How OutCheese turned a failed UK pilot into 20+ qualified SQLs for an open banking platform — Case Study
Client Snapshot
Goal
Build a multichannel outbound engine to generate qualified leads, and identify which markets and ICPs were ready to convert into onboarded merchants
Pain Points
Limited internal bandwidth to test multiple markets and ICP hypotheses in parallel
Need to validate which merchant segments and geographies were the strongest fit for Noda's payment platform
No tested outbound playbook for the markets Noda wanted to explore
Results
Phase 1 - UK e-commerce (paused after 2 months)
Week 2 brought one excellent lead, strong enough on its own to cover the entire outreach budget
Then six weeks of nothing - 1 qualified SQL total in 2 months, well below projected volume
We recommended pausing the engagement and handed our segmentation and filtration learnings over to the in-house team, instead of burning more budget
Phase 2 - Baltic markets (were tested in parallel)
20+ qualified SQLs in Low-Risk merchant segments
Multiple merchants signed within 3 months from the SQLs generated
What Made Our Approach Work
Tech stack-based targeting via BuiltWith
For the UK e-com pilot, we identified merchants by their e-commerce platform - Shopify, Magento, WooCommerce, BigCommerce. Each message used the merchant's actual platform as the hook for offering payment integration
Local language, locally validated
For Phase 2, every message was written in Estonian, Lithuanian, or Latvian, and reviewed by a native speaker before sending
Free site scraping to keep enrichment costs low
Scraped target sites ourselves to feed cleaner inputs into AI enrichment downstream, keeping API spend low while improving filter accuracy
Real local senders, not generic agency profiles
Outreach went out from Noda's actual local sales reps' LinkedIn accounts. We tested $750/month rented LinkedIn accounts in parallel - live accounts converted noticeably better, even when their profiles looked less polished
ICP for Phase 2 (Baltics)
01
E-commerce merchants
across Estonia, Lithuania, and Latvia
02
Broader Low-Risk verticals
across all three Baltic markets
Why:
Smaller, more contained merchant landscapes where local presence creates disproportionate trust
Less saturated by international PSP outreach
Noda had local sales reps who could meet in person if a deal called for it
Local-language outbound is rare in B2B here, so messages stood out
Target Personas:
Founders / CEOs
Heads of Finance / CFOs
Payment and operations leads
Buying Intent Signals
Vertical fit: business operating in a Low-Risk vertical Noda's payment platform supports
Local market activity: active sales presence or domain footprint in Estonia, Lithuania, or Latvia
Online transaction volume: evidence of meaningful payment flow worth integrating a new payment partner
Decision-maker accessibility: founder, CFO, or payment lead reachable via LinkedIn or email
Messaging Strategy (A/B)
The shift that moved Phase 2 numbers: writing in the local language, sent from a real local sender. We A/B tested it on the same Latvian audience - same offer, same message structure, only the language changed
A: English language
Initial outreach (Latvian audience)
20% accepted
27% replied
Follow-ups (same audience)
6% opened
0% replied
B: Latvian language
Initial outreach (Latvian audience)
48% accepted (+140%)
47% replied (+74%)
Follow-ups (same audience)
29% opened
24% replied
Sender voice mattered as much as language
Aurimas, Noda's Lithuanian sales rep, wrote his own outreach - longer than we'd usually recommend, and we honestly didn't expect it to land. It did, and it outperformed our shorter copy by a mile. A real local voice beat the "best practice" version.
Implication: even when prospects speak fluent English, a message in their native language reads warmer and less salesy. Pair that with a sender who is actually local, and the lift compounds
What Didn’t Work
UK e-commerce as the opening market - high saturation by PSP outreach without a local sender or language advantage to differentiate
Rented LinkedIn accounts at $750/month - cleaner profiles, weaker conversion
English-only outreach into Baltic markets where local language dominates
Key Takeaways
A failed segment is not a failed channel. UK e-com underperformed; the channel itself wasn't broken. A different market, sender, and language combination delivered 20+ SQLs. Knowing when to stop and re-test is part of the work
Native language is the single highest-leverage variable in compact markets - we measured up to +140% acceptance and +74% reply rate from changing language alone, with even larger lifts on follow-ups
Local trust can outperform perfect-looking infrastructure - real rep accounts beat rented accounts on conversion, even when rented profiles looked more professional
BuiltWith alone isn't enough - 90%+ of raw output is noise without an AI re-filter pass