Tokenomics Launches: A Guide for Projects With a Real Product

There are two kinds of token launch, and almost every guide on the internet is written about the one you probably aren't doing.

A memecoin launch is a distribution event: the token is the product, the chart is the roadmap, and it is largely decided in the first few blocks — which is why that playbook is about bundles, snipers and curve mechanics. If that's your project, read that one instead.

A tokenomics launch is different in kind. You have a product, a protocol, or a business that already exists and already does something. The token isn't the thing — it's an instrument with specific rights attached, and which rights you attach is your decision to make: a share of protocol fees, access to something the protocol does, a vote over named parameters, or some combination. It is not equity, it does not convey ownership of your company, and writing about it as though it did is how founders talk themselves into obligations they never meant to take on. What you're issuing on launch day is not a chart, it's a capital structure: an allocation table, an unlock calendar, and a liquidity plan that you then have to live inside for the next three to four years. Get those three documents to agree with each other and you have a token that survives its own supply. Get them wrong and no amount of product progress will outrun the emissions.

The base rates are unkind. Arrakis Research looked at 125 launches and found 85% of 2025 launches ended the year negative and only 9.4% of tokens that declined in their first week ever recovered. Week one is not a marketing window; it's the week your supply structure meets a real order book, and the outcome is mostly decided before you deploy anything. This guide is about deciding it deliberately.

Step 0: Does your project need a token, and which shape?

The uncomfortable first question. A token is a permanent, public, liquid instrument that trades every fifteen seconds, carries whatever rights you gave it, and brings a constituency with opinions. It should do a job that nothing cheaper does.

There are three jobs that genuinely qualify:

• Distributing a fee stream. Your protocol earns revenue and you want the people who use it to own a share of that revenue. This is the strongest case, because it's the one where a token is doing something equity structurally cannot do at internet scale.
• Bootstrapping a two-sided market. You need liquidity providers, node operators or suppliers before you can pay them in cash, and the token buys their participation against future value. Legitimate — but it is a loan from your future self, and the interest is the sell pressure you'll owe later.
• Governing parameters that are actually governed. Fee rates, collateral factors, listing decisions. If your "governance" is a forum nobody reads, this isn't your reason.

If none of those describe your project, you are issuing a fundraising instrument with extra steps — and in 2026 the regulators have finally noticed. More on that below.

If one of them does, the shape of your launch follows from the shape of the project:

Everything from here assumes the right-hand column.

Start from the float, not the valuation

The single most consequential number in your launch is not your valuation. It's your float: the share of total supply that is actually tradable on day one.

Float and fully diluted valuation are two ends of one lever. Price is set by supply and demand on the float alone, but the headline number the world quotes — FDV — multiplies that price by every token that will ever exist. Launch with 10% float and a $50M market cap and you have announced a $500M FDV to grow into, while 90% of your supply queues behind a calendar everyone can read.

The market has learned to price this. Two measurements of the 2024 cohort, from different sources, bracket the problem: Binance Research put the median market-cap-to-FDV ratio at 12.3% — buyers entering structures where 87.7% of supply was still locked — while Tokenomist's annual report across 378 tokens found the average circulating-supply-to-FDV ratio at issuance had risen to roughly 35% by the end of that year. Neither figure is a target anyone has established as optimal; they are the range the market actually launched in, and the results sorted almost perfectly by size: in the Arrakis sample, tokens launching above $1B FDV finished the year with 0% trading above their listing price and a median drawdown of 81%, while tokens under $100M FDV were three times more likely to sustain positive performance through month one than tokens above $500M.

Read that as a pricing rule, not a morality tale. A high FDV asks the market to underwrite future dilution in advance, at the moment it knows least about you. The cheaper promise is easier to keep.

Three practical moves:

• Pick the float you can support, not the one that flatters the raise. Float is only as real as the depth behind it: work out the liquidity you can actually fund, then set float so the pool can absorb a bad day at that size.
• Express float in days of volume, not percent. "18% circulating" tells you nothing. "The float is 40× our projected daily volume, and the first cliff adds another 25×" tells you exactly how long the market needs to chew through what you're about to hand it.
• High float isn't automatically safe either. Lighter distributed roughly 25% of LIT's supply at its January 2026 launch and the token still fell about 32% within hours. Float relative to depth is what matters; a big float into a thin pool is just a faster drawdown.

