How to Get Clients for a Vertical Agency: A 90-Day System

Inside the execution model for how to get clients for a vertical agency

A 14-person team operated in a tightly defined niche. Monthly recurring revenue sat at $94k across 11 active accounts. Their last new client arrived through a referral four months prior.

The founder led every sales call personally. Service delivery was reliable and pricing was sound. The single point of failure was pipeline generation.

Most founders know this pattern well. January brought two new accounts. February yielded nothing. By March, the founder took sales calls with bad-fit prospects simply to protect cash flow.

That's not a slow season. That's a broken system.

The goal of the engagement was specific. Build a 4-channel motion in 90 days. Cold email plus LinkedIn outbound plus LinkedIn content plus conversion assets. Run it weekly. Report the numbers honestly. Replace the referral-only top of funnel with something that produces pipeline whether anyone refers anyone or not. The next ten sections walk through the actual process. Not the highlight reel. The mid-engagement screw-ups too.

Defining the ideal customer profile for a specialized agency

ICP work happened in week two, live, ninety minutes, on Zoom. Vesselin on one side, the founder on the other. Whiteboard open in Miro.

The output was a written ICP matrix. Three segments. Account criteria for each: company size band, geo, tech stack, employee count, vertical sub-segment. Persona criteria: title, seniority, role-specific pain, who else is in the buying committee. Trigger criteria: recent funding, recent hiring posts naming the role, ad spend changes, leadership change.

Why narrow ICP wins for how to get clients for a [vertical] agency: specific pain language, tighter copy, higher reply quality. Generic ICP produces generic copy, which produces 0.4% reply rates and a founder who thinks outbound is dead.

Across TDD's active agency engagements, the engagements that miss the 90-day target almost always skipped or rushed this step. The good ICP work feels slow at week two and pays back at week six.

Cold email copy that generated qualified pipeline meetings

The losing variant ran first. "Hope you're doing well, I run an agency that helps companies like yours scale [outcome] through [vague feature list]." Reply rate at week one: 0.4%.

The winning variant ran from week three. Specific trigger opener (referenced a job post or a funding round). Named pain point in the buyer's exact language. One proof point with a number attached. Soft CTA pointing to a 15-minute call, not a demo. Reply rate at week six on the winning variant: 3.2%.

Subject lines that worked: three words, lowercase, no punctuation. Subject lines that died: "Quick question" and "[First name], grow your [thing]." Both burned in week two.

When The Demand Department runs cold email for an agency client, the copy goes through three founder-led review rounds before it ships. Draft one is rarely the version that goes live. The version that performs is usually draft three, after the founder reads it out loud and rewrites the opener.

Combining LinkedIn messaging and content with outbound email

LinkedIn ran two prongs in parallel. Outbound DMs to the same ICP, different angle. Founder-led content, three posts per week, written for buyers in vertical-specific agencies (swap vertical) specifically.

The compounding showed up in week five. Prospects who had seen two or three of the founder's posts before any DM landed responded at 2.8x the rate of cold-only prospects. Same person. Same offer. Different priming.

A cold email landed Monday. The connection request landed Wednesday. The post showed up Friday. The reply landed the following Monday.

That's not four interactions. That's one motion. The single-channel providers get one shot at being ignored. Four surfaces compound. Across TDD's active agency engagements, this is the part founders underestimate the most going in. The Demand Department's 4-channel GTM motion doesn't run channels in parallel. It runs them as one system pointed at the same buyer, on a schedule the buyer feels but doesn't see.

Sales conversion assets designed to move pipeline forward

Four assets shipped before week six.

Landing page tailored to the specific vertical. One headline, one outcome, three proof points, one CTA. LP conversion on intent traffic: 18%.

Twelve-minute walkthrough video. Founder on camera. Walked through the methodology with screen-shares. Watch-to-book rate from prospects who watched past 8 minutes: 24%.

Case study PDF, redacted client name, real numbers. Sent in the first reply email after a meeting was booked.

Three-email nurture sequence post-booking. Reduced no-show rate from 28% to 11% inside the first month.

