How we book our own meetings.

The problem

We started Agentic Demand the way most agencies start: with zero clients and no budget to hire salespeople. No SDRs. No outbound team. Just a founder with a decade building demand generation engines for other companies, and a strong opinion about how outbound should work.

The irony was not lost on us. We were building AI-powered outbound systems for SaaS companies, but we needed to fill our own pipeline first. So we did what made sense: we became our own first client.

Building the infrastructure

Before a single email went out, we spent two weeks on infrastructure. We registered three dedicated sending domains (agenticdemand.io, getagenticdemand.com, and tryagenticdemand.com) and set up two mailboxes on each. That gave us six mailboxes sending 30 emails per day each, for a daily capacity of 180 emails.

We only send Monday through Thursday. No weekends, no Fridays. That puts us at 720 emails per week at full capacity. Combined with a 3-email sequence per prospect, we target 150 to 200 new prospects entering the pipeline every week. On the LinkedIn side, we run 20 to 25 connection requests per day Monday through Friday, adding another 400 to 500 touches per month.

None of this happened overnight. Domain warmup took the full two weeks. We started at a handful of sends per day and ramped slowly. Skip that step and your emails land in spam. There is no shortcut.

How the pipeline works

Every prospect that enters our system goes through a multi-stage pipeline. It starts with research. AI agents pull data from LinkedIn profiles, company websites, recent news, and job postings to build a picture of who this person is and what their company is doing right now.

From there, each prospect gets scored against our ICP. We are not just looking at title and company size. We are looking for urgency signals: recent funding rounds, new leadership hires, job postings for SDRs, or tech stack changes that suggest they are actively investing in growth. Prospects without urgency signals get cut. That filtering removes about 60% of the initial list, but it triples our reply rates. It is the single highest-value decision in the whole system.

Once a prospect passes scoring, AI writes a unique email based on the research. Not a template with merge fields. An actual email that references something specific about the prospect's company and role. The difference is obvious when you read them side by side, and prospects notice too.

Before anything sends, every email goes through QA. We enforce 15+ rules on every message: no unverified claims, no fabricated details, no pricing in the first two emails, natural opt-out language, and a hard cap of 150 words. If one verifiable detail per email is the goal, "number stacking" (cramming three stats into one sentence) is the thing we watch for and kill. The QA system also checks for banned words, brand voice violations, and compliance issues. Bad emails get caught here, not in someone's inbox.

After QA, we run deliverability checks: domain health, bounce rate tracking, spam score verification. Then the emails enter a coordinated multi-channel sequence. Email Monday through Thursday, LinkedIn Monday through Friday, timed so they reinforce each other. When someone replies positively, they are flagged immediately and routed into our pipeline for follow-up.

What happened

The first few campaigns were rough. Reply rates were low. The targeting was too broad. The emails were decent but not dialed in. That is normal, and it is something we tell every client upfront. Month one is calibration, not celebration.

By the fourth and fifth campaigns, the pattern started to emerge. ICP scoring got sharper. The QA rules caught edge cases we had not anticipated. Email copy improved because we had real reply data to learn from: what resonated, what fell flat, what got ignored entirely.

At steady state, we are seeing reply rates between 3 and 8% and positive reply rates between 1 and 4%. That translates to 5 to 15 qualified meetings per month. For a bootstrapped agency with no sales team, that is pipeline we would not have without the system.

What we learned