The allocation table

Allocation is where founders most often copy a slide from a project they admire without noticing that the project had different obligations. Here are the benchmarks the market currently treats as normal, drawn from Streamflow's vesting design benchmarks:

That last row is the one worth arguing about. Most allocation tables treat liquidity as a rounding error or fold it into "treasury," and then discover on launch week that the treasury is entirely denominated in a token whose price they are trying to defend. You cannot buy your own token with your own token.

Budget a liquidity and market-operations line explicitly, and fund part of it in quote assets — ETH, SOL or stablecoins — before TGE. It pays for the initial pool, the depth you add as volume grows, and the inventory you need to bid when everyone else is selling. Fund it from a private round if that's what you have; just don't let the last stablecoin leave the building before launch day.

One more constraint, even though the law didn't land: the CLARITY Act's "mature blockchain" test carried a hard 20% cap on any single entity's ownership or voting power. The bill stalled in the Senate on 15 September 2026, but the 20% figure is now a de facto analyst benchmark. If your team or foundation holds a third of supply, expect to explain it in every diligence call for the life of the token.

Your unlock calendar is a trading schedule

Here is the reframe that makes vesting design easy: an unlock calendar is the one part of your future supply that the whole market can read in advance. Design it like a trader would read it, because that's who reads it.

Be precise about what an unlock is, though. It makes tokens transferable; it does not make them sold, and plenty of unlocked supply sits still for years. What you know in advance is the date and the size, which is enough to plan against. So every claim below about unlock pressure is conditional: if some share of what unlocks is sold, here is what it meets.

The current norms, and the reasoning behind each:

Four rules do most of the work:

• Nothing from founders, team, or private investors unlocks at TGE. Not 2% "for liquidity," not 5% "for operations." The moment team and investor allocations can be sold on day one, every subsequent conversation about alignment is theatre, and sophisticated buyers will discount your token for it immediately.
• Never let two restricted buckets cliff in the same month. Offset stakeholder groups by at least 60 days. Two cliffs landing together isn't 2× the supply — it's a supply wall that traders can see six months out and position against.
• Prefer streaming to step functions. A monthly cliff is a scheduled event; a per-second linear stream is background noise. Large single-day cliffs, as the 2026 unlock-design consensus puts it, "overwhelm the market and create unnecessary selling pressure," and the market front-runs the date anyway — the drawdown usually starts before the tokens actually move.
• Put it on-chain, and make the contracts match the document. In 2026 a vesting schedule that lives only in a PDF is a promise, not a constraint — and "here is the contract, check it yourself" beats any paragraph of adjectives. Be precise about what it guarantees, though: a vesting contract is only as rigid as its permissions. Cancellation, recipient changes and admin rights are configuration, not physics; Streamflow, for one, makes cancellation a setup-time choice that cannot be added later. So publish the schedule and what the contract allows: who can cancel, who can reassign a recipient, whether it is funded up front, and which keys hold admin rights.

And publish the calendar as a table, with dates and percentages of supply entering circulation. If you don't, someone will publish their reconstruction of it, and theirs will be less flattering than yours.

The airdrop is a sell order you scheduled

An airdrop is the fastest way to create a float and the fastest way to create a problem. The data on how recipients behave is brutal and consistent: in poorly designed drops, 78% of wallets sold inside the first 24 hours, and between 50% and 70% of flat-distributed tokens are gone within 30 days. Across 21 audited protocols, top-quartile day-90 retention was 41% — and the median was 6%.

Then there's the farming problem, and its scale is the point: Linea's drop filtered roughly 517,000 of 1.3 million eligible addresses as sybils — about 40% of all claimants. LayerZero and zkSync are often set side by side here (800,000 addresses filtered and a 16% month-one decline, against 47,000 sybils admitted and a 39% drawdown), but treat that as illustration, not proof: two launches, different tokens, different conditions, no way to isolate filtering from anything else. The reason to filter is mechanical rather than statistical — supply handed to wallets that farmed you is supply with no reason to stay.

