Case Study · Fintech · GTM Engineering

698% Revenue Lift: How Belmoney Scaled with an MBTI-Driven AI Sales Workforce

Results

Revenue Lift

+0%

Closed Deals

+0%

Meetings Booked

+0%

Sales Cycle Duration

-0.0%

Onboarding Fee

+0%

Time Saved

0 hrs/mo

Cost Reduction

€9,500

Operational savings from automation

Time Saved Value

€1,635/mo

131 hours reclaimed per month

Assignment

Belmoney, a fintech platform enabling remittance payments from Europe to the rest of the world, came to us with one directive: bring down Customer Acquisition Cost. They believed marketing was too expensive for the return they were getting.

High CAC wasn't the disease — it was a symptom of a sales engine that needed rebuilding. The story starts with that mandate, and ends with a system that did far more than cut costs.

What we did

Data-Driven ICP

We enriched current customers with 26 data points to find patterns and focus on the highest-leverage audiences. We also built an anti-ICP — a clear profile of leads you should remove from the funnel.

AI Response System

With a sales team in Spain and Brazil, and leads across South America, Africa, and Europe, we needed 24/7 coverage. The AI response agent took over Email and WhatsApp — and booked 26 meetings on its own after launch.

Automated Sales Coaching

To fix the closing rate, we automated the post-call workflow. Meeting transcripts were analyzed to auto-fill CRM fields (saving time) and provide AI sales coaching, helping the team improve their pitch with every interaction.

Obstacles We Overcame

Relevance AI → n8n

We originally built on Relevance AI, but when they raised prices we migrated to n8n — avoiding long-term vendor lock-in and keeping pipeline debugging under our control.

Attribution blind spots

By analyzing the funnel and running workshops, we realized MQL and SQL attribution could be measured more accurately — and that referrals were not being credited well enough.

Learnings

WhatsApp can be a really valuable channel in B2B.
GHL CRM has easy-to-apply but effective AI agents.
Combining Relevance AI + Make let us wire up integrations fast.
An MBTI meeting-prep custom GPT was helpful for the SDRs.

Jasper's tenacious enthusiasm in GTM engineering is a delight to work with. Internally, we described it as building a 'Ferrari growth engine.'

What stood out most was his willingness to have the hard conversations. We pivoted from Relevance AI to n8n — a hard call — but I am very happy he highlighted how this would prevent long-term vendor lock-in and pipeline debugging issues.

What makes Jasper different is that he has been an entrepreneur himself. He understands that sales is more than understanding your customer; you need the right SOPs, or you are capped at how many leads you can serve well. Building is only half the work — testing and managing these AI agents is an ongoing process.

Walter Lopez

Growth and RevOps Manager at Belmoney