Of the six crypto products we've shipped, three have tokens and three don't. The pattern of which ones got tokens is not what we expected going in.
When you actually need a token
- The token is the product — a stablecoin, a memecoin, a governance asset where the asset itself is what users are transacting in.
- The protocol needs incentive alignment between independent parties — liquidity providers, validators, oracles. The token is the alignment mechanism.
- The product requires non-custodial value transfer between users on-chain, and no existing asset fits the unit of account you need.
When you don't (and shouldn't)
You think you need one because investors keep asking. You think you need one because "tokenomics" sounds more like a real business. You think you need one because your competitor has one.
None of these are reasons. They're vibes.
A token adds: securities risk, ongoing custody concerns, exchange-listing politics, treasury management, and a permanent obligation to your holders. Make sure you actually need all five.
The decision matrix we use
For each new crypto engagement, we ask four questions before we'll commit to a token design:
- Is the token the unit of value being transacted, or is it a side effect?
- Could USDC plus a database column do exactly this job?
- Who holds the token after launch — users, the protocol, or speculators?
- What's our compliance posture if the token trades on a secondary market we don't control?
If three of those four answers are "speculators" or "unclear," we recommend killing the token and shipping the product anyway. Half the time, the client agrees. The other half, they go elsewhere.
The tokenless products
All three are non-custodial wallets with smart-contract-based access controls. Users hold their own keys, the protocol takes a small fee on transactions, no token required. Revenue is real, the business is fundable, and the legal posture is dramatically simpler.

Founder. Sets the engineering bar. Runs every audit and signs off on every migration plan.