So design the drop as the supply event it is:

• Points over snapshots. A 60–90 day points program with sybil-resistant scoring rewards sustained behaviour rather than a single block height that mercenaries can trivially farm. Snapshot drops select for people who were paying attention to your snapshot, not to your product.
• Reward verified actions, not balances. Quest-based campaigns requiring multiple verifiable on-chain actions, combined with sybil filtering, have produced retention above 73%; standard giveaways produced 11%. The delta is the entire ballgame.
• Vest part of the drop. A claim streaming over three to six months converts a day-one sell wall into a slope. If you want forfeited portions to return to the treasury, that must be an explicit, disclosed term and a mechanism the contract supports — it does not follow from vesting, and it cannot be retro-fitted on most setups.
• Use a claim deadline. Unclaimed supply returning to the treasury is free float reduction, and a deadline forces the engagement you wanted in the first place.
• Size the drop against your pool, not your ego. The only number that matters on the day is drop size ÷ pool depth. If 30% of the airdrop hitting the market in a day would move your price 60%, your drop is too big or your pool is too thin. Usually both.

And be realistic about the ceiling: you cannot retain a mercenary, only size their exit so it's absorbable. Airdrops that work are the ones where the people who leave don't break the chart on the way out.

Liquidity: size it backward from sell pressure

This is the section most tokenomics guides skip, and it's the one that decides whether everything above survives contact with the market.

The first mistake is the metric. TVL is not liquidity. The question that matters is: how far does the price move on a $10,000 buy or a $50,000 sell? Arrakis's numbers make the difference concrete: a $500K full-range 50/50 position takes roughly 4% slippage on a $10K trade, while $1M deployed as a concentrated position takes roughly 1% on the same trade. And thin books don't just cost slippage — pools with less than $20K in 2% depth showed 237% annualized volatility, versus 171% for pools above $165K. Depth is a volatility control, not a vanity metric.

The second mistake is the direction of the calculation. Most teams allocate liquidity forward — whatever's left after the CEX commitments and the treasury reserve. Size it backward from the sell pressure you have scheduled: your airdrop, your first cliffs, your ecosystem emissions. You know those numbers exactly, because you wrote them. If the capital left over can't absorb the supply you've committed to release, you don't have a liquidity problem — you have an allocation problem, and it's still cheap to fix.

Then the mechanics:

• Own your liquidity; don't rent it. Renting means emissions to mercenary LPs who leave when the rewards stop and sell them on the way out — you pay twice, in dilution and in sell pressure — and rented depth can be withdrawn at the worst moment. Rent only as a bridge to owning.
• Usually full-range first, concentrated after. In the opening hours direction is unknown and any range is a guess, so full-range takes the volatility without going out of range; a team that knows where price should sit — a stable-paired asset, say — may reasonably start concentrated. Once price finds a band, concentrated positions can deliver several times the depth per dollar (3–4× is reported for actively managed ones), but that multiple depends entirely on the range and the path price takes: let price leave the band and the advantage goes to nothing while the position stops quoting on one side. A managed position, not a free upgrade.
• Open at a high fee tier and taper. On day one there is no competing pool for aggregators to route to, so a high fee costs you little of the flow you actually want while taxing the snipers heavily — though it is not free either, since some buyers will simply wait for the taper or not trade at all. Solana teams get this as a primitive — Meteora's DAMM v2 fee scheduler can open a pool at 50% and decay to 0.25% over two hours — and on EVM you approximate it by launching in a higher fee tier and migrating down as trading normalizes.
• Choose the quote asset deliberately. Pairing against ETH or SOL produces less impermanent loss for a governance token, because both sides tend to move together. Pairing against a stablecoin exposes the full price divergence — right for yield-bearing assets and RWAs, punishing for a volatile new token.
• Lock the LP, and make the lock checkable. For a project with a product, a locked pool is a cheap, legible trust signal — it removes the single most common accusation before anyone makes it. Use the lock the venue's own explorers recognize, and link it from your docs.
• Separate the position you lock from the position you manage. You cannot lock liquidity and then migrate its fee tier or re-range it — a lock is precisely a promise not to touch it. Run two: a locked seed position that never moves and is the trust signal, and a managed position alongside it carrying the tapering and re-ranging. The exception is a venue that decays fees inside the pool itself, like Meteora's scheduler, where no capital moves.
• Configure the MEV protections; they are not defaults. Position changes through a private path, price-range guards so a liquidity add reverts if price moved, and dynamic fees during volatile periods. Every dollar a sandwich bot takes from your treasury's own transactions is liquidity budget spent on someone else's infrastructure — the same mechanics we've covered for traders, at larger size.

