Cut sends by 83% and double revenue on replies alone
180,000 emails a month from a scorched domain, a 0.4% reply rate, and invoices landing in spam. Rebuilding around reply rate models monthly revenue from $258K to $533K.
| Measure | Before | After |
|---|---|---|
| Monthly revenue | $258K | $533K |
| Added cash flow per month | — | +$275K |
| Reply rate | 0.4% | 7.1% |
| Qualified calls booked per month | 9 | 78 |
| Monthly sending volume | 180K | 31K |
The problem
In this modelled scenario, the previous setup sent everything from the company's primary domain to a purchased list with no verification. These are assumptions of this scenario, not results: when we audit it, the assumed starting point is that the domain is on two blocklists, roughly 41% of sends are not reaching an inbox at all, and — worse — the domain’s transactional email is failing too, so invoices land in spam and collections slip.
What the model found
The ideal-customer profile is not guessed. A model reads every won and lost deal note from the CRM and returns the firmographic and trigger patterns that separated them, which cuts the addressable list by 82% and writes the first line of each sequence from the pattern matching that account.
The plays
Quarantine the primary domain
Move all outbound to four dedicated sending domains with correct SPF, DKIM and DMARC, warmed over 21 days. The primary domain goes send-silent to recover, and transactional deliverability is restored in week three.
Rebuild the profile from won/lost data
An LLM over two years of CRM notes returns the patterns that actually separate closed deals. 96,000 unverified contacts become 17,400 that match, verified and deduped against CRM.
Write sequences worth replying to
One observation about their business drawn from the matched pattern, one specific claim, one low-friction ask. Four steps, no “just circling back”. Testing on reply rate, not opens.
Instrument deliverability
Seed-list inbox placement monitoring, per-domain reputation and per-mailbox volume caps, with an automatic pause on any domain whose placement drops below 85%.
Fix the handoff
Replies routed by intent into the CRM with a four-hour response target. Half the previous program's losses were replies nobody followed up on.
The lever
Sending 83% less. Every metric that matters improves as volume falls, because the constraint was never reach — it was whether the message arrived and deserved an answer.
How the model was calibrated
Sized against Litmus’ sector ROI benchmarks — 36:1 for software and technology, 42:1 for agencies, 45:1 for retail and e-commerce. Those returns are per dollar spent, not per email sent, which is precisely why cutting volume by 83% is the profitable move rather than the risky one: the cost falls, the deliverability recovers, and the replies that remain are worth more.
Sources
- Litmus — State of Email / ROI of Email Marketing — Return per $1 spent by sector: retail, e-commerce and consumer goods 45:1; agencies 42:1; software and technology 36:1.
- Semrush — AI Search & SEO Traffic Study — ~90% of ChatGPT search citations point to pages ranking 21+; AI search visitors convert at 4.4× ordinary organic; 50% of ChatGPT links go to business and service sites.
Want the same read on your category?
Send a brief and we will run twenty prompts before we talk.