The death of the templated outbound email
How per-prospect generation has quietly killed the mail-merge era and what it means for reply rates.
Mail-merge outbound had a good twenty-year run. Insert first name, insert company, insert one sentence of 'personalisation' scraped from a LinkedIn bio, fire 5,000 a week, expect 0.6% reply. By the start of 2026 that motion is functionally dead in any vertical a property developer operates in. The replacement is per-prospect generation, and it has reset the floor for what counts as outbound.
What changed in eighteen months
Two things, mostly. Generation cost on a per-message basis collapsed by roughly two orders of magnitude through 2025, which made it economically rational to generate a unique 80-word email per prospect rather than substitute three tokens into a template. And inbox providers - Gmail and Outlook both - quietly got much better at clustering near-duplicate sends and routing them to Promotions or worse.
The net effect is that the cost curve and the deliverability curve crossed. Per-prospect generation is now cheaper, in delivered-reply terms, than mail merge.
What 'personalised' has to mean now
In 2023, personalisation meant 'I noticed you posted about X'. By 2025 every prospect had seen that opening a thousand times and it had become a stronger negative signal than no personalisation at all.
In 2026, useful personalisation is structural: the message references the specific scheme type the developer focuses on, the geography they operate in, the average ticket size on Companies House filings, and the financing pattern visible on the last three projects. None of that fits in a merge tag. All of it fits comfortably inside an 80-word email written by a model that has read the prospect's last twelve months of public footprint.
Reply rate numbers
Across Capital Edge accounts in Q1 2026, templated cohorts (the ones we still run as control) sit at a blended 0.4% reply rate. Per-prospect cohorts on the same audience sit at 4.1%. The booked-call conversion off those replies is also higher, because the conversation starts already grounded in what the developer actually does.
That is not a marginal improvement. That is the difference between an outbound channel that works and one that doesn't.
Templated outbound is not coming back. The economics no longer support it, the inbox providers no longer tolerate it, and prospects no longer reply to it. The interesting work in 2026 is at the other end of the spectrum - how much further per-prospect generation can be pushed without losing the structural discipline that makes the channel work at all.
Capital Edge publishes one note a month on UK bridging finance, paid acquisition, and AI-led outbound. Written for brokers, by the team running the playbook.