Email Analytics: The Metrics That Actually Matter
Beyond Open Rates
Open rates have been unreliable since Apple's Mail Privacy Protection (MPP) launched in 2021. MPP pre-fetches email content for Apple Mail users, inflating open rates by 20-40%. Relying on open rates as your primary metric leads to poor decisions.
Instead, build a metrics framework that captures the full email funnel:
- Deliverability — Are your emails reaching the inbox?
- Engagement — Are subscribers interacting with your emails?
- Conversion — Are email interactions leading to desired actions?
- Revenue — What is the financial impact of your email program?
Each layer builds on the previous one. Perfect engagement metrics mean nothing if 30% of your emails are landing in spam.
Deliverability Metrics
Monitor these to ensure your emails are reaching the inbox:
- Inbox placement rate — percentage of emails that land in the primary inbox (not spam, not promotions). This is the most important metric most companies do not track.
- Hard bounce rate — should be under 0.5%. Higher indicates list quality issues.
- Soft bounce rate — temporary failures. Monitor trends — increasing soft bounces may signal reputation issues.
- Spam complaint rate — must stay under 0.1% for Google, 0.3% for most other providers.
- Unsubscribe rate — healthy range is 0.1-0.5% per campaign. Higher suggests content-audience mismatch.
- Authentication pass rates — SPF, DKIM, and DMARC should all be at 100% for your legitimate sends.
Modern ESPs like Brew provide these metrics in their built-in dashboards. For deeper analysis, supplement with tools like Google Postmaster Tools (for Gmail-specific data) and DMARC reporting services.
Engagement Metrics
Engagement metrics that remain reliable post-MPP:
- Click rate — the most reliable engagement metric. Calculate as clicks/delivered (not clicks/opens).
- Click-to-open rate (CTOR) — less reliable post-MPP but still useful for non-Apple segments.
- Reply rate — the strongest positive signal for deliverability. Especially important for cold email and 1:1 campaigns.
- Forward/share rate — indicates content quality and advocacy.
- List growth rate — net new subscribers minus unsubscribes and bounces, as a percentage of total list size.
- Read time — available through some ESPs, indicates content engagement depth.
Segment engagement metrics by mailbox provider (Gmail, Outlook, Yahoo, Apple Mail) to identify provider-specific deliverability issues early.
Revenue Metrics
Ultimately, email marketing must drive business results:
- Revenue per email sent — total email-attributed revenue divided by emails delivered. The north star metric for e-commerce email programs.
- Revenue per subscriber — lifetime value of an email subscriber for budgeting acquisition costs.
- Conversion rate — percentage of email recipients who complete a desired action.
- Attribution model — understand how your attribution model (first-touch, last-touch, linear) affects reported revenue.
- Campaign ROI — revenue attributed to a campaign minus costs (ESP fees, design time, content creation).
Be honest about attribution. Email often plays an assist role in multi-touch journeys. Claiming full credit for every conversion a subscriber makes inflates email's contribution and leads to misallocated budgets.
Building Effective Dashboards
Build three levels of email analytics dashboards:
- Executive dashboard — revenue per email, list growth, overall deliverability health. Updated monthly.
- Campaign dashboard — per-campaign performance with engagement and conversion metrics. Updated per send.
- Technical dashboard — bounce rates, complaint rates, authentication pass rates, inbox placement by provider. Updated daily.
Set up automated alerts for anomalies rather than checking dashboards manually. You should be alerted when bounce rates spike, complaint rates exceed thresholds, or engagement drops significantly from baseline.
Review metrics in context — holiday seasons, product launches, and list growth events all affect baselines. Compare against your own historical performance, not industry benchmarks that may not reflect your audience.