The pattern matters. The website wasn't a vanity asset. Each piece had a job and a measured downstream effect. When the post-meeting nurture went live, no-show rate dropped within ten days. That's the kind of cause-and-effect a founder can audit without a marketing dashboard.

Initial outcomes and metrics from the first 30 days

Week 1: kickoff, ICP plan, sender domain procurement.

Week 2: live ICP workshop, TAM file built (4,200 accounts in segment one), copy drafts.

Week 3: launch.

Week 4: 480 emails sent, 16 positive replies, 8 meetings booked, 5 qualified.

First closed deal landed on day 38. $7,400 MRR recurring, on a 12-month engagement. The founder messaged at 9:47pm on a Tuesday. Two words. "We won."

Mid-month iteration: subject lines swapped on day 19 after the first batch underperformed. ICP segment two activated on day 31 because segment one's funding-round trigger was running thin and we needed a second motion to keep volume up without diluting quality.

The point of week-four reporting wasn't celebration. It was calibration. The numbers told us where to push and where to back off.

The compounding returns observed at the 60-day mark

Month two ran wider. Two ICP segments live. Three sender domains warmed. Five sequences active across email and LinkedIn.

Numbers at day 60: 24 meetings booked, 16 qualified, 3 closed, 6 proposals out. Reply rate stabilized at 2.9%. Cost per qualified meeting: $510.

The first inbound signal landed in week eight. Two warm replies came from prospects who had read three or four of the founder's LinkedIn posts before any cold touch. They didn't know we'd been targeting them. They thought they'd discovered him.

That's the moment compounding starts to feel real. Across TDD's active agency engagements, the warm-inbound signal usually shows up between day 50 and day 70 once the content cadence has run consistently for six weeks. Before that, it's outbound carrying everything. After that, content earns its keep.

The final 90-day pipeline numbers and revenue snapshot

Day 90, all-in:

34 meetings booked. 22 qualified. 8 closed deals cumulative across the engagement. Total pipeline created: $280k+ at face value, with $59k in MRR signed across the eight closed deals.

Cumulative engagement spend: $27k. Retainer plus tooling. ROI crossed positive on day 58. By day 90 the math wasn't subtle.

Content traction: 280k impressions across 36 posts. 42 inbound DM conversations. 7 of those converted to qualified meetings, 2 closed.

The cumulative pipeline chart has three lines. Cold email starts strong and plateaus. LinkedIn outbound ramps slower then steepens at week six. Content-sourced warm starts flat and bends upward at week eight.

Three lines. Stacked. Compounding. That's what a working motion looks like on paper.

Mid-campaign adjustments that improved response rates

Three iterations mattered.

Week 7, sending domain rotation. Deliverability dropped on the original three domains (one mailbox got flagged after a vacation auto-responder caused a sending pause and a re-warmup loop). Two new domains got spun up over the weekend. Reply rate recovered inside four days.

Week 8, micro-ICP segment added. Inside the broader vertical, a 200-account underserved sub-niche surfaced from a Clay enrichment run. Specific copy, specific angle. Highest reply rate of the engagement at 4.1%.

Week 10, content framework restructured. The first six weeks ran "tips" posts. They got impressions, not DMs. Week 10 forward, posts shifted to "stories from inside the agency." Same author, same hour, different format. DM volume tripled.

None of those iterations were guesses. Each one came from the weekly ops review where the numbers were already telling us what to do.

Practical playbook elements you can implement today

Steal the ICP discipline. Live workshop, written matrix, three segments, refused to launch without it.

Steal the 4-channel motion. Cold email plus LinkedIn outbound plus LinkedIn content plus conversion assets. Not email-only. Not LinkedIn-only. The compounding only happens when all four feed each other.

Steal the copy approach. Trigger opener, named pain, proof point, soft CTA. Test subject lines weekly. Kill anything under 1.5% reply rate by day 14.

Steal the iteration cadence. Weekly ops review. Three numbers. What worked, what didn't, what changes Monday. No quarterly retros. Every Monday.

If you're running vertical-specific agencies (swap vertical) and your pipeline is referral-dependent, this playbook closes the gap. The engagement details are here because they're replicable. Some of you will run them yourselves. Some of you will hire The Demand Department to run them. Either way, the math is the same.