A Complete Cold Email Campaign Built in 15 Minutes With Claude Code
From raw company list to a live campaign draft in 15 minutes using Claude Code, Apollo, Lima Data, and Instantly.
Continue exploring this workflow
The Strategy
Cold email used to take entire afternoons: manually scoring leads, writing copy, verifying addresses, and uploading campaigns. Claude Code just made 80 percent of that work irrelevant overnight. Alex Vacca, founder of ColdIQ ($7M+ ARR, Clay's number one global partner), has sent over 23 million cold emails across more than 2,000 campaigns and he ran the full workflow live in 15 minutes. He breaks cold email into three eras. The first was brute force with SDRs manually verifying emails and writing every message by hand. The second was Clay, which centralized enrichment and cut days of work into hours but still required a human to manually configure each workflow through a visual interface. The third era is Claude Code connecting directly to APIs, taking natural language prompts, and executing the entire workflow from lead scoring to copywriting to campaign creation without a GUI bottleneck. The setup at ColdIQ uses three markdown files that encode institutional knowledge: a copy framework with rules for cold email (body copy between 50 and 80 words, specific CTA structures, subject line formatting), a scoring criteria file that classifies companies into Tier 1, 2, and 3 based on industry fit, employee count, funding stage, revenue, and tech stack, and instruction files that tell Claude Code how to sequence the steps. Three APIs are connected: Apollo for contact discovery, Lima Data for email verification, and Instantly for campaign creation and sending. The total monthly cost is $200 for Claude Code's Max plan plus existing API subscriptions. Vacca ran the full workflow live, starting with 122 companies and ending with a campaign draft in Instantly in 15 minutes. The same process used to take him an entire afternoon when delivering for clients.
How It Works
Create three markdown files in a folder: a copy framework file with your cold email rules (word count limits, CTA structures, subject line formatting, fluff cutting rules), a scoring criteria file with point values for industry fit, employee count, funding stage, revenue, and tech stack with tier classifications, and instruction files that define the step sequence for Claude Code.
Connect three APIs: Apollo for contact discovery, Lima Data for email verification, and Instantly for campaign creation and sending.
Give Claude Code access to the folder with all three files and API connections.
Prompt Claude Code: "Go through my list of companies and start filtering and tiering based on ICP criteria." Claude Code scores every company against the criteria file and classifies them into Tier 1, Tier 2, and Tier
Companies that do not meet minimum thresholds are filtered out entirely.
Prompt Claude Code: "Find me the right contact person at all of those companies." Claude Code searches Apollo for GTM and sales leaders (Head of Sales, VP of Sales, VP of Revenue) at each qualifying company.
Prompt Claude Code: "Find the correct email address for every single prospect." Apollo finds most emails. Lima Data fills in gaps for contacts Apollo cannot verify. The two data sources layered together produce a 90%+ find rate.
Claude Code pulls the copy framework file and writes personalized email sequences for each prospect. It flags its own issues before you have to, such as catching that a variation does not work for non VC backed companies, recommending fixes, and applying them.
Prompt Claude Code: "Create the email campaign in Instantly directly." Claude Code creates the campaign in draft mode with a two email sequence, three day gap between touches, properly named and organized for review.
Results
Vacca ran the workflow live on 122 target companies. Scoring produced 84 Tier 1, 22 Tier 2, 4 Tier 3, and 12 filtered out. Contact search found prospects at 87 of 106 qualifying companies, pulling 298 total prospects. Email verification produced 85 verified addresses out of 87 contacts, a 90%+ find rate by layering Apollo and Lima Data. Total time from raw list to live campaign draft: 15 minutes. ColdIQ has sent 23M+ cold emails across 2,000+ campaigns and reached $7M+ ARR. Cost comparison: an SDR costs $4,000 to $6,000 per month and builds 3 to 4 campaigns per week. A Clay based workflow runs $500 to $1,000 per month in tooling and still needs hours of human configuration. Claude Code costs $200 per month for the orchestration layer.
