Mercury Approval Calculator: How We Built It

Quick answer

Our Mercury Approval Calculator takes four inputs - country of residence, business model, annual revenue, and documentation quality - and returns an estimated approval likelihood informed by our own experience helping non-resident founders apply to Mercury since 2025. The estimate is illustrative and approximate, not audited statistics, and approval is always Mercury's decision. It is a directional estimate, not a guarantee. It is free, needs no signup, and is meant to be used before you submit to Mercury.

Form your Wyoming LLC + EIN + Mercury intro for $397 (Wyoming state fee included)

Why we built a calculator instead of a blog post

Every week, founders email us the same question in different words: "Will Mercury approve me?" They are from Bangladesh, Nigeria, India, the Philippines, Brazil, the UAE. They have a half-built SaaS, a dropshipping store, an agency, a newsletter. They have heard Mercury rejects people, and they do not want to form a US LLC, wait for an EIN, and then get rejected at the banking step.

A blog post cannot answer that question for an individual. The honest answer always depends on the combination of who you are, what you do, how much you earn, and how well you can document it. So instead of writing one more generic "Mercury approval tips" article, we turned our hands-on experience into a small model and put it behind four dropdowns. You input your situation, it returns an estimate, and you can decide whether to fix something (your business description, your web presence) before you apply.

This post explains exactly how that calculator works. We believe in showing our methodology because the alternative - claiming a precise number with no explanation - would be dishonest. A model is only as trustworthy as the data and the assumptions behind it, so here they are in full.

The calculator lives at wyomingllc.xyz/tools/mercury-approval-calculator. It is free and requires no email.

The underlying data

The model is built from our hands-on experience helping non-resident founders prepare Mercury applications, combined with Mercury's published eligibility and prohibited-country guidance. It is an approximate, illustrative tool rather than an audited dataset, and it does not represent a count of facilitated applications. We are a Wyoming LLC formation service for non-US founders; banking preparation is part of what we do, so we have a practical sense of what tends to help or hurt an application.

Conceptually, the model considers four attributes and one outcome:

• Country of residence of the founder
• Business model (SaaS, agency, e-commerce, content, services, crypto-adjacent)
• Annual revenue band at time of application
• Documentation quality (website, revenue history, clean business description - present or absent)
• Final outcome: approved, rejected, or sent to extended KYC review

Anything tied to an individual founder is anonymized and aggregated to the country-and-segment level. No individual founder is identifiable in the model, and we never publish raw records.

An honesty note on confidence. Our sense is firmer for high-volume countries - India, Bangladesh, the UK, the UAE, Brazil - and weaker for small markets. Where we have little to go on for a given country, we say so rather than pretend otherwise. For thin countries, the calculator falls back to a global average instead of inventing a precise-looking number. We would rather show you an honest "we do not have enough signal here" than a fake 71.3%. Either way, approval is the provider's decision and is never guaranteed.

The four inputs, and why these four

We considered the obvious candidate variables and kept the four that, in practice, most affect outcomes. Mercury does not publish its decision logic, so these are our practical observations, not Mercury's internal weights. Here is what each input does in the model.

Input 1 - Country of residence

Country is the single largest driver. Mercury runs sanctions and risk screening on every applicant, and a founder's country shifts the likelihood of approval substantially before any other factor is considered. Approval is always the provider's decision and is never guaranteed. As a rough qualitative ordering:

The bottom row is not our judgment - it is United States law. Mercury, like every US financial institution, must comply with sanctions administered by the Treasury's Office of Foreign Assets Control (OFAC). Founders resident in comprehensively sanctioned jurisdictions cannot be onboarded, full stop. Mercury maintains a public prohibited countries list that reflects these restrictions, and it changes over time. We re-check it against our own outcomes when we recalibrate.

The middle tiers are where the calculator earns its keep. A Bangladesh or Pakistan founder is not auto-rejected - far from it - but is more likely to be routed to extended KYC review, which means additional document requests and a longer wait. Knowing that in advance changes how you prepare.