How the tokens actually reach the market

Distribution mechanism is a smaller decision than founders expect, but the options have diverged since the last cycle:

On venue sequencing, the current consensus is straightforward and worth following: open on a DEX, then negotiate the CEX listing with real data in hand. A DEX pool has no gatekeeper and no listing fee, and thirty days of organic volume is a far better negotiating position than a deck. Listing fees in 2026 run from around $15,000 at smaller venues to $500,000+ at tier one, with a credible mid-tier campaign landing between $100,000 and $300,000 all-in once liquidity, market making, audits, and legal are included. Arriving with proven volume is how that number comes down.

The market maker question

At some point in the month before TGE, a market-making desk will send you a term sheet. Understand what you're signing, because the standard structure has incentives pointing in a direction founders consistently misread.

The common loan + call option model works like this: you lend the desk 0.5–2% of supply so it has inventory to quote with, and the desk holds a call option to buy those tokens later at a preset strike. Two things follow mechanically. First, to hedge the inventory risk and get the stablecoins it needs to quote the bid side, the desk is structurally required to sell a portion of the loan. That isn't malice; it's what the position requires. Second, the option pays out above the strike — which ties a chunk of the desk's compensation to where the price goes rather than to how orderly the book is. You wanted tight spreads and you have accidentally bought a directional bet.

The documented failures are variations on one theme. Movement Labs routed 5% of supply through a middleman under side agreements its own leadership didn't fully understand; wallets tied to the market maker dumped $38 million on listing day and the company filed for Chapter 11 the following year. Lighter's January 2026 launch was cleaner but still cost it a trust cycle: the disclosure that Jump Crypto had received more than $24 million of LIT for market-making, plus an ongoing fee position, landed after the fact. The arrangement was ordinary; the sequencing was the mistake.

Binance now publishes what it watches for, and the list doubles neatly as your own diligence checklist. In its March 2026 trading-rules update the exchange flagged: selling that doesn't line up with unlock schedules, persistent one-sided flow leaning toward sells, large coordinated deposits and sales across multiple exchanges, volume that's high while price barely moves, and violent swings in thin books. It also told projects to avoid profit-sharing or guaranteed-return arrangements with desks, and to define asset usage explicitly in any token loan.

If you do hire a desk, the terms that protect you are unglamorous:

• Pay a retainer in cash or stablecoins. The market has been drifting this way for a reason — it decouples the desk's pay from your price, which is the whole problem with the option model.
• Define what the loaned inventory may and may not do, in writing, with the unlock calendar attached.
• Demand independently verifiable reporting. If the only record of performance comes from the counterparty, you don't have reporting, you have a claim.
• Disclose the arrangement before launch, not after. The structure is normal; being found out is what costs you.
• Keep the wallets visible. You should be able to watch the desk's addresses in your own tooling and see exactly what they did in your pool.

And consider the option that didn't exist a few years ago: doing it yourself. The reason projects outsourced market making was never that the strategy was esoteric — it's that the execution infrastructure was. That's changed, and it changes the negotiation. A team that can run its own book is a team that can walk away from a bad term sheet.

Value accrual: what makes the next unlock survivable

Every token you've scheduled to unlock is supply. The only durable answer to scheduled supply is scheduled demand — which is why 2026's defining tokenomics trend is projects routing revenue back into their own token.

The numbers are no longer small. Crypto projects spent a record $638 million on token buybacks in 2026, with Hyperliquid (~$370M) and Pump.fun (~$200M) alone about 90% of it. The fee-switch debate is effectively over: Ethena's September 2026 vote passed with zero votes against, routing 95% of net revenue to buybacks once USDe supply reaches $7.5B — and Pendle (80% of swap and yield fees), GMX V2 (27% of fees to stakers) and Maple (10–30% of revenue, scaling with it) all run versions of the same idea.

The reference case remains Hyperliquid, instructive because the tokenomics were decided before the token existed: over 70% of supply to the community, 31% airdropped to roughly 94,000 users at TGE, no VC round at all, and an assistance fund buying HYPE on the open market with the overwhelming majority of trading fees. No discounted early-round supply means no cliff overhang; revenue-funded buying means a structural bid that appears on no emissions schedule.

