Nash Routes Each Delivery to the Cheapest Available Courier

01 /The Concept

Nash surfaced without warning, announcing a $20M current round alongside a previously undisclosed $7.8M round led by a16z, plus the standard small check from Y Combinator – which backed it the prior year.

The platform solves a specific friction in local commerce: deciding which delivery service to use for each order, at each moment, is a manual decision most businesses make poorly. Nash connects to delivery services – from large national networks like Uber, Lyft, and DoorDash to smaller local couriers – through their APIs, and aggregates availability and pricing data in real time. A business can see, at any given moment, which services have drivers available in their area, what they're charging, and what their recent performance history looks like.

Clients can evaluate and choose manually, or define selection criteria and delegate the decision entirely to Nash's algorithm, which picks the optimal service automatically for each order. A unified dashboard shows the status of all in-flight deliveries regardless of which carrier is handling them. The interface can also be surfaced to end customers, letting buyers select their preferred delivery option at checkout based on current availability and cost.

Nash claims that clients using the platform have seen delivery volumes increase by 2.5x. Two mechanisms plausibly explain this: first, customers find a suitable delivery option from that specific store because more options exist, and complete the purchase rather than looking elsewhere; second, even at unexpected peak times, some carrier in the network has capacity available.

02 /Why It Matters

An earlier review [covered here](/review/365-tysjach-kurerskih-sluzhb-uh-ty) looked at SafeDigit, which offers delivery companies integration modules to automatically receive orders from e-commerce stores. That creates lock-in for the delivery service but not for the merchant: a single integrated carrier leaves the merchant exposed when that carrier's drivers are all busy, or when a different carrier is offering better rates.

Nash flips the dynamic in favor of the merchant by making carrier choice dynamic and continuous. But the more interesting conceptual shift is what this represents for marketplace architecture.

The traditional mental model of a marketplace is a storefront: a directory of suppliers that buyers browse, compare, and select from. Nash represents a different model – the marketplace as a black box. The mechanics are:

- aggregate APIs from a large set of commodity suppliers, - unify them into a single interface or API, - embed that interface invisibly into the client's own workflows and systems.

The client's system calls the Nash API; Nash handles the real-time selection; the client's team never needs to think about which carrier is running any given delivery. The marketplace machinery is invisible.

This architecture already underlies some of the largest fintech infrastructure. Stripe is, in essence, a marketplace of payment and financial instruments from dozens of partner providers, exposed through a single API. Customers of a Stripe-integrated store either choose a payment method explicitly or have one selected automatically by rules running underneath the surface.

Smaller implementations of the same idea exist in e-commerce: Stockly (a virtual cross-merchant warehouse that auto-fulfills out-of-stock items from partner stores), 35up (cross-merchant upsell recommendations at checkout), and Carro (storefronts that show partner-supplied inventory). Each of these is a distributed marketplace where supplier selection happens invisibly inside another business's processes.

03 /Opportunities

The black-box marketplace model is already working at scale – it just isn't recognized as a distinct category, which means it isn't being deliberately used as a template for new businesses.

Nash itself is a strong model to copy into adjacent markets. Delivery is compelling: industry data puts the number of local delivery services in the US at 365,000 and growing at roughly 8% annually, which means the supply fragmentation that makes Nash useful will only increase over time.

The more general question is where else the pattern applies: large fragmented supplier markets offering functionally similar services, where real-time selection among those suppliers would produce better outcomes than manual choice or long-term lock-in. Insurance placement, freight brokerage, cloud compute resource allocation, and subcontractor networks in construction all share these characteristics to varying degrees.

The strategic entry angle is identifying a vertical where a merchant or operator currently manages supplier relationships through a combination of personal relationships, spreadsheets, and inertia – and replacing that with a live selection API. The market already exists; the infrastructure connecting it intelligently usually doesn't.