Valle del Sol × ChangeEngine — Email Engagement QBR

Opens use the platform definition (a recipient who opened or clicked counts as an open — click-without-open happens with image blocking). The headline click rate is link-only and scanner-filtered: computed over moments known to contain a link (≥1 recorded click) and excluding clicks landing ≤60 seconds after delivery (security-gateway scanner signature; 20.1% of all recorded clicks land in that window). Raw variants are shown beneath each tile. Engagement Score = 0.3 × open rate + 0.7 × scanner-filtered click rate.

Scope decision: 2 personal-email moments (178 deliveries, 0.4% of all engagement rows) were excluded per the request. No other channels exist in the export.

Deliveries by month

Open and click rates by month

Monthly rates use per-delivery denominators: a recipient of a recurring moment (e.g. Time Sheets reminders) counts once per delivery, in the month delivered. October 2026 covers 9 days. 2025 coverage starts June and is thin until November (platform ramp-up).

Bars show Engagement Score (0.3 × open rate + 0.7 × scanner-filtered click rate). Hover a row for deliveries and rates. Department is still missing on ~19.5% of deliveries, bucketed as Unknown, not dropped — the HRIS join shows these are almost exactly the recipients who are not in the current employee roster (likely departed staff), so the Unknown bucket largely reads as "former employees". Management level has two values (Employee / Manager); where the HRIS roster disagrees with the engagement file, HRIS wins (10 people corrected). Tenure bands were validated against HRIS tenure_days (99.6% agreement). Programs under 200 deliveries are grouped into the table but read them with caution.

Top 10 by Engagement Score

Bottom 10 by Engagement Score

Per-moment numbers are per-delivery and match the platform. Ranked tables require ≥25 deliveries (test sends excluded). Scanner marks the 5 moments where click bursts match a link-scanner signature — their click metrics are inflated even after filtering; read with caution. Recurring marks scheduled/automated series. CTOR = raw clicks ÷ opens.

Broadcast sends only: 25 automated/triggered series (work anniversaries, start-date journeys, timesheet reminders) and 2 one-recipient test sends are excluded. Hours are the scheduled local send hour. Cells mix different content and audiences — treat as directional, not causal.

Moments sent per month

Moments by day of week

Each moment is counted once, on its first send day. Recurring series therefore appear once, not per repeat delivery.

New hires by month (current roster)

From the HRIS employee file (snapshot Oct 5, 2026; 269 rows). It is a current-roster snapshot: people who already left are absent and the termination fields are empty apart from 4 scheduled voluntary departures later in October 2026, so historical attrition by month still cannot be computed — no leave dates or types exist for past leavers. The 155 email recipients not in the current roster (37% of all-time recipients; their last delivery is typically around Mar 2026) are consistent with departures during the window, but that is an inference, not measured attrition. The same survivor bias means hire counts for earlier months understate actual hiring. Recipient-level comparisons stay on the four agreed cuts; no demographic cuts are shown.