Fixing Cold Email Not Working: A 90-Day Agency Case Study

Client background before the outbound rebuild

The agency operates with nine team members and reached $87k in monthly recurring revenue before our partnership. Their clients pay an average of $3,400 monthly with an eight-month retention period. Referrals brought in 65 percent of new business, LinkedIn DMs accounted for 30 percent, and a self-managed outbound campaign generated the remaining 5 percent over eleven months.

The founder concluded that cold email was not working, and his internal data supported that stance. Over the preceding quarter, 14,000 sent messages produced only 47 total responses. Just 11 replies were positive, yielding four qualified sales conversations and zero closed revenue.

He faced three distinct paths. Abandon outbound completely, attempt a thorough overhaul, or pivot to a new target audience. He gave our team a ninety-day window to evaluate the channel and deliver clear answers.

Budget. $8.5k retainer plus tooling. Close rate baseline on warm referrals: 31%. Sales capacity: founder plus a part-time partner, combined 10 calls per week available.

What follows is the actual run. We changed names. We did not change numbers.

First 30 days: auditing and resetting the baseline

Week 1 was diagnosis. We pulled his last 14,000 emails. Open rate sat at 38% (below benchmark). Reply rate at 0.34% (bottom quartile). Bounce rate at 7.2% (broken).

The diagnosis came back fast. ICP too broad ("DTC e-commerce brands $1M-$50M"). Infrastructure on a single sending domain (his agency's primary) for 11 months. Copy starting with "Hope you're doing well, my name is..." across all 14,000 sends.

Week 1-2 rebuild. Three new sending domains, six mailboxes, warmup running by Wednesday. ICP narrowed to "DTC e-commerce brands $5M-$25M ARR, in beauty or wellness, hiring a Content Lead in the last 90 days." Tight. Findable. Actionable. Sequence rewrite by Friday week 2.

Week 3 launch. 1,800 contacts loaded. Volume ramp 30/50/80 per inbox.

Week 4 metrics. 3,400 emails sent. 56% open rate. 28 replies, 9 positive. 6 qualified meetings booked. 1 proposal out. 1 closed-won at $4,200 MRR by day 28.

The founder's first reaction in the week 4 review. "I've been running this for 11 months. I just produced more pipeline in 28 days than the last 90."

Days 31 to 60: testing initial market feedback

Month 2 numbers. 6,200 emails sent. 31 qualified meetings booked. 8 proposals out. 2 closed-won at combined $7,800 MRR.

Channel performance. Email produced 70% of qualified meetings. LinkedIn outbound (layered week 6) produced 20%. Founder content (started week 6) produced 10% by way of warm DMs.

Mid-engagement iteration. Added segment 2 (DTC wellness brands hiring a Senior Marketing Manager) in week 8. Refreshed copy variant on segment 1 in week 7 because reply rate had compressed by 22% from week 4 baseline.

What broke. Week 7 deliverability dipped. Open rate dropped from 56% to 44% on two sending domains. Diagnosis: a specific subject line pattern was triggering Gmail spam filtering on enterprise gmail accounts. Rotated the variant. Open rate recovered to 51% by end of week 8.

By the end of month 2, the founder had stopped questioning whether cold email not working was the diagnosis. The conversation shifted from "is the channel viable" to "how do we scale the variant that's working."

Days 61 to 90: scaling the winning variations

Month 3 numbers. 7,400 emails sent. 26 qualified meetings booked. 14 proposals out (proposals lag meetings by 1-2 weeks). 3 additional closed-won at combined $11,400 MRR.

Compound pipeline effect. By week 10, the founder was getting warm DMs on LinkedIn from prospects who'd seen 3-5 of his posts and self-identified as a fit. Two of those converted to closed deals by day 90 without ever touching the cold email sequence.

Full funnel. 63 qualified meetings booked. 23 proposals out. 6 closed-won. $23,400 in new MRR. Active pipeline at day 90: $185,000 in stages 3-5.

