I joined @KaitoAI as Marketing Manager.
for the past two months, i've been helping shape how Kaito’s story gets told. this role brings together so much of what I love.
but more importantly, it feels like play to me, because i truly believe that the intersection of crypto, AI, and data will shape the next chapter of the attention economy.
very grateful to @Punk9277@Tianqi_Wang5@blingoh, and the rest of the team for this incredible opportunity to create my best work while working on the industry-defining product.
New in Kaito Pulse: Activity Feeds.
A card in your X sidebar that puts two signals in one place:
- Calls: strong bullish and bearish takes from X and Substack
- Trades: what the people you follow are trading on Hyperliquid and Polymarket
Trade idea generation, on easy mode ↓
i think this sums up the creative process so beautifully.
the sheer number of iterations it takes for an idea to finally click is massively underestimated.
often, what separates good vs great work is the willingness to keep pushing long after most people would've stopped.
so vital in the ai era.
a full blueprint to building a frontier tech brand nobody remembers
(save this before you go to market/your next rebrand):
1. don't define what we are. let the market figure it out. they're smart, they'll get there eventually.
2. write a positioning line so technical that nobody outside the dev team understands it. then put it in the X bio AND the hero section of the website.
3. use the word "revolutionary" at least three times in the hero section.
4. make sure the website hero, pitch deck, and X bio all contradict each other slightly. consistency is for boring brands.
5. skip the ICP work entirely. talk to "everyone." retail, devs, institutional, VCs, the dog walker. one message for all.
6. avoid having any big ideas
7. pretend our category is unique. don't bother with competitive positioning. we have no competitors because nobody understands what we do.
8. when asked what makes us different, lead with the tech stack. mention the consensus mechanism within the first 10 seconds.
9. write our manifesto as a list of features. skip the "why we exist" part entirely.
10. let every team member describe the project in their own words on their own profile. zero alignment and shared language.
11. change the master narrative every quarter to sound fresh. previous quarter's content becomes orphaned, but I mean, who reads old tweets anyway?
12. don't bother with a brand voice doc. let every content writer interpret the tone freely.
13. avoid telling the founder's story. people care about the protocol, not the humans.
14. ignore the principle that brand-comms should sit next to the product and the pain we're solving for our ICP. treat it like a marketing deliverable you check off.
15. don't run the "explain it to my non-crypto friend" test on any positioning line.
16. let LLMs lump us into a commoditized basket of competitors. don't articulate what we do differently. ChatGPT & Claude will figure it out.
17. assume positioning is a one-time exercise. ship it once, never revisit it, never pressure-test it against new market context.
18. don't translate technical features into stories. the deck IS the narrative.
19. ignore that brand-comms drives cost of users, cost of capital, cost of partners, and cost of talent. it's just a marketing thing. it has nothing to do with revenue.
20. skip building a messaging hierarchy across audience phases. talk to pre-TGE retail and post-mainnet institutions with the same exact deck.
21. burn through brand-comms work every time we onboard a new partner agency. start from scratch every quarter.
22. write copy that hedges and add "leveraging," "synergies," + "supercharge" wherever possible. make sure 100 other projects could swap their name into our website and nothing would change.
23. don't audit how the project is being talked about externally. if KOLs, partners, and journalists are calling us something different than we call ourselves – that's their problem.
24. don't track which lines from our messaging actually get repeated by our community.
25. ignore that brand-comms compounds. underinvest now, scramble later, blame the market.
a few things i’ve learned about content creation lately:
> your first sentence matters more than you think
> stop trying to appeal to everyone
consistency matters more than perfection
> your personality is your biggest advantage
> make your content easy to read and understand
> pay attention to what gets people talking
>don’t be afraid to experiment with different formats
> share your own experiences, not just information
> create something you’d actually want to see on your timeline
3 extra tips for video creators:
> hook your audience in the first 3 seconds: skip long intros.
> use captions: a lot of people watch videos without sound.
> cut unnecessary pauses and keep your videos moving.
you don’t need to go viral to create good content.
sometimes getting the right 10 people to care is worth more than getting hundreds of random impressions.
Raising a round? Start with the investors already talking about your space.
Kaito’s VC leaderboard shows which funds and investors are attracting attention, with rankings and mindshare trends over time.
See who’s active in your vertical, read how they think, and build a sharper outreach list ↓
back when i was interviewing for my role at @kaitoai, i had to collect recommendations from a couple of the biggest projects i’d worked on.
it reminded me of something: clients don’t really remember your single cleverest idea.
i expected the praise to be about the creative. the narratives, the campaigns, the big ideas.
however, what they all pointed at was:
> “nothing slipped”
> “everything landed when it was supposed to land”
> “made every project feel under control, no matter the complexity”
nobody had to chase an update or wonder what was happening next.
that’s the part people actually remember: the feeling of a project being under control.
especially in tech/crypto, where everything moves fast and most teams already run on some level of chaos.
your content and marketing skills might be why someone hires you.
the follow-through, the peace of mind of knowing someone has it covered, is why they trust you and keep you around.
a lesson for anyone inspiring to go full-time into marketing in here.
I feel like we’re sitting on two extremes on tech Twitter rn.
it’s either the most disgusting ai-generated design you’ve ever laid your eyes on OR
the most captivating story-driven motion design/cinematic edit you wanna bookmark instantly.
nothing in between as of lately.
the big fugazzification of AI agents
the more ai agent demos i see, the more i think the gold rush analogy is almost too perfect.
california, 1849. everyone runs west because they think they’re about to strike gold.
but only 0,01% of people does.
the people who make consistent money are the ones selling shovels, maps, gear, and the promise of “this is where you should dig.”
ai right now feels very similar.
to be clear, the gold is very real - i can't imagine my workflows without llms some very specific agents bring real value.
but the panic around it?
> “you’re already behind.”
> “this replaced 10 employees.”
> “claude code just did 3 weeks of work in 4 hours.”
that part is the product.
and a lot of people selling the product also happen to sell the shovels.
courses. templates. agency offers. done-for-you automations.
i’m not saying they’re lying.
i’m saying their incentives are not perfectly aligned with your objective reality.
they need urgency.
because urgency makes you click, subscribe, buy, repost, and feel like you can’t afford to think clearly.
imo the boring truth is this:
agents are useful for narrow, repeatable tasks where the output is easy to verify and mistakes are cheap.
for instance, i connected @KaitoAI's mcp to claude and set a weekly routine to pull the smart followers data for the recap, so i can take over editing the draft from there.
but that is very different from “you can replace skilled people” or “you don’t need to learn the craft anymore.”