But buybacks are not a substitute for supply discipline. Castle Labs examined six protocols — Aave, Aerodrome, Hyperliquid, Pump, Sky and Uniswap — generating $726 million in H1 2026 revenue between them, and found that after emissions, unlocks and incentives, net value flow to holders turned negative for several. A buyback returning less than your emissions remove is a press release, not a policy.

So keep one equation on the wall — but write it in units that survive contact with an analyst. Everything below is dollars per quarter at the current price, and every token is counted exactly once:

Net flow = (buybacks + fees paid to holders) − (newly sellable supply × the share of it you assume is sold)

Three things make that harder to fool yourself with than it looks:

• One bucket, two sources. Tokens you mint (emissions) and tokens that merely become transferable (unlocks) are different events — one dilutes, the other only changes who may sell — but both put supply within reach of the market, so for this purpose they belong together.
• Count each token once. Liquidity mining, points rewards and partner incentives are almost always paid out of emissions; subtracting them separately double-counts them.
• The sell-through share is an assumption you own. There is no correct value — a farming cohort behaves nothing like a team cliff — so run several and state the number whenever you quote the result. An equation with a hidden assumption is worse than no equation.

If net flow is negative across plausible sell-through assumptions, the token is being supplied faster than it is being bought, and no narrative fixes that. If it holds positive even at pessimistic ones, you have something rarer than a story: a reason for the price to exist.

The legal layer in 2026, briefly

This is background, not legal advice. Whether your token is a security, where you may offer it and who may receive it are questions for securities counsel in every jurisdiction you touch — before the allocation table is final, not after. What follows is only the shape of the ground.

The legislative route stalled: the CLARITY Act passed the House with 294 votes in July 2025, then failed its Senate cloture vote on 15 September 2026 — 49 for, 50 against, eleven short of the 60 needed to break a filibuster. The regulatory route moved instead: in August 2026 the SEC proposed Regulation Crypto Assets, a rulemaking that would create the agency's first purpose-built offering regime for tokens. In the accompanying statement, Chairman Atkins describes a startup exemption of up to $5 million over four years, a fundraising exemption of up to $75 million annually with audited financials at certain thresholds, and an investment-contract safe harbor under which, if the issuer certifies it has ceased the essential managerial efforts it promised and satisfies certain other conditions, the Commission would no longer deem the asset subject to an investment contract. Those other conditions will decide whether the safe harbor is usable, so read the proposal rather than a summary of it.

Two design implications, regardless of how the proposal lands:

• The safe harbor rewards genuine decentralization on a deadline. If "we will eventually hand this over" is in your materials, build the plan that makes it true and dated. That is now a regulatory asset, not just a value.
• Concentration is what gets measured. Between the CLARITY 20% test and the disclosure-heavy direction of everything else, an allocation where team and investors hold a third of supply is one you will be defending for years. Cheaper to set a distribution you can stand behind than to explain one later.

In the EU, MiCA has made compliant public sales a product rather than a risk — which is exactly why rails like Legion and Kraken Launch exist. If you have European users, that path is worth pricing.

The practical summary for a founder: don't self-assess your token's status, don't design your distribution to route around a rule, and don't let the tokenomics get locked before counsel has seen it. Retro-fitting a vesting contract is expensive; retro-fitting a disclosure you should have made is worse.

Write the trading policy before you need it

Most tokenomics guides omit this entirely. Your team's selling is potentially a matter of securities law, exchange listing agreements and — at minimum — the disclosure standard your own investors hold you to. The safe posture is the boring one public companies converged on decades ago: decide the rules in advance, publish them, then execute mechanically.

What that looks like in practice:

• A written treasury and trading policy, published before you trade. What the treasury sells, why, on what schedule, within what limits, and who authorizes it. Publishing it costs you nothing — the market has already assumed you will sell something — and it converts every future sale from a discovery into a disclosure.
• Pre-scheduled rather than opportunistic. A plan adopted while you hold no undisclosed material information, executing on dates or rules set in advance, is what keeps a routine sale from looking like a timed one. Well-worn in equities; still a differentiator in tokens, which is why adopting one is cheap.
• Blackout windows around material news. No treasury or team selling into the run-up to an announcement, a partnership, an exploit disclosure, or an earnings-equivalent update. This is the rule whose violation ends projects, and it is entirely within your control.
• Never trade on information your holders don't have. Knowing things the market doesn't is the permanent condition of running a protocol, not an edge to be used. Put that in writing, for everyone with access.
• Keep an auditable record. Logs of what was executed, when, by whom, and under which authorization. The point of a policy is being able to demonstrate you followed it.
• Disclose your market-making arrangements. Before launch, not after — see the Lighter example above. Same for any supply lent to a desk.

