What do crypto market making services actually do?

Then the listing manager sends back a short message with three conditions before going live: minimum daily volume, a tight bid-ask spread, and a signed agreement with a market maker. You open the proposals you collected during your fundraise — and the language in them is dense. Two-sided quoting. Depth SLA. Inventory risk. Cross-exchange hedging. Loan structure. Retainer.

This is where most token projects stall. Not on the marketing, not on the exchange relationship, but on the operational decision of how to keep the order book alive once the listing is public. The market making conversation is technical on the surface, but underneath it is a question about how much friction your project is willing to absorb — and who is going to absorb it.

In this piece we want to walk you through what crypto market making services actually do, how the two dominant pricing structures differ in practice, what risk management looks like when a token is moving, what success looks like when measured honestly, and where the line sits between legitimate liquidity provision and the kind of activity that quietly erodes trust.

Market making is operational plumbing — the work of staying quoted through quiet hours, volatility, and exchange downtime so that your token behaves like a tradable asset rather than a one-off event.

How automated liquidity provisioning actually works on the order book

The mechanical core of crypto market making services is simpler than the marketing decks suggest. A market maker is a participant — almost always an algorithm, often running across many exchanges simultaneously — that continuously places both a buy order and a sell order on the order book at prices around the current mid-price. The buy side is called the bid; the sell side is called the ask. The gap between them is the bid-ask spread.

When a market maker is active, the spread narrows. On a healthy pair you might see the bid-ask tighten from a starting range of several percent — six percent is a common starting point for a thin new listing — down toward a fraction of a percent, often below half a percent on active pairs. The orders sitting on the book, waiting to be filled, are typically clustered within roughly 0.5% to 2.5% of the mid-price. That cluster is what people in the industry call depth, and it is what your community actually feels when they try to transact.

This is what the work actually is: continuously quoting two-sided markets, in real time, across changing conditions. Algorithms adjust order sizes, reposition quotes, widen spreads during volatility, and tighten them when markets calm — all while watching their own inventory, which is the running position of tokens and quote currency the market maker has accumulated from fills. If the algorithm buys more than it sells, it is long inventory and exposed to downside; if it sells more than it buys, it is short. Both states are managed actively, not passively.

The thing worth holding onto here is that the market maker is not generating the price — the price is set by the broader market, and by the trades of real participants. What the market maker does is keep the order book tradable at any moment, so that when a genuine buyer or seller shows up, they can transact immediately at a price close to the last traded value. That distinction matters when you are trying to read a proposal and separate the parts that describe real work from the parts that describe volume theater.

Retainer versus loan structures — and why the choice is about alignment

Most market making engagements fall into one of two structures, and the choice between them is more important than it looks. Both can produce a working order book, but they put your project in very different positions, and the alignment of incentives shifts in ways that show up months later.

Under the Retainer Model, sometimes marketed as Market Making as a Service (MMaaS), the token issuer pays a fixed monthly fee and provides both sides of the liquidity — typically the project token and a stablecoin like USDC. The market maker supplies the algorithm, the exchange connectivity, the risk management, and the operational uptime, but does not put its own capital on the line. The project retains full control of the inventory sitting on the book. Because there are no option mechanics involved, there is no dilution pressure on the token's float from the market making relationship itself.

Under the Loan/Option Model, the market maker funds the quote side of the book itself — usually USDC or USDT — and the project loans the market maker a quantity of its tokens to be sold into the market. Compensation does not come from a monthly invoice; it comes from call options written on the token, or from spreads captured on the activity, or from a combination of both. The project does not pay an upfront retainer, but it does give up inventory and accepts dilution through the option component, which can be material depending on the strike and the size of the option grant.

Here is a compact comparison of how the two structures differ in practice:

The alignment question is real. In a retainer, the market maker is paid to keep the book alive regardless of price action — its incentive is uptime and SLA compliance, which is what you want when your priority is operational continuity. In a loan structure, the market maker is compensated more when the token performs — and that compensation shape can be a feature or a risk depending on how it is calibrated. Neither structure is intrinsically better; the right answer depends on your token's float, your treasury policy, and how much dilution you are willing to absorb in exchange for not paying a fixed retainer.

What risk management actually looks like when markets move

The phrase "risk management" is everywhere in market making proposals, and it is worth unpacking because it is where most of the operational quality lives — and where most projects underestimate what is required to keep an engagement healthy through a full market cycle.

The core risk a market maker manages is inventory risk. If the algorithm has accumulated a long position in your token because buyers kept appearing and sellers did not, the market maker is now exposed to a drop in price. If it has accumulated a short position because sellers dominated, it is exposed to a rally. In either case the algorithm has to neutralize that exposure without breaking the order book you have hired it to maintain. Doing nothing is not an option, because the inventory bleeds quietly into the P&L of the engagement and shows up one day as an uncomfortable conversation.

There are three tools that come up most often, and you should expect any serious firm to be running all of them as part of their standard kit.

First, cross-exchange arbitrage. When your token trades on multiple venues, the algorithm watches the price differential between them in real time. If one venue is trading meaningfully above another, the market maker can buy on the cheaper venue and sell on the more expensive one, capturing the spread and flattening inventory at the same time. This is the most routine of the three tools and the least controversial, and it is also the one that quietly does the most work for a multi-listed token.

