Compliance, Sold Cold
23 campaigns selling anti-money-laundering software to the compliance officers who buy it.
- Timeline
- 18 months · 23 campaigns
- Role
- Built the workflows and ran the campaigns, with an Outlyst team
- Status
- Client engagement
- Primary stack
- Instantly · Smartlead · Clay · Apollo · LinkedIn Sales Navigator
Headline metrics
- What it tackles
- Selling AML and KYC compliance software to compliance officers across Europe, in a market where every buyer has already been contacted, the decision maker has no fixed job title, and one wrong email lands in a competitor's inbox.
- What it delivers
- An 8.7% reply rate held across 23 campaigns and 18 months, 163 positive replies, and 96 named companies engaged, on a qualification system reused in every outbound build since.
The problem in full
The client is a European compliance-software platform selling the screening, monitoring and customer due diligence tooling that regulated businesses are legally required to run. The buyers are compliance officers at banks, fintechs, crypto exchanges and payment providers across Europe, and that inbox is among the most heavily prospected in B2B. The market is small, the buyer list is well known, and everyone on it has been pitched repeatedly. A newer name has to earn its first sentence before it is allowed a second.
The second problem pulls against the first: the person who buys compliance software has no standard title. There is no VP of Compliance the way there is a VP of Sales. Depending on the company, the decision sits with a Chief Compliance Officer, an AML Director, a KYC Director, a Head of Financial Crime, a Chief Risk Officer, or the COO. Filter on one title and you miss most of the market. Filter on all of them and you email four people at the same company.
And there is one company type you can never contact: another regtech vendor. Selling AML software to an AML vendor is the fastest way to look like you did no research at all, and in a market this small, word travels.
System design
Three workflows in a fixed order, and the order is the design: qualify the company, rank the people inside it, then and only then write something about one of them. Scoring a company from three web pages is the cheap step; pulling every contact and writing a researched line is the expensive one. Running the gate first means the expensive work never fires on a company that was never eligible. The engagement ran with an Outlyst team; I built the workflows and ran the campaigns, from a starting list of about 300 contacts through 23 campaigns.
Company qualification came first. Three inputs per company, the official website, the LinkedIn company profile and the SEO meta description, were aggregated and scored by a model against criteria written specifically for AML and KYC relevance. It returned exactly five fields: SCORE, a 1 to 10 on AML relevance; TYPE, bank, fintech, crypto exchange, payment provider or other; VALUE, an assessment of likely value; EVIDENCE, the specific reason this company needs compliance tooling, drawn from their own content, which forces the model to point at something real rather than return an opinion; and COMPETITOR, a flag that removes rival regtech firms entirely. COMPETITOR is a gate, not a score. A company that trips it leaves the run, and no scoring elsewhere brings it back.
Then the people. Rather than filtering on job titles, the system ranked every available contact at a qualified company against a hierarchy: compliance leadership first, then financial crime, AML and KYC roles, then risk, then the executive line. The three highest-ranked contacts went forward. Where a company had a named compliance function, the top groups carried it; where it did not, the ranking fell through to the Chief Risk Officer and then the COO, so a company with no compliance department still received a correctly addressed email rather than no email at all. That contextual ranking is what let one run address a licensed bank and a 15-person payments firm correctly on the same day.
The copy stayed in human hands. Only the opening line was generated, built from the company profile and the individual's own background, and it referenced something real: one opener cited a Belgian platform's position as a pioneer in ecological crowdlending, another a prospect's role as founding chair of an Islamic finance advisory group. The body and follow-ups were written by hand and held constant across a segment, so any change in reply rate traced to one variable. Even the sender identity was tested rather than assumed: one audience, one message, three variants across three sending accounts. 567 leads, 101 replies, 17.8% combined, and the best variant at 19.6%, which set the sender for every campaign that followed.
The calls that made it work
Rank titles, never filter them
Compliance has no standard buyer title, so a hard filter either misses most of the market or emails four people at one company. Ranking every contact against a hierarchy, compliance leadership down to the executive line, surfaces the right person at each company and falls through gracefully when no compliance function exists. This is the workflow I have reused most; every outbound system I have built since carries a version of it.
Make competitor exclusion a gate, not a score
Emailing a rival regtech vendor burns a small market fast. The COMPETITOR field is binary and final: a company that trips it leaves the run entirely, automatically, and nothing downstream can bring it back. The gate is invisible when it works, which is exactly the property it needs.
Demand evidence, not verdicts
The EVIDENCE field asks the model for the specific reason this company needs compliance tooling, drawn from their own published content, instead of a yes or no. That forces the qualification to point at something checkable, and it hands the copy step a fact worth opening with.
Test the sender instead of guessing
Sender identity was resolved with a controlled test: one audience, one message, three variants, nothing else changed. 567 leads and 101 replies later, the winning variant at 19.6% set the identity for the rest of the engagement. Finding out costs one campaign split three ways; guessing costs every campaign after it.
The results in full
Across 23 campaigns and 18 months the system contacted 17,709 compliance decision makers with 43,168 emails and held an 8.7% reply rate, in a market where the buyer is among the most heavily prospected in B2B. 1,542 people replied and 163 of those replies were logged as positive: 42 interested, 44 who requested a meeting, 1 who booked one, 25 requesting information, 9 opening other opportunities, 26 custom follow-ups and 16 asking to be contacted later. That 163 counts every logged reply except the refusals, which is this study's own convention; counted on stated interest alone the tally is 121. Ninety-six companies engaged that can be named, including Raiffeisen, Aroundtown, Kartesia, Sun Finance, Centtrip and Atos Bank.
The most durable output is the system itself. The qualification gate, the title ranking and the evidence-driven opener were written here first, and every outbound build since carries a version of them. This was also the longest engagement on my register, 18 months of continuous campaigns for one client, which is its own result in a market that punishes careless senders quickly.
Continue reading
One System, Two Payment Rails
Two payments platforms sold to regulated finance across Europe on one shared system.
Outbound in Norwegian
A 25.26% reply rate across 18,062 Norwegian prospects, in a language I do not speak.
The Listing Auditor
A pipeline that audits a brand's live Amazon listing and writes the email about what it found.