🔖 Mark: building AI bookmark

The best consumer founders know the game. 35K+ of them read Consumer Startups every week.

Stay ahead. Get the playbook behind today's breakout startups.

Read time: 4 mins 21 seconds

Last month, I grabbed coffee with Eason Tang. 

He's 21, and he was getting ready to meet execs at a major fashion house to discuss a potential collab: limited-edition drops of his product.

His product is Mark, an AI bookmark. It's a physical highlighter and bookmark for real paper books. You scan a line that hits you, whisper a thought, and slide it back in to keep reading. In the background, AI organizes everything you saved, summarizes it, and makes it searchable. 

This past May, Mark II launched with millions in funding and a video that did over 11 million views in a week, selling out the entire first batch of thousands of preorders.

I am excited to share Eason’s story today. Keep on reading to find out:

Eason has been obsessed with design since middle school, back when he was flipping sneakers and collecting playing cards. In 2020 he fell into Blender through the famous donut tutorial, eventually logging around 7,000 hours in 3D and doing concept work for teams at H&M, Travis Scott, and SpaceX.

However, making things for other people had a ceiling. He wanted a brand he could carry himself.

The itch that became Mark was personal. 

He wanted to write. Substack was booming, but he's a self-described terrible writer who found it painful to collect and organize his own thoughts. He wanted a tool that could do it for him.

Eason dropped out of USC and bought a one-way ticket from California to Shenzhen, his first real stretch in China since growing up in Wuhan. 

Building hardware there felt more real than any internship: factory floors, banquet dinners, and a crash course in Chinese. His favorite story is the first factory visit, when the boss handed him a cigarette he had no idea how to smoke, then watched him fail to light it and hold it awkwardly until it burned down on its own.

He launched the first version of Mark in February of 2025. It had a screen and a knob and saved lines by voice. 

The launch went viral, accumulating over 7M+ views on X. However, the nightmare scenario came soon after.

Activity flatlined overnight. His co-founder decided to leave. Skeptics said it was just "a highlighter plus a bookmark plus AI" and that your phone already did this.

Instead of defending the idea, Eason went back and ran 200+ user interviews. 

Two key learnings:

1/ Everything you read on paper is stranded from your digital life. People wanted their highlights digitized into something searchable, synced to tools like Notion. Almost nobody was seriously solving that.

2/ The first product was solving it the wrong way. Repeating every line out loud into a voice recorder was simply too much work. 

At the same time, people already have a behavior for marking what matters: a highlighter. More fortuitously, China has a mature scanner-pen technology that's been around for 20-plus years and moves over 20 million units a year. 

Eason realized he could reshape it into a bookmark. At the same time, those scanner-pen manufacturers were desperate for a new idea to expand beyond this aging technology. It was not hard to find a manufacturing partner.

He wanted to position Mark as a lifestyle product instead of an AI tool, thus appealing to the masses. 

In fact, there's no AI on the device at all. It scans text and records voice. All the intelligence lives in the backend: organizing, summarizing, connecting themes across books, and syncing to your tools.

Eason spent the next year working towards Mark II, an upgraded version of Mark I.

Mark II consists of two parts: a customizable bookmark that sits in your book and a separate smart highlighter. When you find an idea worth keeping, you use the highlighter to scan the text or record a quick audio note. The device logs your exact page and syncs the data to the Mark app on your phone.

He launched it this past May and it went even more viral, getting ~11M views on X and they sold out thousands of pre-order units in days. 

Here is what happened under the hood:

1/ Invite-only in-person launch event. Eason spent over a month producing an invite-only event in LA: a scene designer, a rented banquet hall, a full demo environment. The idea was to control Mark's first impression, letting people hold the product and preview the launch film before the internet weighed in. They got an assist from COLLINS, the design studio behind Robinhood, Cash App, and Mailchimp, and one Eason had followed since middle school, offered their office library as the venue.

2/ First X launch flopped. The initial plan was to launch online Tuesday, host the launch event at COLLINS on Friday, and let the hype walk in the door. Tuesday came, and the video did nothing. Barely any views.

3/ Then they brute-forced the algorithm. Over the next two days the team posted three or four different cuts and formats of the launch. On Thursday, one of them detonated: 11M+ views, right before the Friday event. 

One attendee flew in from Seattle, a five-hour flight, to try a bookmark. Many bought on the spot. The first batch of thousands of preorders sold out, purchased by folks like a16z investors and Stripe Press team.

For Eason, taste isn't "looks nice." 

It's history, realness, and depth, the layers you get by studying real objects from the 30s and 40s rather than scraping Pinterest. His reference is Leica: around 1980, while Sony and Canon raced toward smarter shutters and higher resolution, Leica doubled down on mechanical shutters and manual focus. Every shot is a deliberate choice about what's worth keeping. He thinks AI is at the same fork, an ocean of generated text on one side and the words a human chose to keep on the other. Mark is built for the second pile.

It's also why he thinks Humane and Rabbit died. They sold hardware as tech, to tech people.

So Mark is positioned as a lifestyle object, a status symbol. The way Hermès does it, the way Chanel does it.

The playbook from here is organic DTC plus brand collabs and limited drops, similar to Supreme or Pop Mart. Hence the fashion-house meeting after our coffee chat.

If you read my Town piece last time, this will rhyme. 

When AI itself becomes a commodity, the moat shifts to something human. For Town, it was delight and context. For Mark, it’s taste, physicality, and status.

That said, Humane and Rabbit also had taste, hype, and beautiful launch films, and they still died because the product underneath didn’t deliver. A status object that doesn’t work is just an expensive object. A viral launch video can earn the order, but only the product earns the reorder.

Right now, the hype for Mark still feels mostly confined to the tech-bro circle on X. Selling out thousands of preorder units is great, but the people who preorder a beautiful bookmark aren’t always the ones still reaching for it a year later.

Still, as an avid advocate for physical books, I see a ton of potential in Mark. It’s a better way to take notes and accumulate learnings in a way that feels natural and effortless. It’s perfect for readers who like to highlight and annotate, and for academic researchers who have to skim through hundreds or thousands of books each year.

I’m also excited to see more consumer AI products exploring new modalities and interaction patterns where AI isn’t the focal point, but something in the background powering daily workflows.

In a world where everyone seems to be outsourcing their thinking to AI, it’s refreshing to see a product that does the reverse and helps humans think better.

What do you all think? Reply to this email, I read every one :)

Leo

Follow me on X and LinkedIn

📥️ Want to advertise in Consumer Startups? Learn more.