Hi
After 3 months of asylum, we are back with even more practical web3 marketing stuff! So expect a new letter 1x or 2x a month depending on what I can handle — usually on Saturdays at this time.
As you already know, AI is hot now. So much so, that it has become easy to build beautiful nonsense with it.
Those cute tools that don’t solve a real problem, don’t 10x productivity, and don’t increase revenue? That’s beautiful nonsense.
It’s safer to build these kinds of tools when you’re messing around with personal projects. But these experiments become costly in a business environment. And they are addictive. They can put you on a delusional high. So you think building more automatically makes things better. No it doesn’t.
There’s also the school of thought that believes “everything should be automated because we are in the AI era.” Quite frankly, I understand it.
Repeatable tasks must/should be automated whilst tasks that produce new insights should be polished.
But automation should only come in when a process has been clearly understood. One mantra I've corporately lived by over the past few months is “You can't automate a process you don't understand”. So, the struggle shouldn't be to automate everything, but understand the workflow first before deciding how automation can improve the process.
This is what I mean:
If you always input 1 and the output is 2, automate it so you can move on to other things.
For example, part of my routine at Pionex involves pulling data from multiple sources to find pages losing visibility, new citation opportunities, indexing issues and content that needs updating.
I shouldn’t spend every morning manually opening five dashboards, exporting the same data and comparing spreadsheets. The inputs are predictable and the process is repeatable. An agent can easily collect the data, run the checks and surface what needs my attention.
Automate that. But,
If you input 1 and the output could be 2, 5 or 27 depending on what you discover, don’t automate the thinking.
Let’s say the agent notices that searches around a newly disclosed keyword are suddenly rising and it aligns strongly with a brand’s products. The data can surface the opportunity, but it cannot predetermine what I should do with it.
Maybe the opportunity is an article, a tool or a landing page that matches search intent. Or maybe after investigating, I decide it isn’t worth building anything at all.
That part requires context, judgment and new insight.
Use AI to help you think through it. Don’t automate yourself out of the thinking process.
Trump’s stock portfolio
A few months ago, after carefully following the news about Trump’s stock portfolio disclosure, I thought : “This could be a good thing for Pionex to jump on”, knowing fully well that many of the assets are also available as tokenised stocks on Pionex.
NB ⚠️: Tokenised stocks are digital assets that track the price of traditional stocks, allowing you to get exposure to companies like Apple or Tesla through a crypto platform. They are not the same as owning the underlying shares directly.
New to tokenised stocks? Learn more HERE.
Trade Trump's 2026 stock portfolio
Now BACK to Trump
While finalising the campaign, I tried to look back in retrospect and pick out lessons:
1. Automate with intention.
This is the first stage of building anything with AI: the idea stage.
For me,I wanted to build a landing page that would help people discover popular assets from Trump’s portfolio that they could trade with USDT on Pionex.
But having an idea plus AI tokens, doesn’t automatically make the idea worth building. That brings me to the next rule.
2. Automate to solve a problem.
This is the intentional part that separates vibe-coding because you feel the vibe from building because you want to solve something.
For me, I didn’t want to create just another landing page. I wanted to build something that answered actual questions someone interested in Trump’s portfolio might have.
I wanted it to clearly answer questions like:
Can I trade assets from Trump’s stock portfolio on Pionex with USDT?
Which assets from Trump’s portfolio are available to trade on Pionex?
Btw you can trade hundreds of real-world assets on Pionex. Just click HERE to start.
New to bot trading? Try out the Pionex Trading Bot Finder Test to find out which bot you should start with as a beginner.
3. Automate to increase revenue
This is one of the things that differentiates junior marketers from senior marketers.
Solving the immediate problem wasn’t enough. I also had to ask: How does this connect to the rest of the business?
Pionex isn’t only a place to trade tokenised stocks. It’s also known for its crypto trading bots. So it wouldn’t have made much sense to send people to the stock markets and end the journey there. So I added another layer.
Depending on the market and product availability, someone exploring Trump’s portfolio could also discover relevant Pionex bots such as Grid Bot, Futures Grid, DCA Bot and Rebalancing Bot, with links to learn how each one works.
Now one search intent can potentially introduce a user to multiple parts of the Pionex product.
That’s the difference I’m trying to get at. AI made building the page faster. But the value came from deciding what deserved to be built, what problem it should solve and how that solution could contribute to the business.
Trade Trump's 2026 stock portfolio
Some housekeeping
Not sure what trading bot to start with as a beginner? Take this test.
New to tokenised stocks trading? Use this guide.
Trade Over 80 Assets From Trump’s 2026 Stock Portfolio: Apple, Tesla, NVIDIA & others on Pionex
Trade RWA on Pionex either manually or using any of your preferred trading bots
That’s it for my letter to you today. I hope you’re able to try out some of the things you’ve learnt from this.
See you in your inbox again soon