None of this makes your token safer from the market. It makes it safer from the entirely avoidable category of damage: the kind where the trading was defensible and the disclosure wasn't.

The part that starts the day after TGE

Everything above is planning; this part runs for years. What makes it different from the memecoin case is not that one of them stops trading — both run continuous market operations — it's that yours are governed by a calendar. Every 30 days something unlocks. Every quarter a partner, an early investor or a team member has a liquidity need. Your treasury is denominated in a token you're trying not to sell carelessly, and your pool needs depth added as volume grows and defended when it doesn't. Four jobs never stop:

• Preparing for your own supply events. The cliff you published is a date the market has been positioning against for months. Having depth ready means unlock sales meet liquidity instead of an empty book.
• Executing disclosed treasury sales without destabilizing the market. Converting tokens into operating capital is normal and legitimate for a funded project — provided it is disclosed and scheduled rather than opportunistic. Good execution doesn't make a disclosed sale quieter than it should be; it means the sale needn't arrive as one destabilizing block. Sizing orders against live flow, or spreading them over time, lets the market absorb them in the depth that actually exists — the same reason equity desks use scheduled and participation-based algorithms instead of one market order.
• Maintaining depth as the token grows. The pool that was correct at $5M market cap is thin at $50M. Depth is a maintained position, not a launch task.
• Knowing who is actually selling. An unlock recipient distributing quietly, a desk hedging its loan, a sybil cohort finally cashing out, and organic profit-taking all look identical on a price chart and completely different on a labelled trade feed.

The execution layer underneath those jobs matters more than the strategy sitting on top of it — landing rates, MEV protection, who builds your transactions and what they skim, whether your data is streamed or polled. We wrote the full buyer's checklist in how to choose a market making bot, and it applies identically whether you run operations in-house or hire them out.

A 30-day pre-TGE checklist

None of this takes a quarter. The thinking in this guide is the slow part; the execution is a month, and most of it runs in parallel. The only genuinely long-lead items are an audit and — if you're airdropping — the points program that earns the distribution, and neither blocks the rest of the list.

Running both halves on Sumo

The reason this guide ends here is that a tokenomics launch has two halves that are almost always bought from two different vendors — a launch service for day zero, a market-making desk for everything after — and the seam between them is where founders lose custody, information, and money.

Sumo is both halves in one place, running from your own wallets:

• Launch it yourself. Direct pool launches on Meteora DAMM v2 (SOL or USDC quoted), Uniswap V2 on Ethereum, Aerodrome and Uniswap on Base, PancakeSwap on BNB Chain, and more — with the parameters this guide is about as fields in the form: total supply and decimals, the percentage of supply paired into the pool, the pool fee tier, per-wallet supply limits, and an option to lock the LP position once trading opens. No launch dev holds your keys, your authorities, or your timing.
• Distribute it. Airdrop campaigns to up to 50,000 addresses per run — batched, monitored, and targetable at existing holders of another token with balance filters, which is how a real points or usage cohort gets paid.
• Then run the market you just created. DCA and limit orders for the scheduled and price-triggered execution a published trading policy calls for, and Counter Trade for sizing orders to live flow so a disclosed sale is absorbed by the depth that's actually there — all executing from your own multi-wallet treasury across Solana, Ethereum, Base, BNB Chain, Robinhood Chain, Tron and more.
• Keep the record. Trade logs, treasury logs, and audit logs of every sensitive operation. If you've published a trading policy, this is the evidence that you followed it.
• See who's on the other side. The live trading feed tags every transaction in your pool in real time — MEV attackers and victims, arbitrage, holder tier, first buys and full exits, the platform each trade came from, and wallets you've labelled yourself. On unlock day, that's the difference between watching a price and reading a cap table.
• Keep the keys. Seeds and private keys are sealed with envelope encryption and isolated in a dedicated signing service; they never touch the application database or the logs, and no ordinary API response returns one. The deliberate exception is Key Export — off by default, owner-gated, typed confirmation, audit-logged, and it permanently flips that wallet's custody to shared — because your keys should be yours to take. The design is published, which is the minimum any provider holding your treasury should be willing to do.