Input 2 - Business model

Mercury publishes eligibility criteria and a list of industries it cannot serve. Crypto exchanges, money services businesses, gambling, and adult content sit on the restricted side; clean software and services sit on the easy side. Our data mirrors that policy, and the model applies these adjustments to the country baseline:

• SaaS, agency, professional services: baseline applies. These are Mercury's bread and butter.
• E-commerce / dropshipping: baseline −5%. Reviewed more cautiously because of chargeback and fulfillment risk.
• Content creator (YouTube, Substack, newsletters): baseline −2%. Slightly cautious - Mercury wants to see how the business actually monetizes.
• Crypto-adjacent (wallet tooling, analytics, non-custodial): baseline −20%. A restricted-leaning category that needs careful description.
• Crypto exchange / custodial / MSB: baseline −50% or auto-reject. This is policy, not preference.

A note for crypto founders: the issue is rarely "crypto" as a word. It is whether your model looks like a money services business that triggers FinCEN money-transmitter obligations. A non-custodial analytics dashboard reads very differently from an exchange, and describing yours accurately matters more than the category label.

Input 3 - Revenue history

Revenue is a smaller lever than country or model, but it is real. Mercury wants to see a business, not just an idea, and demonstrable revenue is the strongest signal that one exists.

• $0–$10K (pre-revenue or just launched): baseline −5%
• $10K–$50K (early traction): baseline applies
• $50K–$250K (established): baseline +3%
• $250K+ : baseline +5%

If you are pre-revenue, the penalty is small and recoverable - strong documentation (input 4) offsets most of it. You are not locked out for being early.

Input 4 - Documentation quality

This is the input you fully control, and in our data it is the one founders most often get wrong. Mercury's own document checklist lists what you must provide; the calculator captures how well you provide it.

• Business website + revenue history + LinkedIn presence: baseline +5 to +10%
• Business website only: baseline
• No web presence at all: baseline −10 to −15%
• Vague business description: baseline −10 to −20%

The single most common rejection reason we see is a vague business description. "I run an online business" tells a reviewer nothing. "I run a Shopify store selling premium yoga mats to US customers, sourced from a Vietnam manufacturer, fulfilled via ShipBob in Dallas, expecting $15–30K monthly through Stripe" tells them everything. Same business, very different outcome. The documentation input exists to make that gap visible before you submit, while you can still fix it.

How the model combines the inputs

The mechanism is deliberately simple: country sets a baseline, and the other three inputs adjust it up or down. We start from a country baseline, then add or subtract the business-model, revenue, and documentation modifiers, and clamp the result to a sensible range (a high-tier founder is capped well below certainty; a restricted country stays at 0% regardless of other inputs). The output is an estimate, never a guarantee.

We chose an additive, transparent model over a black-box machine-learning classifier for three reasons:

1. It is explainable. Every founder can see exactly which factor helped and which hurt, which is the entire point of the tool.
2. It is honest about its precision. A complex model trained on a few hundred rows would overfit and produce false confidence. A simple additive model with ±5% honesty bands does not pretend to know more than the data supports. The principle here is the same one statisticians use when validating any prediction model - calibration matters more than apparent sophistication. A model is well-calibrated when, across all the cases it scores at "70%," roughly 70% actually get approved. Calibration, not a fancy algorithm, is what makes a probability trustworthy (scikit-learn documentation on probability calibration explains the concept well).
3. It degrades gracefully. When an input is missing or a country is thin, an additive model can fall back to the baseline instead of producing nonsense.

Any accuracy band we cite is only a rough sense of confidence within well-populated segments. It is wider, implicitly, for thin segments - which is exactly why we route those to the global average rather than quoting a tight number.

What the calculator deliberately does NOT do

Being transparent means being clear about the limits. The calculator cannot model:

• Reviewer assignment. Mercury applications are reviewed by humans (and automated checks) whose disposition we cannot observe. Two identical applications can land with different reviewers and different first impressions.
• Current policy state. Mercury tightens and loosens onboarding periodically, sometimes in response to fraud waves or regulatory pressure. A model built on the last 12 months can lag a policy change made last week.
• Application timing and history. If you have applied and been rejected three times in six months, Mercury's risk model deprioritizes you, and our calculator has no way to know your prior history.
• The truth of what you enter. The model trusts your inputs. If you select "established revenue" without revenue you can document, the estimate is meaningless. Worse, a mismatch between your stated activity and your documentation is itself a rejection trigger.

This is why the output is framed as a likelihood, never a guarantee. Anyone selling you a guaranteed Mercury approval is selling you something that does not exist.

