For a business building software in late 2024, crypto’s most interesting questions may be fairly ordinary ones. Can a payment move with less friction? Can different systems recognize the same digital item? Can a customer use the result without first becoming the family blockchain correspondent?
The October 2024 State of Crypto report from a16z crypto puts stablecoins and infrastructure improvements prominently in the conversation.1 Its perspective deserves context: a16z is an investor in the sector and has a commercial interest in its development.
Still, infrastructure is a more useful starting point than assuming every business needs a token. A system should earn its complexity through the job it performs. The pitch deck is allowed to be excited; the order ledger needs to remain composed.
Payments: compare the complete journey
A payment rail moves value between parties. For a merchant, the actual task begins before that transfer and ends after someone has reconciled it with an order.
Stablecoins aim to maintain a reference value, often against a currency. That intended stability does not remove the need to evaluate the particular asset, provider and network. Nor does it make a payment workflow automatically familiar to the customer.
A low network fee is only one cost
Suppose a fictional supplier and buyer are evaluating a digital payment route. Both need a way to obtain, hold, send and ultimately use or convert the asset. There may be provider fees, conversion costs and operational work around those steps.
Compare the full process with the existing alternative. Include handling exceptions and returning money, not just the successful transfer. If the customer needs outside help to complete a routine payment, that support belongs in the comparison too.
The a16z report itself notes that some quoted onchain costs may exclude entry and exit costs. That is a useful reminder to follow the money all the way through rather than stopping the stopwatch at the most flattering point.
When off-the-shelf stops fitting, see how we approach custom development.
Settlement needs a business definition
An order system needs explicit rules about when it considers a payment received, which network and asset it accepts, and how it handles underpayments, duplicates and delayed confirmation.
A successful transfer record also needs matching to the right customer and order. Without that connection, the business has money somewhere and a spreadsheet developing a personality.
None of this is unique to crypto. The unfamiliar rail simply makes the surrounding accounting and exception handling harder to take for granted.
Digital ownership: identify who needs the shared record
A token can represent a digital item or entitlement, but the business still has to define what it means and which systems will honor it. A record alone cannot make an external service deliver the promised benefit.
For a fictional membership program, ask whether members need to carry an entitlement between independent organizations. If the same company controls every benefit and every participating system, a conventional account record may be simpler to operate and recover.
If this was useful, try Stablecoin subscriptions have arrived, with some important small print.
Lower infrastructure costs change some possibilities
Ethereum’s Dencun upgrade in March 2024 introduced a different way for rollups to publish certain data, supporting lower data costs for those scaling systems. That creates room to explore applications where expensive transactions previously made little sense.
It does not mean every transaction on every network becomes cheap, or that the customer experience looks after itself. Wallet access, recovery, permissions and the link between a token and a real benefit still need a design.
I’m interested when a shared record solves a coordination problem that otherwise requires awkward trust between separate parties. I’m less persuaded when blockchain is added to a loyalty program that one business could already run quite comfortably.
The decision can stay modest. Choose one operation, compare a complete conventional solution with a complete crypto-based one, and include what happens on a bad day.
If the newer infrastructure makes that operation clearer or more practical, there may be something worth building. If its strongest benefit is giving the About page another futuristic noun, the conventional database will probably cope.