That doesn’t mean open rates are useless. It means they measure “did an inbox client somewhere on the internet touch this email”, which is not the same thing as “did a human read this”. Yet we still hire, fire, and greenlight campaigns on them.
What I actually track
Three things, in this order:
Reply rate. A human writing back is unfakeable. If a sequence drives replies, it’s working — even the negative replies (“not the right person, try X”) are actionable data. This is the only vanity-proof top-of-funnel metric I trust.
Meetings booked per 100 sent. The only conversion that ties to revenue without a debate. Roll it up per persona, per sequence variant, per week. Everything downstream is CRM data.
Time-to-first-touch. How fast does a real reply get to a human on our side? If the answer is “the SDR checks the inbox once a day”, the sequence is bleeding.
Open rates as a diagnostic, not a KPI
I’ll still glance at open rates when something looks broken — a sequence that used to work suddenly reply-rates at zero, and open rates crashed too? Probably a deliverability issue, not a copy issue. That’s legitimate diagnostic use.
But as a leaderboard metric, tracked in Slack every Monday, celebrated in QBRs? No. It’s the outbound equivalent of measuring website “time on page” and calling it engagement.