Straight version, no spin. Here's what cold email actually drove in month one, how I counted it (and what I refused to count), why it came in low, and exactly what's being built to fix it. Plus something beyond our scope I'm willing to take on, because I'd rather over-deliver through a slow month than explain one.
Net-new paid came in under 10, so per the de-risk clause we set, I'm refunding $1,000 of the $2,000 I billed this cycle. That's the clause working as designed, not a renegotiation — and I'm actioning the refund myself rather than waiting for you to ask.
| Stage | Count |
|---|---|
| Positive replies (cold) | 46 |
| Signed up (cold-attributed) | 7 |
| … paid (the 4 below) | 4 |
| … signed up, not yet paid | 3 |
The 46 cold conversations are low — deliverability + offer fatigue, below. The 3 who signed up and haven't paid are immediate follow-up targets. And across the whole engagement, ~96 cold repliers signed up and never paid — a warm, already-interested pool the follow-up machine and win-back go after.
| Customer | Replied to us | Signed up | How matched |
|---|---|---|---|
| abhishek@reclips.ai | 20 Apr | 18 May | exact email |
| z.zoonglwe@roitriage.com | 06 Jun | 15 Jun | exact email |
| jules@scrapely.co | yes | 04 Jun | direct (known lead) |
| prospeo.io | hugo@ replied | joseph@ onboarded | same-company (colleague replied, company pays) |
Each is traceable end to end — the reply and the signup are both in your own data. Prospeo shows how cold compounds: a colleague (hugo@) replied to our email, the company evaluated, and joseph@ came on and pays.
The contract says: cross-reference paid Stripe customers against our send list, 45-day lookback. The honest problem with month one is that exact-email matching catches very little — people reply from a work address and sign up with a personal Gmail, or we email info@ and the founder signs up. So I looked at whether a looser "same company domain" match would be fair.
It isn't, and I won't use it. I tested the domain-match rate across customer cohorts, including ones who signed up years before we ever ran a single email for you:
| Signed up | Domain-matches our send list | Could we have caused it? |
|---|---|---|
| Pre-2025 | 70% | No — predates our work |
| 2025 | 46% | No |
| 2026, pre-deal | 31% | No |
| Deal window | 37% | Maybe |
Customers from 2023 match our send-list domains 70% of the time. That's not us working — it's that we email the entire cold-email niche, which is exactly who buys Inframail. So a "domain match" proves almost nothing, and the deal-window rate (37%) is actually below the coincidence baseline. I'd rather show you this than quietly inflate the number with it. The 4 above are what I can actually stand behind.
One confirmed cause, and two hypotheses I'm treating as things to test, not excuses.
Positive replies fell off a cliff: ~492 in January to ~68 in June. The root cause was .info domains getting mass-blacklisted on SURBL, which buried sends in spam. It hit hard around 1 June and we caught it later in the month. The .co rebuild fixes this one directly.
We've run essentially the same offer for ~8 months. Offers decay; what pulled in January likely resonates less now. Worth testing fresh angles rather than assuming the same message still lands the way it did.
We've touched much of the cold-email niche ~3 times since January. Even if people don't consciously remember us, a large share have seen the emails before — same offer into the same inboxes means a shrinking pool of people it still lands on fresh. Two levers fix that: sharpen the offer so it hits harder, or widen the audience to operators newer to cold email. Probably both.
These compound — a tiring offer into an over-touched list, then a deliverability hit on top. All three are fixable, and they point at the same plan: new offer angle + fresh audience + the .co rebuild.
| Fix | What it does | Status |
|---|---|---|
| 500 fresh .co domains | Replaces the blacklisted .info infra — clean deliverability | warming (~2 wks) |
| Follow-up machine | Daily surfacing of every interested lead who hasn't converted, so none go cold | live |
| 2 re-engagement campaigns | (1) everyone who replied positively but never signed up · (2) everyone who signed up but never paid | building |
| Unique per-lead tracking | A distinct signup link per lead, so every future paid signup traces to an exact person — no domain guessing, no argument | proposed |
You're sitting on ~1,994 churned customers and ~7,463 free signups who never paid — roughly 9,400 people who already know Inframail. Reactivating even 4% of the churned base is ~80 customers back on the books.
This is outside what we agreed — my mandate is cold acquisition, not your existing base — but I'm willing to take it on and run a win-back across that whole dormant base, using the same follow-up machine and sending infra. It's revenue that's just sitting there, and I'd rather push on it through a slow month than coast. If it lands, we fold reactivation into the engagement properly.
Rather than settle all of this over a doc, here's what I'd propose we walk through live:
Month one came in under the line. I'm telling you straight, showing my math, refusing to pad it, refunding the difference myself, and taking on work beyond scope while I rebuild the channel. That's the version of this I can stand behind.