Already launched elsewhere? Import the token and the operations half works the same — the strategies don't care where the token was born.

FAQ

What is a tokenomics launch? A launch where supply, distribution and release schedule are designed in advance — an allocation table across team, investors, community, treasury and liquidity; a vesting calendar with cliffs; and a funded liquidity plan. The counterpart to a fair launch, where effectively all supply hits the market at once by intention.

What percentage of supply should be circulating at launch? Higher than the market's recent habit. Published measurements of the 2024 cohort range from about 12% to about 35% depending on what is measured, and none is an established optimum — but the low-float, high-FDV end clearly performed badly: tokens above $1B FDV finished that year with none trading above listing price. Set float from the depth you can actually fund, and check it in days of projected volume, not as a bare percentage.

How long should team tokens vest? A 12-month cliff followed by 36–48 months of monthly release is the current norm for founders and core team, with nothing unlocking at TGE. Investors typically take a 12-month cliff (18–24 for the earliest rounds) and 24–36 months linear. Offset the different restricted buckets by at least 60 days so two cliffs never land in the same month.

How much liquidity do I need at TGE? Enough that a realistic sell doesn't move price more than you can accept — measure impact at $10K and $50K rather than tracking TVL. For scale: $500K full-range takes about 4% slippage on a $10K trade, $1M concentrated about 1%, and pools under $20K in 2% depth run materially higher volatility. Size it backward from the airdrop and cliffs you've already scheduled.

Do I need a market maker for a token launch? You need market making — continuous depth, disclosed treasury sales executed without destabilizing the book, depth on unlock days. You don't necessarily need a market maker: a counterparty holding a loan of your supply with a call option whose payoff depends on your price. If you hire one, pay a cash retainer rather than a loan-plus-option, define what the inventory may do, demand independent reporting, and disclose it before launch.

When can a project team sell its own tokens? Under a policy you wrote and published before trading — schedule and limits set in advance, blackouts around material announcements, adopted while you hold no undisclosed material information. The failure mode that ends projects isn't selling; it's selling on a timetable that, in hindsight, lines up with news only the team had. Whether your token carries formal restrictions is a question for securities counsel in each jurisdiction you touch.

Should I launch on a DEX or a centralized exchange first? Usually DEX first. There's no gatekeeper and no listing fee, you control the opening price and depth, and thirty days of organic volume is a far stronger position from which to negotiate a CEX listing — which in 2026 runs from roughly $15,000 at smaller venues to $500,000 and up at tier one.

Do buybacks fix bad tokenomics? No. They're the strongest demand-side tool available — a record $638M industry-wide in 2026 — but a buyback returning less than emissions and unlocks put into reach is cosmetic. Six protocols generating $726M in H1 2026 revenue still showed negative net flow to holders. Fix the supply schedule first, then fund the bid.

Can I launch a token with real tokenomics without a launch dev? Yes, and you should want to. Custom token contracts are where the catastrophic bugs live, and the operational steps — deploying, seeding the pool, distributing, running strategies — are tooling, not expertise. The reason to keep it in-house isn't cost; it's that nobody else ends up holding your keys, your authorities, or advance knowledge of your unlock timing.

A memecoin lives or dies on attention. A token with a product behind it lives or dies on whether its supply schedule and its demand schedule were designed by the same person, on the same day, with the same spreadsheet open. That's the whole discipline. Everything in this guide is one of those two columns — and Sumo runs both halves of it, from your own wallets, with nobody else holding the keys.

Related Sumo docs

• Launch a new token — the wizard that covers supply, decimals, pool liquidity, fee tier, and LP locking.
• Airdrops — batched distribution to up to 50,000 addresses, with holder targeting and balance filters.
• Strategies overview — Counter Trade, DCA, limit orders, and instant trades against your own treasury.
• Live trading and surveillance — per-trade MEV, arbitrage, holder-tier and platform tagging; tracked wallets for desks and unlock recipients.
• Platform fees — what launching and trading on Sumo costs.