The market has gone through retracements after retracements, with liquidity spreading thinner and thinner. New agents that recently launched successfully experienced a max MC of ~$10M. By “successfully,” I mean a product having PMF, providing value to actual users, and generating (or about to generate) revenue.

This is a stark contrast to 3-4 months ago when agents with PMF could go as high as ~$100M+ MC, especially if they positioned themselves as an agent + framework/launchpad token. For example, $AVA as a 3D agent also captures value from its launchpad and the projects it supports through its audiovisual layer.

The Old Playbook: Agent as a Framework

The playbook back then was to launch an agent to showcase its capabilities, attract demand from devs wanting to build their own agents, and require those devs to hold/burn/pay with the agent token to access the framework. The problem? CT assigns too high a premium to framework tokens while these “framework agents” often lack differentiation. In many cases, they don’t even have a product—they just yap on Twitter, hoping the token price goes up.

The first iteration of agents made the conversational agent itself the product. This is unique to crypto because we prioritize community building—akin to founder-led marketing (where the founder yaps to generate mindshare). Having an agent yap to generate mindshare for your project seemed like a good idea—it worked for a solid month when it first arrived in November 2024. Now, with 420,690 agents yapping non-stop, most are unsophisticated, repetitive, and, frankly, annoying.

The New Playbook: Agent as a Business

Here’s how you should think about launching an agent—

Launching an agent means you’ll be running a startup, managing up to three products at the same time:

1. Core Product (The Actual Business)

Your core product should solve real problems. It should not be just a conversational agent but an actual product.

Examples:

• Prediction models that improve betting odds, helping users win more in sports betting (e.g., @AskBillyBets).
• Crypto asset prediction models that enable better trading, minimize impermanent loss, and maximize LP returns (e.g., @Cod3xOrg, @gizatechxyz, @Almanak__).
• AI Agent research search engines that aggregate insights from top alpha sources like Cookie, Kaito, Nansen, Messari, Aixbt, CG, Dexscreener, and Bubblemaps to aid investment decisions (no team has cracked this yet—we need a Perplexity for AI Agents).

The core product should be every team’s first priority before launching a token. You need to ensure there’s an actual market for the product and that users are willing to pay for it. Otherwise, you’ll be stuck in the crypto version of the “Valley of Death,” which can be far worse than its traditional startup counterpart:

• High operational expenses.
• Paying customer acquisition costs (CAC) with a token.
• Token price tanks → reputation tanks → nobody cares about your project.

If your token dumps too much, it becomes a curse. Most people in this space won’t care about your project, regardless of progress or how strong your core product is.

Instead of relying on token incentives, focus on onboarding customers with your product. Figure out a monetization model that balances growth & revenue generation.

@KaitoAI’s playbook is a great case study:

• They built an enterprise product—a crypto search engine focused on social/sentiment/narrative—and charged $$$ to users, projects and ecosystems, providing real value.
• They introduced Mindshare Dashboard, which became the standard for tracking narratives and trends.
• They doubled down with the Yapper Leaderboard, making KOLs organically share it everywhere as a status symbol.
• They tripled down with NFT WLs and $KAITO airdrops, incentivizing Twitter engagement with tangible rewards.

Hard to replicate, but the lesson is to Find PMF first. Generate revenue. Make people excited before launching a token. Once you have attention (hype) and revenue, then take it to the next level.

Also, communication is key. Many projects have solid products but poor comms. If no one knows what you’re doing, no one will care.

2. The Token (Alignment Tool)

We’ve moved from “VC coins” to “fair launch” celebrating high float, low FDV coins. But fair launches aren’t truly fair—every token strategy has trade-offs.

If you’re launching an agent token with a high float, low FDV structure, you won’t be able to raise from VCs and angels (due to the low valuation). However, you can use the token as a marketing tool to bootstrap mindshare.

Many teams launch two tokens:

• Agent token → Bootstraps mindshare.
• Ecosystem token → Raises capital from VCs and angels at a higher valuation.

But this creates expectation misalignment—the community expects an airdrop, and when the ecosystem token launches, capital rotates from the agent token to the ecosystem token, crashing the former.

Managing core product + agent token + ecosystem token while ensuring value accrual to each is extremely complex.

In an ideal world, there should be one token that accrues all value from the core product. Historically, projects that generate revenue and redirect it back into the token (via buybacks or revenue distribution) survive long-term.

The token should be complementary to the core product—not a necessity.

For a deeper dive into agent token strategies, check out @VaderResearch’s breakdown of @virtuals_io’s agent token playbook:

3. The Agent (The Complementary Product)

The “agent” refers to conversational agents built using frameworks like ElizaOS, G.A.M.E, ARC, Pippin, etc.

While these agents integrate on/off-chain capabilities, they should be a complementary product to the core product.

An agent should enhance the core product by shifting the user funnel:

• Instead of users finding & using your product, the agent brings the product to them.
• This could mean: Using the agent to showcase the product directly on Twitter via text/video.
• Using the agent as an AI companion, changing how users interact (ChatGPT-style abstraction).
• The agent acting as the interface itself, executing tasks behind the scenes.

Exceptions exist. @aixbt_agent is an example—offering real-time social & sentiment insights from Twitter, , allowing users to get access to real-time alpha signals before anyone else. Aixbt became the #1 KOL on CT by consistently delivering alpha, showcasing the capabilities of the terminal. In this case, the agent itself is the product.

However, this is incredibly hard to replicate. Most should focus on strengthening their core product first.

A great product-first case study is @cookiedotfun:

• Started with a free AI Agent dashboard to acquire users.
• Transitioned to a freemium model, where locking $COOKIE unlocks premium insights.
• Monetized via an API for projects & agents.
• Launched @agentcookiefun to bring insights directly to Twitter.

Bringing it all together

In 2020-21, you needed Solidity knowledge to launch a token. Now, platforms like pumpdotfun make it easy to tokenize anything.

This has shifted the mindset—instead of building real products, people just launch tokens. It’s garbage in, garbage out—capital rotates to the next garbage.

We need to change this.

To build something sustainable, treat an agent project like a startup. Instead of farming CT, VCs, and angels, build something with lasting value—not for the next 6 months, but the next 6 years.

Innovate. Solve real problems. Create actual businesses—not just the next speculative token farm.

The future of Crypto AI agents depends on it.

Disclaimer:

1. This article is reprinted from [0xJeff]. All copyrights belong to the original author [0xJeff]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
3. The Gate Learn team does translations of the article into other languages. Copying, distributing, or plagiarizing the translated articles is prohibited unless mentioned.