McKinsey · Tech Consulting · Hard · 35-45 min

US Regional Bank — Join FedNow or Stay on ACH

US Regional Bank — Join FedNow or Stay on ACH is a hard McKinsey tech consulting case interview that runs 35-45 min. A US regional bank — $22B assets, 1.4M retail and small-business customers, core in the Midwest and Southeast — still moves money on ACH and wires. A strong answer works through 5 phases: Separate the customer jobs; Compare FedNow, RTP, and 'do nothing on ACH' on reach, cost, irrevocability, and time to first live payment; Size the fraud and ops cost of irrevocable payments; Decide build-on-processor vs deeper core integration, and what the core vendor will and will not support; Recommend a lead rail, a 18-month sequence, and the kill criteria if adoption or fraud misses the plan.

Last updated 2026-09-05

The brief

A US regional bank — $22B assets, 1.4M retail and small-business customers, core in the Midwest and Southeast — still moves money on ACH and wires. Instant payments are live: the Federal Reserve's FedNow and The Clearing House RTP. A neobank just launched instant payouts to contractors in the bank's footprint, and treasury clients are asking why payroll still settles next day. Joining both rails costs about $9M over two years (processor, fraud stack, core integration) plus a running fraud-loss budget the CRO has not signed. The CEO wants a go / no-go and which rail to lead with. You advise the CTO.

How to approach it

  1. Separate the customer jobs — consumer P2P, small-business payouts, treasury payroll — and say which one actually moves deposits or fee income
  2. Compare FedNow, RTP, and 'do nothing on ACH' on reach, cost, irrevocability, and time to first live payment
  3. Size the fraud and ops cost of irrevocable payments — authorized push payment, account takeover — before treating speed as free
  4. Decide build-on-processor vs deeper core integration, and what the core vendor will and will not support
  5. Recommend a lead rail, a 18-month sequence, and the kill criteria if adoption or fraud misses the plan

What a strong answer does

  • Picks the use case by deposit and fee economics, not by 'instant is the future'
  • Treats FedNow and RTP as overlapping but not identical — receive vs send, bank reach, and who already sits on the processor
  • Prices fraud loss and the dispute model as a P&L line, not a risk appendix
  • Sequences receive-first or a narrow send use case rather than turning on every payment type on day one
  • Names what the CEO tells treasury clients on Monday if the answer is 'not yet'

Red flags interviewers score down

  • Recommends both rails immediately because competitors have them
  • Ignores irrevocability and treats instant payments like a faster ACH
  • Skips the core-vendor constraint and assumes a greenfield integration
  • No number on fraud loss, adoption, or payback

Cases are written in each firm's style, written and reviewed by working consultants; they are not the firms' own published cases.

More McKinsey cases