A Series A dev-tools company selling to engineering leaders came to us stuck at a 0.8% reply rate, mostly negative replies at that. Ninety days later they were at a 9% positive reply rate and booking 10-12 qualified meetings a month. We changed four things, in this order:
- Narrowed the ICP from 'anyone in engineering' to a specific role, company stage, and tech stack.
- Fixed deliverability: new domains, proper authentication and warm-up, off a burned primary domain.
- Rewrote the opener around a real trigger event instead of a product pitch.
- Added a LinkedIn + phone layer so the email wasn't carrying the whole campaign alone.
None of this was clever. It was disciplined. Here's what each change looked like, including the parts that didn't work and the three weeks where nothing seemed to move.
Change 1: We narrowed the ICP hard
Before: the client was emailing anyone with 'engineer,' 'developer,' or 'DevOps' in their title, from 20-person startups to 5,000-person enterprises. The list was huge and the message had to be generic to fit all of them, which meant it fit none of them.
After: we cut the list to platform and infrastructure leaders (heads of platform, staff/principal infra engineers, VPs of engineering) at post-Series-A to Series-C companies running a specific cloud-native stack the product actually plugged into. The addressable list got about 70% smaller, and far better.
This was the biggest single lever. When every person on the list plausibly has the exact problem you solve, your copy can be specific, and specific copy gets replies. Our full method for this is in how to define your ICP.
Change 2: We fixed deliverability from the ground up
Before: they'd been sending from their primary domain at high volume with no warm-up. Bounce rate was near 8%, and seed tests showed roughly half their mail landing in spam. A chunk of that 0.8% reply rate wasn't disinterest: it was invisibility.
After: we moved sending to fresh secondary domains, set up SPF, DKIM, and DMARC properly, warmed the mailboxes for two weeks before real volume, and cleaned the list to get bounces under 2%. Inbox placement went from about 50% to over 85% on seed tests.
This is unglamorous and it's the step people most want to skip. You can't out-write a spam folder. The exact steps we followed live in our cold email deliverability checklist.
Change 3: We rewrote the opener around a trigger
Before: the first line was about the product: 'We help engineering teams ship faster with...' Every prospect had read a hundred versions of that sentence and deleted all of them.
After: the opener led with a specific trigger we could observe: a recent funding round, a public job posting for the exact infra roles that signaled the pain, or a migration the company had blogged about. The email referenced the trigger, connected it to a concrete problem, and asked a short question. The product came later, if at all.
What didn't work: our first rewrite leaned too hard on flattery, citing things like 'loved your recent post about...', and it read as fake. Replies actually dipped. We stripped it back to a plain, factual reference to the trigger and a direct question, and that's when the opener started earning replies.
Change 4: We added a LinkedIn and phone layer
Before: email-only. If someone didn't reply to the sequence, they were gone.
After: we wrapped the email cadence in a light multichannel motion: a LinkedIn connection and a soft touch mid-sequence, plus a phone call for the highest-fit accounts that had gone quiet. Several of the best meetings came from a prospect who ignored three emails and then picked up the phone.
The channels reinforce each other; a name that's already familiar from LinkedIn gets a warmer answer on a call. How we structure that cadence is in our multichannel outreach sequence.
The 90-day timeline
Honest version: the first stretch was slow, and we told the client to expect that.
- Weeks 1-3: new domains warming, list rebuilt, low send volume on purpose. Reply rate barely moved and it felt discouraging. This is the part clients hate; it's also non-negotiable.
- Weeks 4-6: deliverability stabilized, narrowed list live, first version of the trigger opener shipping. Positive replies climbed past 3%.
- Weeks 7-9: opener refined after the flattery misfire, multichannel layer added. Positive reply rate crossed 6% and meetings started stacking up.
- Weeks 10-12: full motion humming: 9% positive reply rate, 10-12 meetings a month, and a pipeline the founder could forecast against.
Why it worked (and why the numbers were so large)
A jump from 0.8% to 9% is big precisely because the starting point was broken in several ways at once: bad targeting, bad deliverability, and a generic message, all compounding. Fix three compounding problems and the gains multiply.
A campaign that's already reasonably targeted with clean sending has far less headroom, so we don't promise everyone a 9%. What we do repeat is the diagnosis and the order of the fixes. The failure patterns behind that original 0.8% are the same ones we see constantly, and we catalogued them in why cold email campaigns fail.
Frequently asked questions
What changed the client's reply rate the most?
Narrowing the ICP had the biggest impact. They'd been emailing anyone with 'engineering' in their title across every company size. Once we cut the list to platform and infrastructure leaders at post-Series-A companies with a specific stack, relevance jumped and replies followed. Deliverability fixes and the trigger-based opener compounded from there.
How long did it take to see results?
The first three weeks were slow and honestly discouraging: we were rebuilding infrastructure and warming domains, so volume was deliberately low and the reply rate barely moved. The real climb came in weeks four through twelve as deliverability stabilized and the narrowed targeting and new opener compounded. By day 90 the campaign hit a 9% positive reply rate.
Are these results typical?
A move from 0.8% to 9% is strong and not something we promise every client. This account had many fixable problems, which is why the upside was large. A campaign that's already well-targeted with clean deliverability has less room to improve. We share it because the diagnosis and sequence of fixes are repeatable, not the exact numbers.
Want a teardown of your current campaign?
If your outreach is stuck and you can't tell why, we'll do the same diagnosis on your campaign that we did here, covering targeting, deliverability, opener, and channel mix, and tell you honestly where the biggest gains are. No pitch required. Book a call and share your current numbers and we'll walk through them with you.