The list is healthy and the top of the funnel works — a 29% open rate on a 15.9M-send volume is a brand people still want to hear from. Everything breaks at the click. A 0.86% click-to-open means that for every 100 people who open, fewer than one acts. The emails are earning £4 per thousand sent when the same brand's tighter, product-led sends earn three to four times that. The database is being treated as one undifferentiated list of 15.9M — the four economies the CRM-OS defines (Champions, Mid-Tier, At-Risk, Dormant) are written down but not being used. The revenue isn't missing because the audience is cold. It's missing because the mechanics after the open are built for broadcast, not for a person.

Nearly half of all sends earn under £3 per thousand. A handful of product-led and tight-list sends clear the US benchmark. The account average sits low because the losing half drags it down — not because there's no ceiling.

29.1% average open across 15.9M sends is genuinely strong for a base this size. Deliverability is clean (99.7%), hard bounces are negligible, and unsubscribe sits at 0.20% per send. The brand equity is intact. This is the foundation everything else gets built on — do not burn it with more undifferentiated volume.

Product-led sends out-earn editorial £5.17 vs £2.94 per thousand — a 1.8× gap. Comparing like-for-like mass sends (50k+ recipients), the top earners are almost entirely product-led; the bottom are almost entirely story/editorial.

Opens are near-identical top and bottom (27–29%). The entire gap is in the click and the offer — the story sends open fine, then give the reader nothing to do. The rule isn't "editorial is bad": it's that product must arrive as the answer, or the open never becomes a click.

You don't need a test to prove segmentation works. It's already visible in the regional splits:

And the UK "Part 2" resends to prior openers hit 82–88% open rates. Warm, relevant, targeted sends are already winning inside this account. The playbook isn't unknown — it's just not being applied to the 15M UK volume.

29% open, 0.86% click-to-open. Only 0.031% of opens become an order. The subject lines are doing their job; the email body, the CTA, and the offer are not carrying the opener across the line. In Central Brain terms, the emails win the trigger but never build the desire or close the action — they inform and then send people away instead of making one clear, rewarded ask.

The CRM-OS is explicit: the base is four psychological economies, each needing different language, timing, incentive and offer. The email data shows the opposite — 45 of 61 sends went to 250,000+ recipients, near-universal blasts. Champions are getting the same generic editorial as Dormant. That is the single biggest structural miss, and it's costing money at both ends:

31,911 people unsubscribed over the period — a permanent asset being spent to send low-relevance mass editorial. Every blast to the 88% Dormant pool that isn't sharply reactivation-led is trading a future customer for near-zero revenue today. The bottom 10 sends (all 50k+, all editorial/story) earned £0.86–£2.05 per thousand — they cost list health and returned almost nothing.

Best-day-to-send logic (purchase intent on the eve of the peak buy day; awareness earlier in the week) doesn't appear to be governing the calendar. Sends are cadence-driven, not intent-timed. This is a fixable overlay, not a rebuild.

This isn't a cold-audience problem, so the fix isn't more sends. The UK book runs at £3.91 per thousand; the tighter US sends run at £7.64. Modelling the UK volume (15.7M sends, £61,315 direct over the 6.5-month period) at improved rev/1k, annualised ×1.85:

Assumptions, stated plainly: US rev/1k is treated as the achievable ceiling because it's the same brand, same products, tighter list; the conservative case assumes we close only half that gap; annualisation scales the 6.5-month actual to 12 months at flat volume. No extra sends, no extra media — this is reallocation and segmentation of what already goes out.

This export contains sends, opens, clicks and revenue — not a per-send CRM segment field. So "Champions are getting generic blasts" is inferred from send size (a 340k send cannot be segment-targeted), not read from a segment label. It's a safe inference, but to make it airtight for the client we'd confirm it with the DotDigital audience/segment attached to each campaign. Everything else in this document is computed directly from the file.