For merchant fee audit and interchange specialists

Brands overpaying on card fees, asking for an audit.

We find ecommerce brands with enough card volume for an audit to matter and write to the person who signs off. The pitch: lower fees, same processor.

The problem

You sound likeevery processor.

  • Problem 01

    A crowded inbox.

    Finance inboxes get processor pitches every week, and yours looks the same.

  • Problem 02

    Volume you can’t see.

    Brands below a certain card volume can’t pay for an audit, and you can’t see volume from outside.

  • Problem 03

    A hard buyer to reach.

    The buyer is the CFO or controller, a hard person to reach with a generic list.

The brand’s side

The fee nobodyreads line by line.

  • Record fees.

    US merchants paid $198.25 billion in card fees in 2025, up 5.9% from 2024. The average Visa and Mastercard credit fee was 2.36%.

    Sources:Merchants Payments Coalition (opens in a new tab), Nilson Report (opens in a new tab)

  • A small fix, not yet final.

    The 2025 Visa and Mastercard settlement cuts average credit interchange by about 0.1 point for five years, and only after final court approval.

    Source:Reuters (opens in a new tab)

  • Hidden in flat pricing.

    Brands on flat or blended pricing never see the interchange breakdown. Brands on interchange-plus see markups and downgrades they can’t parse.

  • Switching is painful.

    Changing processors means new integrations, payouts and risk reviews, so brands put up with the fees.

Who we target

Brands with enough card volumeto audit.

Segments

  • Brands with high card volume (roughly $5M+)
  • High order values: furniture, electronics, outdoor
  • Wholesale and B2B card payments where Level 2 and 3 data matter
  • Subscription brands
  • Brands with more than one processor

Signals we use

  • Payment apps installed on the store
  • Average order value
  • Wholesale pages
  • Subscription apps
  • Revenue band

Who we write to

  • CFO, controller or head of finance
  • Founder below about $10M
How it works

From your offerto brands that raised their hand.

  1. Step 1: Map

    We pull brands by payment setup, order value and estimated volume, and drop the ones too small for an audit.

  2. Step 2: Pick

    Together we choose the segments where your offer is strongest, and the person to write to at each brand size.

  3. Step 3: Reach

    We write short emails about the one problem you solve and send them from dedicated inboxes at low volume per inbox.

  4. Step 4: Hand off

    Interested brands go to your sales team in Slack or your CRM, with the thread and context. Your team takes the conversation from there.

Sample angles

How we’d open the conversation.

Sample emails show the style we write in for a client. They are illustrations, not sent copy.
  1. Keep your processor.

    “We audit what you pay today and fix it where you are.”

  2. Order value.

    “At a average order, a few basis points is real money.”

  3. Wholesale card payments.

    “B2B card payments without Level 2 and 3 data pay the highest rate.”

What we run on

The tools behind every campaign.

We pay for the data and the sending stack. You don’t buy seats, credits or inboxes.

Brand data

StoreLeads and our own crawl of ecommerce stores.

  • StoreLeads

Account search

  • Apollo
  • Sales Navigator
  • Clay
  • Prospeo

AI qualification

An AI qualification step reads each company’s website and confirms the fit before anyone gets an email.

Contact finding

  • Wiza
  • Icypeas
  • Findymail
  • Enrow

Email verification

Anything older than 60 days gets checked again before it’s sent.

  • MillionVerifier
  • BounceBan

Sending

On dedicated domains we set up and warm for you.

  • Instantly

Operations and alerts

  • n8n
  • Modal
  • Slack
  • Claude

Hand-off

Your CRM or a shared Slack channel.

  • HubSpot

For fee audits we read installed payment and subscription apps and average order value.

Results

What this looks like with our 3PL clients.

Our longest-running clients are 3PLs selling to ecommerce brands. Same buyer, same engine. Names withheld.

We haven’t published results for payment fee auditors yet. These are from 3PLs selling to the same brands.

  • Case study 01Closed deal

    Specialty 3PL with FDA-registered warehouses, US

    • Before: A previous agency produced nine leads in nine months.

    • What we did: Rebuilt the list from their own sales calls, wrote campaigns per product category, and warmed fresh domains for three weeks.

    • After: Qualified brand meetings from the first weeks, and a six-figure brand closed in month one.

    “We spent a ton of money and maybe got one customer out of it over the course of nine months.”

    The client’s CEO, on their previous agency

    Qualified brand meetings from the first weeks, and a six-figure brand closed in month one.

