For US agencies, 4 structural picks.

Digital agencies, consultancies, and professional-services firms run materially different cash flows from SaaS or e-commerce. The structural differentiators: multi-client retainer cash flow (3-30+ active clients each contributing monthly retainers), 1099-contractor-heavy operating shape (10-100+ recurring contractors versus full-time employees), per-client P&L tracking (gross margin per client matters operationally), and retainer-receivable timing (project / monthly retainer revenue rather than transaction-volume revenue like e-commerce or subscription revenue like SaaS).

The structural decision pivots on which of three patterns matches your agency's operating shape. Per-client cash visibility (Relay's sub-accounts) is the deciding factor for agencies running formal per-client P&L tracking — each client gets a dedicated sub-account and the close-the-books friction at month-end compresses materially. Retainer cash + treasury yield (Mercury) is the deciding factor for agencies above $1M revenue with material retainer-receivable parked cash — the IntraFi sweep covers FDIC pass-through to $5M and Mercury Treasury yields on the parked-cash leg. Corporate-card spend depth (Brex paired with Brex Cash, or Mercury paired with Ramp) is the deciding factor for agencies with material vendor spend (SaaS tools, freelance contractors via card, content / marketing spend).

Mercury ranks first for agencies above $1M revenue where retainer cash management plus treasury yield is the structural problem — $5M IntraFi sweep, Treasury yield on parked retainer cash, paired with Ramp for the corporate-card layer. Relay ranks second for agencies running formal per-client P&L tracking — up to 20 sub-accounts under one login is structurally unique. Brex ranks third for agencies wanting the consolidated card + cash + spend-management stack in one platform with the YC-cohort-style onboarding paved. Ramp ranks fourth as the paired spend-management + Bill Pay + 1099-contractor-payment layer stacked alongside whichever operating-account choice fits.

What is the structural fit for a digital agency with 5-50 employees?

Mercury for the operating account (FDIC pass-through to $5M via IntraFi, Treasury yield on retainer-receivable parked cash) plus Ramp for the corporate-card and 1099-contractor Bill Pay surface is the most-common pattern. For agencies running per-client P&L tracking, Relay's sub-account architecture (up to 20 sub-accounts under one login) is the structural alternative — each client gets a dedicated sub-account for inflows / outflows / margin tracking. The decision between Mercury and Relay typically pivots on whether per-client cash visibility is the structural problem (Relay wins) or whether retainer cash management plus treasury yield is the structural problem (Mercury wins).

How do these platforms handle 1099 contractor payments?

All four platforms ship 1099-friendly contractor-payment workflows. Mercury, Brex, and Rho route via standard ACH with W-9 collection at the contractor onboarding step; Ramp Bill Pay is the most-automated surface (OCR vendor onboarding, automatic 1099-K / 1099-NEC tracking, year-end 1099 filing integration). For agencies running 20+ recurring 1099 contractors, Ramp Bill Pay typically reduces close-the-books friction materially. The structural caveat: cross-border contractor payments (non-US W-8 contractors) settle via SWIFT correspondents with intermediary-bank fees — none of the four platforms offer Wise-level FX on cross-border contractor payouts. For agencies with material international contractor flow, stack Wise Business or a corresponding-bank relationship alongside.

Why does sub-account architecture matter for agencies?

Agencies with multi-client retainer cash flow benefit from per-client cash visibility: separate buckets for inflows (retainer receipts), outflows (contractor + vendor payments), and margin reserves per client. Relay ships up to 20 sub-accounts under one login, which is structurally distinct from Mercury (1 main + 1 Vault) or Brex (1 main + Brex Cash). For agencies running 10+ active retainer clients, Relay's architecture compresses the close-the-books reconciliation friction — each client's cash position is visible at the sub-account level rather than reconstructed from QuickBooks class-coding at month-end. The trade-off: Relay's sweep ceiling ($3M via Thread Bank) is lower than Mercury's ($5M) and Brex's ($6M), so agencies above $3M idle cash should either split balances across multiple sweep networks or pair Relay with Mercury Treasury.

Do these platforms integrate with QuickBooks Online, Xero, and Sage?

All four ship QuickBooks Online and Xero integrations with bidirectional transaction sync, class / department / location coding (relevant for per-client P&L in QBO Plus or Advanced), and vendor-record mapping. Sage Intacct integration is available on Mercury IO, Ramp Plus, Brex Premium, and Rho. For agencies running QuickBooks Online Plus with class-coded per-client P&L, the integration depth is roughly equivalent across all four; the differentiator is more about which platform's native UX matches the agency's operating shape (Relay's sub-accounts vs Mercury's sweep + Treasury vs Brex's consolidated stack vs Ramp's spend-management depth).

For agencies running formal per-client P&L tracking, Relay with its 20-sub-account architecture is the structural fit — each client gets a dedicated sub-account and close-the-books friction compresses materially. For agencies above $1M revenue where retainer cash management + treasury yield is the structural problem, Mercury with its $5M IntraFi sweep + Treasury yield — typically paired with Ramp for the corporate-card and 1099-contractor Bill Pay layer. For agencies wanting the consolidated card + cash + spend-management stack in one platform, Brex. The two-platform pattern (Mercury + Ramp, or Brex + Brex Empower) is increasingly common at growth-stage agencies above 25 employees; the single-platform pattern (Relay alone, or Mercury alone) is the structural fit for smaller agencies under $1M revenue.