Our Take
This is the most concrete Claude Code workflow we have seen for outbound sales. The numbers are specific and the live demo adds credibility. What makes this stand out is not just the speed but the architecture: encoding institutional knowledge into markdown files that Claude Code can reference means the quality of output scales with the quality of your frameworks, not your team size. The cost comparison is devastating for traditional SDR models. The main caveat is that Vacca acknowledges the copy is not award winning yet and still needs human review. But the gap between AI generated and human written copy is shrinking while the gap between teams using this approach and teams doing it manually is growing faster. Best suited for outbound agencies, sales teams running high volume cold email, and anyone currently paying for Clay who wants to understand what comes next.
Frequently Asked Questions
The practical questions a builder or operator is likely to ask before trying a strategy like this.
What does this marketing agencies AI sales agent actually do?
This marketing agencies AI sales agent is a real workflow where the agent takes on an operational job, not just a brainstorming task. A Complete Cold Email Campaign Built in 15 Minutes With Claude Code shows what that looks like in practice. From raw company list to a live campaign draft in 15 minutes using Claude Code, Apollo, Lima Data, and Instantly. The practical value comes from the agent handling repeatable business work with enough autonomy that a human only steps in after context has already been gathered.
Who should use a marketing agencies AI sales agent like this?
This example is most relevant for marketing agencies operators. It is especially relevant for businesses where speed to lead, after-hours coverage, or consistent intake quality directly affects revenue. The category here is Marketing & Sales, which means the best fit is a team looking to turn a manual bottleneck into a repeatable system with a marketing agencies AI sales agent.
Which tools are used in this marketing agencies AI sales agent setup?
The source names Claude Code, Apollo, Lima Data, Instantly, Clay. That matters because one of the strongest signals in this directory is whether the operator shared the actual stack. Named tools make a marketing agencies AI sales agent strategy far more useful than vague claims about “an AI system” doing the work.
How hard is it to implement a marketing agencies AI sales agent like this?
Intermediate difficulty is the current read. The listing suggests a launch window of days. Startup cost is listed as $200+/mo. We were able to extract 9 concrete workflow steps from the source. We would treat a marketing agencies AI sales agent like this as a workflow that needs real business context, testing, and exception handling rather than something you should copy blindly from one prompt.
What results can a marketing agencies AI sales agent produce?
Vacca ran the workflow live on 122 target companies. Scoring produced 84 Tier 1, 22 Tier 2, 4 Tier 3, and 12 filtered out. Contact search found prospects at 87 of 106 qualifying companies, pulling 298 total prospects. Email verification produced 85 verified addresses out of 87 contacts, a 90%+ find rate by layering Apollo and Lima Data. Total time from raw list to live campaign draft: 15 minutes. ColdIQ has sent 23M+ cold emails across 2,000+ campaigns and reached $7M+ ARR. Cost comparison: an SDR costs $4,000 to $6,000 per month and builds 3 to 4 campaigns per week. A Clay based workflow runs $500 to $1,000 per month in tooling and still needs hours of human configuration. Claude Code costs $200 per month for the orchestration layer.
How credible is this marketing agencies AI sales agent case study?
Right now the evidence comes from an X post. That is enough for us to study and curate the workflow, but not enough on its own to treat this marketing agencies AI sales agent like an audited case study. We look for named tools, concrete results, and enough workflow detail to understand what was actually deployed, then we add our own editorial judgment on top.
Related Strategies
More AI agent strategies you might find useful
Fully Autonomous Meta Ads Manager Built on OpenClaw
A fully autonomous Meta ads operation running on OpenClaw for $0 per month that …
SaaStr Replaced a 10 Person Sales Team With 20 AI Agents and 1.2 Humans
A sales team of 10 SDRs and AEs was replaced by 20 AI agents managed by 1.2 huma…
Running a Company Entirely With AI Agents: What Worked and What Broke
Seven AI agents ran a company on a single VPS for a week. They pivoted the busin…
Want more strategies like this?
Get weekly AI agent case studies in your inbox.