Second, perpetual futures hedging. When the market maker is long inventory and the market starts to fall, the algorithm can short an equivalent notional on the token's perpetual futures market. The short position gains as price falls, offsetting the loss on the inventory held for spot market making. This is a standard tool, but it requires that a healthy perp market exists for your token — which is not always the case for newer or smaller-cap projects. If your token does not have a liquid perp, ask the market maker what they do instead.

Third, dynamic spread widening. During sharp moves — a token unlock, a hack announcement, a coordinated sell — the algorithm will deliberately widen the bid-ask spread. This is not a failure mode; it is the correct response. A tighter spread during chaos means the market maker is absorbing toxic order flow at unfavorable prices and will not stay quoted for long. A wider spread during chaos means the market maker is still present, but at prices that reflect the new information. Communities sometimes read this as the market maker "disappearing," when in fact it is the market maker doing the work responsibly.

If a market maker promises you a tight spread through every kind of market event, they are either lying or they will not be your market maker for long.

The takeaway for founders is that risk management is not a back-office function you can ignore once the agreement is signed. It is the part of the engagement that determines whether your order book is there for you in week six and month six, or whether it quietly disappears the first time your token has a bad day. You do not need to understand the technical mechanics, but you should understand enough to ask intelligent questions about what your market maker is actually doing when the market is stressed.

Measuring success honestly — SLAs, depth, and spread

A market making engagement that cannot be measured is an engagement that cannot be held accountable. The firms that have been doing this work for years will agree to service-level commitments in writing. The exact numbers vary, but the categories are stable, and you should expect them to appear in any serious proposal you receive.

Uptime is the share of the trading day that the algorithm is actively quoting two-sided markets on your pair. Serious operations target the high nineties; anything in the eighties should be a conversation about why and a commitment to fix it.

Order book depth is the size of resting limit orders clustered around the mid-price. A common benchmark is depth within ±2% of mid — meaning, how much can be bought or sold without moving the price more than two percent. Your mileage will vary by market cap and listing tier, but a meaningful depth figure should be specific to your pair, not borrowed from a larger token's profile.

Maximum spread is the upper bound on the bid-ask spread the market maker commits to under normal conditions. During volatility events this can loosen temporarily; the SLA defines what "normal" means and what happens — communication, remediation, fee adjustments — when it does not.

We have seen projects skip this step because they trust the market maker's reputation, or because the listing window felt urgent, or because negotiating felt confrontational. None of those reasons survive a six-month review meeting when the order book looks thin and the proposal says only "we will provide liquidity." Specific, measurable, time-bound commitments are how you turn a service relationship into a working partnership — and how you protect your community from being the ones who discover the absence of depth during a volatile hour.

A useful exercise is to pull a chart of resting bids and asks across a 24-hour window every month and look at it with the same attention you would give a community sentiment report. You are looking for continuous depth at multiple price levels, not a single spike followed by emptiness. That single chart, more than any deck the market maker sends you, tells you what your engagement is actually producing.

Legitimate market making versus the activity that erodes trust

There is a version of this work that erodes trust — and it is worth naming it directly because the phrase "market making" gets attached to activities that look superficially similar but produce very different results on the order book.

Legitimate market making is visible on the order book. It produces tight spreads, meaningful depth at multiple price levels, uptime through quiet hours and volatile ones, and OTC execution for larger tickets. It does not need to advertise, because the order book is the advertisement. Its compensation structure — retainer or option-based — is disclosed in the agreement, and the inventory it holds is visible to anyone who bothers to look.

The other thing — sometimes called wash trading, sometimes called volume generation, sometimes wrapped in marketing language that obscures what is actually happening — produces trades that exist primarily to be counted. They show up as volume on aggregators. They do not show up as resting depth, because the orders are not designed to rest; they are designed to execute against each other or against a counterparty who knows the activity is coming. The spread during the activity is often wider than during quiet periods, which is the opposite of what real liquidity provision looks like. When the activity stops, the order book goes back to whatever it would have been without the engagement.

The signal that distinguishes the two is durable order book depth across normal trading hours, not volume figures quoted in a deck. If you pull a chart of resting bids and asks across a 24-hour window and the book is empty most of the time and briefly full, you are looking at the second kind. If the book has continuous depth at multiple price levels with a tight spread that occasionally widens under stress, you are looking at the first.

For projects that care about the long-term credibility of their token — and most do, even when short-term pressure makes the alternative tempting — this distinction is the difference between a market making relationship that compounds trust and one that quietly burns it. The trust deficit from a thin or manufactured order book is not something you recover with a follow-up announcement; it shows up in the next listing conversation, the next partnership pitch, and the next community thread where someone asks why the chart looks the way it does.

Closing thought

The question we keep coming back to is not "should we hire a market maker" — for most listed tokens, the answer is yes — but "what kind of alignment are we signing up for, and what does success look like twelve months from now."

A market making engagement is a long-running operational commitment dressed up as a service agreement. The retainer model puts the cost on your P&L and gives you control over inventory. The loan model puts the cost on your token float and gives the market maker a stake in performance. Either can be the right answer. Neither replaces the work of building genuine demand. And neither protects you from the trust deficit that comes when the order book turns out to be thinner than the proposal suggested.

So before you sign, ask the firms on your shortlist the same three questions: what is your uptime SLA, what is your depth commitment at ±2% of mid, and what is your maximum spread under normal conditions. If the answers come back in numbers, you are in the right conversation. If they come back in adjectives, you are not.

The order book is the most honest scoreboard your token has. Treat the engagement that builds it with the seriousness it deserves — and the sustainability of everything else you are building will tend to follow.