    Specialty 3PL with FDA-registered warehouses, US

    Campaigns from April 2026

    24k+
    Verified targets
    24+
    Active campaigns
    <1mo
    To first closed deal
  • Case study 02Closed deal

    Founder-led 3PL, Winnipeg, Canada

    • Before: Their own cold email had stopped working, and no new brand had signed since March 2026.

    • What we did: Turned the extras they already gave brands for free into one offer, wrote to one product category at a time, and sent from 20 new domains warmed for 21 days.

    • After: 7 brand meetings in the first 5 weeks. One of those brands referred another, and the client started fulfilling for it in early October.

    7 brand meetings in the first 5 weeks. One of those brands referred another, and the client started fulfilling for it in early October.

    Founder-led 3PL, Winnipeg, Canada

    First email August 24, 2026. First brand meeting four days later.

    7
    Brand meetings in 5 weeks
    4 days
    To first brand meeting
    1
    New brand live
  • Case study 03Closed deal

    International 3PL, multi-region

    • Before: No structured way to win new clients. They came from referrals.

    • What we did: Ran outbound campaigns to apparel, pet and home brands.

    • After: Their first client closed from outbound came out of the first three weeks of campaigns.

    See if you’re a fit (opens in a new tab)

    Full case study available on request

    Their first client closed from outbound came out of the first three weeks of campaigns.

    International 3PL, multi-region

    First three weeks of campaigns

    6
    Qualified meetings
    1
    Closed deal
  • Case study 04

    Multi-warehouse 3PL, US

    • Before: Sales relied entirely on inbound, with no outbound system.

    • What we did: Ran campaigns across five product categories, including pet, toys and home, with CRM sync live by week two.

    • After: First qualified meeting within 48 hours of launch, across five product categories.

    See if you’re a fit (opens in a new tab)

    Full case study available on request

    First qualified meeting within 48 hours of launch, across five product categories.

    Multi-warehouse 3PL, US

    First 48 hours of campaigns

    48h
    To first meeting
    1,200
    Daily sends at peak
    5
    Verticals targeted
Compared with the usual options

Why not hire an SDR or a generalist agency?

Already have an SDR?

Put them on the sales calls. We start the conversations.

  • Knows which ecom brands fit your offer

    Hire an SDR
    Learns on the job
    Generalist cold-email agency
    Buys a list after you sign
    Romar AI
    Starts from our own map of classified ecommerce brands
  • Who gets the email

    Hire an SDR
    Whoever they find
    Generalist cold-email agency
    Job-title filters
    Romar AI
    The person who owns the problem at that brand size
  • Sending setup

    Hire an SDR
    One inbox
    Generalist cold-email agency
    Shared or limited
    Romar AI
    Dedicated domains and inboxes, monitored daily
  • Personalization

    Hire an SDR
    Manual, per lead
    Generalist cold-email agency
    AI lines per lead
    Romar AI
    Written for the segment and the problem
  • What happens to replies

    Hire an SDR
    They handle them
    Generalist cold-email agency
    Forwarded to you
    Romar AI
    Interested ones passed to your team with the thread and context
  • Learning loop

    Hire an SDR
    Their notes
    Generalist cold-email agency
    Monthly report
    Romar AI
    Weekly keep/cut decisions from live results
The engagement

What we handleso your team sells.

  • Market map: every brand that fits your offer, sized and split into segments
  • Account check and contact finding, with verified emails
  • Sending domains, inboxes and warm-up
  • Campaign copy for each segment, tested a few variables at a time
  • Daily sending and inbox health checks
  • Interested replies handed to your team in Slack or your CRM, with the thread
  • A weekly review: what got answered, what we cut, what runs next
Who you’ll work with

You work withRodrigo and Margaux.

Romar AI is run by its two co-founders. Rodrigo Zabalaga builds the market map, the sending setup and the testing. Margaux Malarme runs the campaigns day to day and the hand-off of interested replies to your team. The people who design your campaigns are the people you talk to every week.

FAQ

Frequently asked questions.

Can you tell which processor a brand uses?
Often, from the payment apps installed on the store. Not always.
Can you filter by volume?
We estimate it from order volume and order value. You set the floor.
What about Shopify Payments brands?
Flat pricing limits the savings. We usually exclude them or use a different angle, your call.
Do you only write to finance?
At smaller brands the founder decides, so we write to the founder.
What do you need from us?
Minimum volume, fee model, processors you work with, example savings you can share publicly.
What does it cost?
Start with a trial, then a retainer priced on the call. Before the first email goes out, we agree in writing on the target.
Your market map

How many brands are worth auditing?

On the call we size your market from our database, show the first segments we’d test, and the person we’d write to at each one.