Building Chopcast a targeting system that produced 12 meetings in 3 weeks, with zero historical data
Chopcast case study
Oriol SerraFounder & CEO
Most outbound engagements start with data. Past campaigns, reply history, a list of angles that worked and a longer list that didn’t.
Chopcast had none of it.
They had a product that sold well once a conversation started, and no reliable way to start one. No campaign history to learn from. No proven angle. No baseline.
That constraint shaped the entire build. When you can’t learn from past performance, you have to encode the targeting logic upfront and let the data do the qualifying before a single email goes out.
Here’s how we did it.
The company
Chopcast turns long-form content into short-form clips. Podcasts, webinars and recorded sessions come in, editable social clips come out.
Their clients are companies already producing long-form content who don’t have the internal resources to cut it up and distribute it.
That detail matters, because it became the targeting mechanism.
The situation
Chopcast wasn’t struggling to close. They were struggling to get in the room.
After an initial surge, growth flattened. Inbound produced traffic, not conversations. They had tried building an outreach engine twice before and neither attempt produced what they expected.
The diagnosis they arrived with was a copy problem. It wasn’t.
Both previous attempts failed for the same reason. There was no system underneath deciding who to contact and why. Without that, copy is just guessing at scale.
Reaching a cold audience with no data and no infrastructure isn’t a messaging exercise. It’s an architecture problem.
What we built
The engine had 6 layers, built in sequence. Each one had to hold before the next was worth building.
This is where the system lives. Everything above it is input, everything below it is output.
The targeting logic
The play came from a simple observation. Chopcast’s buyer is identifiable by behaviour, not by firmographics.
Someone posting a podcast episode or a webinar recording on LinkedIn has already declared the thing that makes them a buyer. They produce long-form content. The only open question is whether they have the resources to repurpose it.
So we targeted the behaviour, not the company profile.
- 1Source people posting long-form content on LinkedIn
- 2Filter for content formats Chopcast can actually work with
- 3Enrich to identify employer, role and seniority
- 4Apply ICP filter on role
- 5Detect company size, exclude companies outside ICP
- 6Classify what the post is actually about (AI layer)
- 7Normalize company names and job titles
- 8Waterfall email discovery and verification
- 9Write the personalization variable
Normalization deserves a note, because it’s the step everyone skips. Raw data gives you “Chief Executive Office at Apple Inc.” If that string lands in a first line, the email is dead and the prospect is burned. Normalizing job titles and company names is what separates personalization from something that reads like a mail merge.
Every record that made it through carried 31 enriched fields. The email was written from those fields, not from a template with a blank in it.
The copy layer
The offer was the lead magnet: we’ll cut your last recording into clips, on us, so you can see how they perform.
That works for one reason. It solves the exact problem the prospect has right now, and it’s verifiable in a week. A guide or a report asks the prospect to believe something. A finished set of clips from their own podcast asks them to look.
The structure held across every send. Specific reference to their content, a relevant observation about the format, a question about their constraint, then the offer.
High personalization is easy to claim and expensive to produce manually. The point of the enrichment layer is that it’s produced automatically, at volume, without a human writing 900 first lines.
The results
1,894 emails sent. 76 replies at 7.6%. 18 of those positive, which is 24% of everyone who replied.
That last number is the one that matters. Reply rate tells you the list was accurate. Positive reply rate tells you the offer was relevant to the person who got it. A campaign can produce replies and still be wrong about who it’s talking to.
12 meetings in the first 3 weeks, worth $70,000 in pipeline.
Chopcast doubled MRR over the following 4 months. The system produced the conversations. Their team closed them.
What they own now
The campaign is the visible part. The asset is underneath it.
Chopcast ended the engagement with a sourcing and enrichment engine they can point at a new segment without rebuilding anything. The targeting logic is encoded in the tables, not in someone’s head. The infrastructure is warm. The copy structure is documented and tested.
New angle, new vertical, new market: the work is a re-run, not a fresh project.
That’s the difference between hiring a campaign and installing a system. One stops producing the day it’s turned off. The other keeps producing as long as someone maintains it, and the maintenance is testing copy and managing the inbox, not rebuilding the machine.
Who this is for
Companies that close well and can’t reliably get in the room.
Where inbound produces traffic but not conversations. Where previous outbound attempts underperformed and nobody can explain exactly why. Where there’s no campaign history to optimize against, so the targeting has to be right the first time.
If the problem were copy, a copywriter would fix it. It usually isn’t.
Want this built for you?
30 minutes. We map where your GTM is leaking and what to fix first. You leave with the answer whether you work with us or not.
Request a Strategy Call
