CallPage - five hypotheses in five weeks
Taking a new product to market - an AI voice agent. 41 leads in the first 5 weeks of sending.

| 41 | Leads in 5 weeks |
| 5 | Hypotheses tested in parallel |
1. About the client
Among other things, CallPage provides a website widget that connects visitors with a salesperson by phone - the visitor leaves a number and the company calls back. It also offers call routing and meeting scheduling. More than 1,500 companies use it.
The new product is CallPage AI - a voice agent that answers calls when a person cannot, handles customer questions and books meetings. CallPage was taking it to market and looking for an agency to help launch and scale it. They chose Scaling Labs.
2. Challenge
CallPage started from hypotheses. The new product had its first customers, but the company had little experience selling it and no verified customer profile.
A new product on the market
For most businesses an AI voice agent is a new category. The recipient first has to understand what it is, and only then decide whether it is for them.
A customer profile to verify
The first deployments gave signals, but not enough to point to one type of company that buys most readily.
Many open paths at once
A good customer could be almost any business that receives a lot of inbound calls from customers. Each path needed checking separately.
3. How we take this product to market
Five hypotheses, five segments and a results guarantee
We proposed focusing on five segments and testing each as a separate hypothesis.
Most segments were different kinds of local businesses - companies that get frequent customer calls and have no one to answer them.
On top of the tests we offered a results guarantee and outcome-based billing. Even without a precisely defined customer profile, and with a new product, we know that in selected segments, mainly local businesses, we can deliver results.
A map of several thousand companies per segment
We mapped each segment in full - several thousand companies per segment, with a separate database for each.
We combine company registries, industry registers, Google Maps and business databases. Wherever it was possible, we addressed emails personally to the owner or the person running the business. Every address was verified before sending.
Copy built on each segment's pain points
We worked through each segment - its biggest pain points, its challenges and how the voice agent helps solve them.
A workshop loses the customer who couldn't get through to book a service. A language school loses the parent who called about enrolment. A developer loses the buyer who rang on a Saturday. Each segment got its own copy.
Lead magnets and parallel tests
We tested everything in parallel, including different lead magnets, to see what works best and in which segment.
One of the lead magnets let recipients call the agent and try it for themselves.
Interested replies went to CallPage's team as they came in, with the full email history, in the same thread.
4. Results
Interested replies from companies matching CallPage's profile, counted from the start of sending in mid-October.
Five segments, each with its own database, copy and lead magnets, compared on the same terms.
The first test results showed which segments, messaging angles and data sources work best.
5. Lessons
Three principles from this campaign that we apply when taking a new product to market.
Hypotheses tested in parallel
Five segments tested at the same time give a comparison within a few weeks.
Copy built on the segment's pain point
The same product solves a different problem in each industry. A message about the recipient's specific situation is easier to grasp than a description of a new product category.
The agency carries the testing risk
A results guarantee with outcome-based billing lets the client test several markets at once without paying for hypotheses that do not hold.
Five weeks in, CallPage decides where to scale based on replies from the market.
6. Summary
A new product, five segments and a results guarantee. A full map of each market, copy built on each industry's pain points and lead magnets tested side by side produced 41 leads in 5 weeks, plus first data on where to start scaling.
Let's talk about your future clients.
Talk to us →A new playbook every two weeks for more meetings from prospecting
Strategies from the campaigns we run right now, straight to your inbox.
Unsubscribe in one click.
More case studies
More stories from similar markets.
Mapping a niche Polish vertical and meeting owners for a Nordic PE acquirer
An engagement with a Nordic PE fund evaluating Polish acquisition targets in healthcare distribution: 300+ companies mapped, a target list of 91 and 41 owner meetings booked.
Medidesk - testing over the summer, scaling from September
Building an outbound sales channel in healthcare. From 0 to 30+ qualified leads a month.
SmartLunch in Czechia - cold outreach on a market labelled as “closed”
46+ qualified leads in 6 months on a market the industry called closed for cold email.


