Most service businesses that run email automation never actually check whether it’s working — they set it up once and assume it’s fine because emails are still going out. Here’s what the actual numbers say, and what to do with them.

Start With Open Rate — With One Important Caveat

Open rate is still the right first place to look. If people aren’t opening the email, nothing downstream matters yet — the problem is either deliverability (the email is landing in spam or a promotions folder instead of the inbox) or the subject line and sender name aren’t earning the click.

The caveat: open rate has gotten less reliable as a standalone number since Apple introduced Mail Privacy Protection in 2023. Apple Mail now pre-loads images for privacy reasons, which can register as an “open” even when a person never actually looked at the email. It’s still worth checking first — a genuinely low open rate is still a real signal — but click rate and click-to-open rate have become the more trustworthy numbers for judging actual engagement, because a click requires a real human action in a way an automatic image pre-load doesn’t.

For context on where a business actually stands: cross-industry average open rate sits around 19-20%, average click-through rate around 2.4%, and average unsubscribe rate around 0.9% — though unsubscribe benchmarks vary quite a bit by source, with some putting a healthy rate closer to 0.1-0.3% and flagging anything above 0.5% as worth investigating.

The Second Number: Unsubscribe Rate — Checked Two Ways

This one needs two separate reads, not one. First: is the unsubscribe rate trending higher than new subscriber growth? If so, that’s an active, urgent problem — the list is shrinking faster than it’s growing, full stop. Second: if unsubscribe rates are elevated but the list is still growing overall, that’s a quieter but real problem — it usually means the content itself needs work, even though growing list size is masking it from a quick glance at the dashboard.

A Realistic Example

Picture a 4-email onboarding sequence for a home services business. Email 1 opens at 44% — solidly above the industry average. By email 4, the open rate has dropped to 21%, unsubscribes have climbed with each send, and new signups have stayed completely flat month over month. Looked at in isolation, email 1 looks like a genuine success story. Looked at as a full sequence, the pattern is clear: something after email 1 is losing people, and without tracking the sequence as a whole rather than as individual sends, that problem stays invisible indefinitely.

Testing Isn’t a One-Time Event

The most common mistake: run a single A/B test, pick the winner, consider the job done. In reality, B beating A once just means B is the new baseline to beat — not the final answer. Real testing is ongoing champion/challenger: whatever’s currently winning stays live in the sequence while a new challenger gets built and tested against it, on a repeating cycle rather than a one-time event.

Because most service businesses have real seasonality — demand shifts by time of year in ways that affect how content lands — quarterly review is the floor, not the target. Monthly is the actual recommendation. A sequence that performed well in spring can quietly underperform by fall without anyone noticing, if nobody’s looking on a regular enough schedule to catch the shift.

The Sign a Sequence Has Gone Stale

A slow, steady decline across every metric at once — open, click, and reply — usually doesn’t mean email marketing stopped working. It means a message built once and never revisited has stopped matching who’s actually on the list receiving it now.

This article was co-authored by Brandon Sheriff and SPARK, Intelligent Analytics’ AI Intelligence Layer for CRM and marketing automation.