ROI math. Cumulative spend: $25,500 retainer plus $4,200 tooling. Total $29,700. Cumulative closed revenue: $23,400 in MRR (annualizes to $280,800 ARR). Pipeline still in flight: $185,000 ACV. ROI turned positive on day 67.

Content traction. 38 LinkedIn posts published over 90 days. 740,000 cumulative impressions. 32 inbound DMs. 14 of those qualified.

The founder's quote at the week 12 retrospective. "I almost killed the channel. Turns out the channel was fine. My execution was broken."

The core changes that restored campaign performance

Lever 1: ICP tightness. The 1,800-account list outperformed the previous 14,000-account list by 4.2x on reply rate and 6.5x on qualified meeting rate. Same channel. Same retainer. Different ICP. ICP was the single biggest move by far.

Lever 2: infrastructure rebuild. Moving off the primary domain to three secondary domains lifted open rate from 38% to 56% in the first 30 days. The previous domain reputation was salvageable but not from the founder's primary. The lesson: do not run cold email from the same domain you run client communications from.

Lever 3: copy rewrite. Trigger-specific opener (referencing the specific Content Lead hire) replaced the generic "Hope you're doing well" opener. Reply rate jumped from 0.34% to 1.6% inside 30 days.

Lever 4: 2-hour reply SLA. Founder previously responded to positive replies in 18-36 hours on average. Inside 2 hours, meeting book rate jumped from 28% to 41%.

Lever 5: founder content cadence. Three posts per week from week 6 produced the warm DM pipeline that closed two deals by day 90.

Missteps and wrong turns along the way

Three mistakes worth naming.

One. We held the original subject line eight days longer than we should have. By week 5 it was producing flat numbers and we kept it through week 6 because we were optimizing the opener instead. Cost us about a week of stale performance and roughly four qualified meetings.

Two. We added segment 2 in week 8 instead of week 9 or 10. The founder had momentum on segment 1 and wanted to scale. We agreed too quickly. Reply time on segment 1 drifted from 1.6 hours average to 4.2 hours for ten days. Caught it on the week 9 review. Pulled segment 2 to half volume.

Three. The first week of content cadence was generic. The founder wrote three posts about "how to think about creative testing" with no specific examples. Average impressions: 2,400 per post. Once he switched to posts referencing specific shoots, specific numbers, specific outcomes, average impressions jumped to 14,000. Two weeks lost to generic content.

Each mistake is documented so the next agency runs without it.

How to evaluate outbound performance at your agency

Self-identify against the inputs.

If you've concluded cold email not working after 6+ months of execution, your diagnosis is most likely wrong. The channel works. The execution is the variable.

If you're at $50k to $200k MRR, sell to a similar buyer (B2B, 50-500 employees), have a founder who can publish 3 posts per week, and have sales capacity for 8-12 calls per week, the directional shape of this case study should hold within 30%.

If you've been running cold email from your primary domain for 6+ months, the infrastructure is your first lever, not the copy. Most "cold email not working" diagnoses are actually "I burned my primary domain" diagnoses.

If your ICP is broader than 5,000 accounts, the ICP is your bottleneck. Tightening to 1,500-2,500 accounts with a real shared trigger lifts every downstream metric.

This case study is not a guarantee. It's a pattern. Your niche, your offer, your founder visibility, and your sales process all swing the math. The directional shape is consistent across most TDD engagements.

Executing this rebuild on your own terms

Yes. Two prerequisites.

15-20 hours per week of operator focus. The diagnostic, the rebuild, the daily reply triage, the content cadence, the iteration cycles all require attention. Founders who try to fit this into 5 hours per week produce 30% of the case study numbers.

Specialist capability across copywriting, infrastructure, list building, and reporting. The case study agency had founder strength in sales and content, weakness in infrastructure and list building. The execution-only outsourcing fit because we plugged into the gaps.

In TDD's engagements with agency founders, the team is what compresses the 12-week learning curve and prevents the second cold email not working diagnosis. If you have the time and specialist range, run it yourself. If you don't, the math on outsourcing is straightforward.