Mercury
Each of these is a real decision Mercury is working through, written so you can make the call yourself — incomplete information, someone who pushes back, no answer key.
Mercury is a US-based fintech platform delivering digital-first business banking, treasury, credit, and financial operations software for startups, tech-focused SMBs, founders, and VC-backed companies, operating on top of partner banks, not as a chartered bank itself. It offers checking and savings accounts, debit cards, payment services, cash management tools, and analytics. It is growing into new segments including eCommerce and New Verticals such as Marketing Agencies, Consulting, and Staffing. Mercury is building Mercury Bank, N.A., a proposed nationally chartered bank focused on serving ambitious businesses, and has obtained a provisional national bank charter while taking proactive steps toward final approval. Mercury operates an SEC-registered investment advisor and a FINRA-member broker-dealer, and its Investments team builds products that oversee billions in customer assets. It builds a personal banking product covering personal money management, banking, savings, investments, money movement, and, in the future, lending, and is building its first consumer credit card. Its infrastructure is built primarily on AWS, including EC2, RDS, OpenSearch, ECS, and EKS.
Your first one is free. No card.
28 live right now.
The rest
Clarify payment choices before founders rebuild the comparison
Founders are leaving the payment flow to compare timing and cost in spreadsheets. You have to weigh a simpler decision against honest limits on what each payment route can promise.
Choose the next campaign test before weak joins waste spend
Strong signup campaigns are producing uneven funded and payment-active accounts. You must weigh faster testing against whether the joined evidence is strong enough to guide message and spend.
Win lean founders while software costs keep climbing
Lean startup founders are leaving before they reach spending controls, even as software costs rise. You have to weigh a focused activation loop against claims and audiences that can withstand review.
Shape a cash routine while founders watch every dollar
Newly funded founders are asking how to manage cash and employee spend as costs rise. You have to weigh repeatable adoption against the boundary between practical guidance and financial advice.
Forecast cash pressure before payroll turns into a spreadsheet scramble
Founders cannot tell whether upcoming payroll and pending payments will outpace available cash until they rebuild the picture themselves. You must weigh an early, useful forecast against uncertainty that could make a confident answer misleading.
Decide who joins a pilot for recruiter cards without loosening controls
Staffing firms cannot consistently tell whether recruiter card spend is operating expense or client-funded activity. You must weigh pilot learning against controls that remain usable for distributed recruiters.
Choose one integration for agency cards before the first cohort starts
Agency operators are mixing client campaign charges across employee cards and separate accounts. You must weigh a low-friction platform experience against controlled card use and sustainable partner economics.
Set cash rules for Command without surrendering founder control
Founders want Command to propose movements of idle operating cash, but the approval and evidence boundary is not yet defined. You have to weigh faster cash management against consent, auditability, and partner-bank confidence.
Set limits for agentic payments while approval records stay incomplete
Some agentic payments cannot be tied cleanly to a customer instruction and approval. You must weigh continuity for active businesses against a record gap that partner banks cannot accept.
Recheck Command access when staged payments reach approval
A staged payment can remain available after a finance admin removes the requester’s access. You must weigh delegated payment speed against proof that authority is current at approval.
Surface AI spend without mislabeling ordinary software costs
Founders are discovering AI and token charges only after monthly card review, when controls arrive too late. You must weigh useful category visibility against the risk of putting a misleading control in front of an account owner.
Prepare risk cases before founders repeat their story
Specialists reconstructing account history delay clear responses when founders need payment access explained. You must weigh faster preparation against the evidence and judgment that must remain human.
Cut EDD prep from 90 to 45 minutes while investigators decide
EDD preparation takes about 90 minutes per matter because prior evidence is scattered across review records. You have to weigh faster investigator context against citations and judgments that must stand up to scrutiny.
Assemble closure evidence without letting AI close the issue
High-risk issue packages are being reopened because the proof and procedure do not line up. You must weigh faster manager review against preserving independent human validation.
Turn platform discovery into next steps before rivals shape the deal
Only 34% of qualified platform opportunities get a reviewed next step within three business days. You must weigh faster follow-up against the seller’s need to verify every integration claim and commercial promise.
Sort invitation feedback without letting AI set the roadmap
Founders keep raising different concerns about inviting finance teammates, and PMs cannot quickly separate a recurring permissions problem from channel noise. You have to weigh faster evidence synthesis against keeping every product judgment with the PM.
Draft cited context for fraud alerts while engineers own diagnosis
Engineers spend a median 42 minutes assembling context for each suspicious card alert. You must weigh faster investigation setup against evidence that engineers can independently verify.
Confirm the payee without stalling a wire
The partner bank requires payee confirmation on large domestic wires within 45 days, and replays show 0.4% with mismatched details, mostly urgent payments. You have to design the check so payroll still goes out on time.
Settle how unclaimed balances are handled
A partner-bank review found 37 balances from closed accounts handled under inconsistent dormancy rules, and set a 75-day deadline. You weigh a migration you cannot reverse against the cost of classifying a balance wrong.
Move 180 customers before the wire cutoff
A partner bank retires the old wire workflow in 60 days, and 180 customers have saved beneficiary details that won't carry over. You decide who gets validated first, without stalling time-sensitive supplier payments.
Decide if the model is ready for intake
A partner bank set a fixed date to retire manual review, and replay shows the errors cluster in applications with several owners. You have to weigh unsafe approvals against delays for real founders.
Switch over wire screening before the deadline
The partner bank retires the current wire screening in 75 days, and 0.6% of replayed wires still disagree on name matches and payee data. You weigh keeping payments moving against being sure the controls hold.
Decide what to move onto the new authorizer
Routing has to move off the old provider within 90 days, and the new one approves more but disagrees on some merchant and limit rules. You have to pick a scope you cannot cheaply undo.
Fix a channel that approves but rarely funds
One program brings 31% of approvals but only 18% of accounts that fund and pay within 60 days. You have three weeks to reset targeting and qualification without cutting off the volume growth needs.
Decide which digital-asset accounts can stay
The partner bank set a deadline to replace informal case guidance with a written eligibility and monitoring standard. You have to say which activities remain eligible, knowing a restricted label is costly to undo.
Settle the referral terms before the conference
The partner wants a co-branded referral program announced at its annual conference, and the draft revenue-share and data terms would be hard to unwind. You have to weigh that launch date against economics you can live with.
Move the ledger without breaking balances
The legacy store retires this quarter, and shadow writes show a small mismatch on reversals and on late partner events. Once you start the cutover there is no clean rollback, so you have to decide how much proof is enough.