Email Analytics Review Checklist

Check Google Postmaster Tools and Microsoft SNDS for the sending domain and IP. Capture domain reputation, IP reputation, and spam-rate trend over the last 30 days. A drop from High to Medium is the early signal — don't wait for Low.

Run a DMARC aggregate report (EasyDMARC, Valimail, or dmarcian) and confirm pass rates above 98% for authorized senders. Watch for new sub-senders that appeared this period — a forgotten transactional service sending unauthenticated mail is the usual cause of alignment failures.

Hard bounces above 2% or complaint rates above 0.1% (1 per 1,000) trigger ISP throttling. Break out by Gmail, Yahoo, Outlook/Hotmail, and corporate domains separately — Yahoo and Gmail tightened thresholds in February 2024.

Spot-check 5 sends from the period: physical postal address present, accurate from-name, non-deceptive subject, working unsubscribe honored within 10 business days. For Canadian recipients, confirm express consent records exist and identification disclosures are in the footer.

Roll up the prior four checks into a single answer. If reputation, authentication, bounce/complaint, or compliance triggered any concern, mark Yes — the remediation plan step downstream will fire.

Apple Mail Privacy Protection inflates opens for ~50% of recipients. Report opens as a directional metric only and lean on click and conversion data for true engagement. If you have a separate MPP-excluded segment, pull that view as the cleaner read.

Pull CTR and CTOR by campaign from the MAP (HubSpot, Marketo, Klaviyo, Iterable). Compare against your rolling 90-day baseline, not industry benchmarks — the latter are too noisy to act on.

Filter GA4 by source=email and verify the conversion event fires on actual purchase or form submit, not on email-blur or page-load. UTM source/medium drift is the most common reason email conversions look wrong — confirm every send used the documented UTM convention.

An unsubscribe rate above 0.5% per send signals list/content mismatch — usually a re-engagement push to a stale segment. Pull the top three sends by unsub rate and identify the common factor (segment, subject line, cadence).

Healthy = active engagement >25% in trailing 90 days, low bounces. Watch = engagement 10-25% or bounces creeping up. Poor = under 10% engaged or compliance/reputation risk. Poor triggers the sunset and re-engagement workflow downstream.

Pull tested sends from the period and confirm winners reached statistical significance, not just a directional lead. Document the winning patterns (length, personalization, emoji, question vs. statement) in the subject-line learnings doc so they feed next month's tests.

Compare click distribution across hero CTA, mid-content CTA, and footer CTA using click maps in the MAP. If the hero gets <60% of clicks, the offer or copy isn't strong enough — flag for the next creative iteration.

Compare engagement on personalized vs. fallback variants. Watch for token-failure sends where {{first_name}} rendered literally — these crater trust and inflate unsubs. Set fallback defaults on every dynamic field.

Image-to-text ratio above 60% raises spam scores, especially on B2B sends where image-blocking is the default. Confirm alt text on every image, real text for headlines (not flattened images), and color contrast meets WCAG 2.1 AA.

Pull CTR, conversion rate, and unsub rate by ICP persona. If one persona sits at half the engagement of the others, the messaging-house mapping for that persona needs a rework — not the cadence.

Look at recency-frequency segments (engaged-30d, engaged-90d, dormant-180d). Engagement should step down predictably across the buckets — if dormant-180d converts as well as engaged-30d, the segmentation logic is broken in the CDP.

Compare segmented sends against the most recent broadcast control. If targeted variants don't outperform broadcast by at least 20%, the segmentation isn't earning its operational cost — consolidate.

Mark Yes if any persona or behavior segment underperformed, has stale criteria, or no longer maps to a real GTM motion. This triggers the segment refresh planning step in Reporting.

Run the active templates through Litmus across Outlook (desktop + 365), Gmail web, Apple Mail, and Gmail/Apple iOS. Outlook desktop is where templates break — VML fallbacks for buttons and background images are the recurring fix.

50-70% of opens are mobile depending on audience. Send a live test to a real iPhone and Android device — emulators miss tap-target sizing and dark-mode color inversion. Hero text should be readable without zoom; CTAs should be at least 44px tall.

Confirm source=email, medium=email (or marketing-automation), campaign matches the campaign-naming convention, content disambiguates the link position. Drift here is the #1 reason multi-touch attribution reports look wrong at QBR.

For EU and California recipients, click-tracking and open-tracking pixels must respect the recipient's prior consent state per GDPR and CPRA. Confirm the MAP suppresses tracking for non-consenting recipients — OneTrust or Cookiebot integration on the MAP is the usual gating.

Document the specific issue (reputation drop, DMARC failure, bounce spike, compliance gap), the root cause, the fix owner, and the deadline. If reputation is the issue, plan an IP warm-down + dedicated-IP review with the ESP.

Build the re-engagement series for dormant-180d subscribers, then suppress non-responders. Don't sunset silently — Yahoo and Gmail track the gap between sends and complaints, and a sudden suppression of a stale segment improves reputation faster than a gradual re-engagement attempt.

For each underperforming segment, document the proposed criteria change, expected size, owner, and the test plan to validate the new segment beats the old one. Loop the CDP/RevOps owner — segment definitions often live in Segment or the data warehouse, not the MAP.

Refresh the Looker / Tableau / Sheets dashboard with this period's deliverability, engagement, conversion, and segment metrics. Annotate any anomalies inline so the WBR audience doesn't have to ask.

Lifecycle lead or marketing manager closes the review. Capture the overall verdict, the narrative summary that goes into the WBR deck, and attach the dashboard export so future reviewers have the audit trail.