Require three numbers from any provider of crypto market making services, each measured on your own token and venues: depth within a tight band of the mid-price, a maximum bid-ask spread, and uptime measured in short windows. Tokenomist’s example of an enforceable clause is a 50 basis point maximum spread, at least $25,000 on each side and quotes live 95% of each month. Use that as the shape of the commitment, then set the numbers for your market.
Who decides this
The head of markets owns these numbers. They write the quoting schedule, agree it with the CFO, who pays for it, and monitor it after listing. They are accountable for one thing the board will ask on listing day: can a holder sell a meaningful position without moving the price far?
Which order book depth measure should the contract use?
Specify depth as the value of resting orders within 1% of the mid-price on each side, not the 2% figure that data aggregators show. Kaiko, a market data provider, defines depth as the sum of limit orders within a set percentage of the mid-price. It reports that 0.1% and 1% are its own most-used bands.
Kaiko gives a reason to avoid 2%. It stopped using 2% depth because that is the level commonly reported on CoinGecko and CoinMarketCap, and it had observed that level is more often manipulated. A commitment written at 2% invites a provider to stack orders where the screenshots look good and the trades never happen.
Write depth per side and per venue. A total across both sides can hide an empty bid. A total across venues can hide a dead pair on the exchange where most holders trade. We would also add a smaller top-of-book size, so the best bid and ask carry real orders, not a token amount.
Size the number from the trades you expect, not from a peer’s contract. Kaiko expresses liquidity as slippage on a hypothetical order, such as a $100,000 sale. Pick the sale a large holder or the treasury might need to make, and require enough depth inside the band to absorb it.
How tight should the bid-ask spread commitment be?
Set the maximum spread together with the depth figure, because a tight spread on a thin book is worth very little. Flowdesk, a market maker, makes the point with an example: a 5 basis point spread on an altcoin may carry only a few hundred dollars of depth. It calls a 30 basis point spread with $100,000 of depth a healthier signal for traders.
There is no market standard for the right spread. Arrakis notes that KPIs are not standardized and terms vary widely between firms. Exchanges’ own programmes give a reference point. BitMEX’s designated market maker programme asks quoters on its Group C contracts, which include several altcoins, to stay within 0.50% of the mid-price at a set size.
That reference applies to derivatives on a large venue, not to a newly listed token. Its value is the structure: spread measured from the mid, paired with a minimum size, on a named instrument. Copy the structure and set the numbers for your own pairs.
What uptime is enough, and how should it be measured?
Require at least 95% uptime, measured in daily or shorter windows, with a cap on the longest single outage. Flowdesk says market makers should generally achieve more than 95% uptime. Tokenomist’s fee comparison uses 98% across five venues in its stronger quote.
The measurement window matters more than the percentage. Tokenomist warns that a 95% monthly target can be met with the book empty for 36 hours in a row. It recommends shorter windows and a cap on the longest outage. Those 36 hours could be the first day and a half after listing.
Define “up” precisely. A quote counts only when both sides are inside the maximum spread and at or above the minimum size. BitMEX tracks its own market makers’ presence tick by tick and asks for quotes at least 90% of market hours. An uptime SLA that counts any order at any price is not an SLA.
Ask about the infrastructure behind the number. Flowdesk points to responsive infrastructure across exchange integrations and teams spread worldwide, since crypto trades around the clock. Neither belongs in the contract as a figure, but both explain why a provider misses uptime on the most volatile days, which are the days that matter.
| Commitment | Weak version | What to require | How to measure |
|---|---|---|---|
| Order book depth | “Deep liquidity” or 2% depth in total | Value within 1% of mid, per side and per venue, plus a top-of-book minimum | Sampled order-book snapshots from a third-party source |
| Bid-ask spread | A tight spread with no size attached | Maximum spread from mid, valid only at the minimum size | Same snapshots, time-weighted |
| Uptime SLA | 95% averaged over a month | 95% or more in daily windows, with a cap on the longest outage | Share of snapshots meeting both spread and size |
| Venue coverage | “Major exchanges” | Named venues, plus a rule for adding new ones | Venue list in the schedule |
| Capital deployed | Not stated | Depth posted relative to tokens and capital lent | Depth snapshots against the loan schedule |
| Reporting | Firm’s own monthly PDF | Live dashboard plus a fixed-format report | Reconciled to the third-party data |
Who measures, and what happens when targets are missed?
The project should measure performance against third-party order-book data, and every missed target should trigger a pre-agreed remedy. Tokenomist recommends that the agreement name who measures, the sampling frequency, the data source and the dispute process. Without an independent verification right, the firm’s own report is the only record.
Independent data is practical. Kaiko reports collecting two order-book snapshots per minute for each instrument it covers, so a provider’s claims can be tested against an outside record. Agree the source and the sampling rate in the contract; do not argue about them after a bad week.
Fix the reporting format at the same time. Flowdesk suggests asking for live dashboards or regular reports on trading activity, depth, spreads, market share and PnL, and asking how often they arrive. We would require the provider’s report to use the same data source and sampling rate as the contract, so the two records never disagree.
Add a capital check. Tokenomist suggests comparing the depth posted with the capital lent. A ratio near 100% means the loan is working; a low ratio means tokens are sitting idle while the firm holds the options. Above 100% means the firm is committing its own balance sheet.
Remedies give the numbers force. Tokenomist lists a fee reduction, a shorter notice period and a right to terminate without penalty. Under a retainer, the fee reduction is the natural lever. Under a loan, the termination right and the return schedule are what count, and those clauses are covered in our guide to market maker contracts and the clauses that matter.
Watch for venue drift. Tokenomist notes that a token listed on a new exchange after signing is not automatically covered. Either the schedule adds venues by rule, for example any venue above a set share of the token’s volume, or the new venue goes unquoted.
What goes wrong: a launch plan that assumed liquidity
The launch plan has a market maker in it but no numbers. The provider’s proposal promised “tight spreads and deep liquidity on major venues,” and the agreement repeated the phrase. The head of markets assumed that phrase meant a book a holder could sell into.
Listing day shows the gap. The spread looks tight on the aggregator, but the bid is a few hundred dollars deep. Depth at 2% looks healthy; inside 1% there is almost nothing. Uptime averages fine over the month, which includes the hours on day one when there was no bid at all.
It can get worse than a thin book. In March 2025, Binance said it had offboarded a market maker that placed sell orders for 66 million tokens a day after a listing while posting only small buy orders. Numeric two-sided obligations, watched from outside the firm, are how a project sees that pattern in hours rather than from an exchange’s announcement.
How Alcyone approaches it
Alcyone treats the quoting schedule as part of the launch plan, where every part has a named owner. Through Alcyone’s business development and partnerships advisory, its partners introduce issuers to exchanges and market makers and review market-making contracts. Partners stay on the line through the launch window, and a post-launch report follows within the week.