A non-resident's checklist before you apply

The calculator tells you where you stand. This checklist tells you how to improve the inputs you control before you submit:

1. Form the LLC and get the EIN first. Mercury wants a real US entity. You will need your Wyoming Articles of Organization and your IRS EIN letter (CP575). We file the LLC within 24 hours and obtain the EIN; no SSN or ITIN is required to bank - your passport is your identity document.
2. Build a one-page website. Even a simple landing page with what you do, who you serve, and how to contact you moves the documentation input meaningfully. "No web presence" costs you 10–15 points.
3. Write a specific business description. Name your product, your customers (ideally US-facing - Mercury's eligibility rules expect you to serve or plan to serve US customers), your fulfillment or delivery method, and a realistic monthly revenue estimate.
4. Gather proof of activity. A Stripe dashboard screenshot, signed customer contracts, or an app store listing all help. They turn "I claim to have revenue" into "here is the revenue."
5. Check your passport expiry. Use a passport valid at least 12 months out. Near-expiry passports trigger avoidable friction.
6. Confirm your country is not on Mercury's current prohibited list before you spend time applying.
7. Have a fallback ready. If your calculator estimate is low, line up Relay or Wise Business in advance (see below) so a Mercury rejection does not leave you bankless.

If your estimate is low: the banking fallback ladder

A non-resident founder is never reliant on a single bank. If the calculator puts you in a lower band, plan the ladder:

• Relay - similar requirements to Mercury but a different reviewer pool. Many founders rejected by Mercury are approved by Relay. Worth trying second.
• Wise Business - the safest fallback for non-residents, with very high acceptance. It gives you USD account details and strong multi-currency rails, though it is a fintech, not a chartered US bank.
• Brex - realistic only for funded or higher-revenue startups.
• Payoneer / Airwallex - high acceptance and good for receiving marketplace and multi-currency payments, with fewer full-banking features.

The point of estimating before you apply is that you can sequence these intelligently instead of scrambling after a rejection.

The tax angle non-residents must not skip

Banking and tax are separate problems, and getting approved by Mercury does not settle your IRS obligations. A foreign-owned single-member US LLC is, by default, a disregarded entity that must file Form 5472 (attached to a pro forma Form 1120) for any year it has reportable transactions with a related party - and capitalizing the LLC or paying yourself counts. The IRS penalty for failing to file is $25,000, and it escalates by a further $25,000 for each 30-day period the failure continues after IRS notice (IRS Instructions for Form 5472).

For the 2025 tax year, the filing is due April 15, 2026, extendable to October 15, 2026 with Form 7004 (About Form 5472, IRS). We mention this in the calculator's results page on purpose: a founder who is thrilled about a Mercury approval and then forgets Form 5472 can face a penalty larger than several years of LLC running costs combined. The bank account and the federal filing are two doors you walk through, not one.

How we keep the model current

We revisit the estimates periodically. Country profiles genuinely shift - a market that was relatively easy can tighten after a policy change, or loosen the other way - and an estimate that does not move with reality slowly becomes wrong. We also re-check Mercury's published eligibility and prohibited-country pages, because policy text sometimes moves before anything else.

We are upfront that these are estimates of how a reasonably well-prepared applicant tends to fare, not a promise of any individual outcome. Approval is the provider's decision and is never guaranteed: treat the framing as "you can improve the inputs you control," not "a given country-and-segment combination is guaranteed a particular result."

The bottom line

The Mercury Approval Calculator is a transparency tool, not a crystal ball. It turns our practical experience into a directional estimate you can use to prepare before you spend time and money applying. Country sets your baseline; your business model, revenue, and - most importantly - your documentation move it. The estimate is honest about its limits: tighter where our sense is firmer, a global average where it is thin, and never a guarantee.

If you are ready to start, we form your Wyoming LLC, file for your EIN, and make the Mercury introduction for $397, with the Wyoming state filing fee included. Run your numbers in the calculator first, fix what you can, then apply with the strongest profile you can build.

Start your Wyoming LLC + EIN + Mercury intro - $397, state fee included

Sources: IRS Instructions for Form 5472; IRS - About Form 5472; Mercury - Eligibility; Mercury - Prohibited countries; Mercury - Gathering your documents; U.S. Treasury Office of Foreign Assets Control (OFAC) sanctions program; scikit-learn - Probability calibration.