Chapter 6: Rewriting the Rule book

When people talk about go-to-market strategy in Shenzhen, they're usually talking about the same playbook. It's a good playbook. It has been refined over decades by some of the smartest operators in consumer electronics, and it works with remarkable consistency for the right kind of product. You manufacture, you list, you advertise, you optimize, you scale. The feedback loop is tight, the metrics are clear, and the path from factory to customer is well-worn enough that you can almost run it in your sleep.

Almost every successful hardware company in this ecosystem has a version of this playbook on the shelf. Most of them built their first ten million dollars of revenue running it.

We tried it. It didn't fit.

Not because we executed it badly. Because Memoket isn't the kind of product the playbook was written for. The playbook assumes a customer who already understands the category, who has already formed the desire, who just needs to find the right product to satisfy it. Memoket was building a category that didn't exist yet. The desire hadn't formed because the gap it filled was one people had been living with so long they'd stopped noticing it.

You can't advertise your way to understanding. And Memoket required understanding before it required anything else.

So we threw the playbook out. And we started writing a new one.

This chapter is about that process. Not the theory of it, but the specific decisions, one by one, that became the first rules of a GTM approach that no one in Shenzhen was running. We didn't design it all at once. We discovered it, the way you discover most things that are actually worth knowing: by being wrong first, then paying close attention to what worked instead.

Rule one: find the right people, not the most people.

To understand where this came from, I need to go back to our first round of beta testers.

After Chinese New Year, with the "try everything" mindset firmly in place, we turned our attention to building our founding member community. The original early bird mechanism was a $5 deposit that secured your place in line for the full $199 package. Standard pre-order logic. Clean, simple, and built on an assumption we were starting to question: that people could understand enough about Memoket from a website to decide they wanted it.

We sat with that assumption and found it wanting. It had taken us, the people who built this product, actually using it to fully understand it. Not reading about it. Not watching a demo. Living with it, recording real conversations, watching the synthesis come back over days and weeks, feeling the moment when the context layer stopped being impressive and started being necessary.

If that was true for us, it was true for everyone.

And that led to a decision that felt counterintuitive by every conventional measure: instead of spending money on expensive advertising to drive pre-orders from people who didn't fully understand what they were buying, we would just give the product to people who wanted it.

Pay five dollars for shipping. Get a Gem.

Not a discounted Gem. Not a trial unit. A free device, yours to keep, in exchange for your willingness to use it and tell us honestly what you found. The hardware units weren't ready yet, so in the meantime we opened the full app, every capability, on both iOS and Android, completely free. No paywall, no limited trial, no gated features. The entire product, available immediately, because the only way to understand what we were building was to experience it the way we had: by actually using it.

We wanted 100 founding members. We surpassed that significantly in the first 48 hours.

That number told us something important. It told us the right people were out there, that they were reachable, and that when you lowered every barrier between them and the product, they showed up. The conventional wisdom says you protect access, create scarcity, make people work to get in. We went the opposite direction and the opposite direction worked.

Lesson: stop trying to convince people. Get the product into their hands and let it make its own case.

Rule two: the community is the infrastructure.

The feedback started arriving through Discord almost immediately.

Discord was its own small education. It's not a platform most people in Shenzhen's tech world use. The tools here are Lark, WeChat, DingTalk: deeply embedded, deeply familiar, the operating system of Chinese corporate communication. Discord was foreign territory for most of our team. Different conventions, different culture, a different way of being present in a digital space.

Our team jumped in anyway. They learned it fast. They showed up daily, read the threads, responded within 24 hours, engaged with founding members not as a support function but as genuine participants in a conversation about a product they were all building together. I watched our IT people navigate a completely unfamiliar environment with a fluency that surprised even me. It took courage, the specific kind of courage that comes from being willing to not know something in public and learn it anyway.

The founding members noticed. A community feels different when the people behind the product are actually in it, reading what you write, responding when you have a question, taking your feedback and coming back a week later to tell you what they did with it.

What came back through those channels was the most valuable market research we could have bought, and we didn't buy it. We got it by being present and listening. The app was strong. It also had rough edges: bugs, flows that confused people, features that needed more investment than we'd given them. All of that was specific, actionable, and grounded in real usage. Not survey data. Not focus group responses. Real people, in their real working lives, telling us what the product was actually doing.

Lesson: build the community before you build the funnel. The community is not a marketing channel. It is your earliest and most honest product team.

Rule three: find the channels where the right people already gather.

Building the community was one thing. Growing it was another.

We didn't have the budget for a major influencer campaign, and more importantly we weren't sure the ROI would be there for a product that required understanding before it required desire. Influencer marketing compresses consideration into seconds. Memoket needed more time than that.

Elisa suggested newsletters.

Not our own newsletter, not yet: the newsletters that already existed, built around communities of people who cared about technology, AI, productivity, and the future of knowledge work. People who read carefully. People who thought about their tools. People who would recognize the gap Memoket was closing because they'd felt it in their own working lives.

Newsletter marketing turned out to be one of the most efficient channels we found. The right readers, in the right mindset, arriving at our community already curious rather than needing to be warmed up. It was the opposite of interruption advertising. It was permission-based, interest-aligned, and cheap relative to what it returned.

No one in Shenzhen was running this play for a hardware product. We ran it because our motto was try everything, and this was one of the things we tried that worked immediately and kept working.

Lesson: go where your people already are. Don't build the audience from scratch when someone else has already gathered them around the right ideas.

Rule four: choose platforms that reward the real thing.

We mentioned Product Hunt earlier in this story, but there's a lot more I want to tell you about what that experience actually taught us.

When Edo first brought up Product Hunt as a potential launch platform, we were familiar with the name but not with the mechanics. The more we understood it, the more it fit the rulebook we were beginning to write.

We had looked at Kickstarter. We stepped back from it. The more we studied how crowdfunding campaigns actually succeed there, the more clearly we saw that the game rewards advertising spend and manufactured momentum as much as it rewards real products. That wasn't the environment we were looking for.

Product Hunt operates differently. It is a daily competition where new products are submitted and the community votes based on genuine interest. There is no mechanism for buying your way to the top that the community hasn't already developed immunity to. The people who participate there are real early adopters, real makers, real founders, people who have chosen to spend their time finding and evaluating new things because they actually care about new things.

For a product in a new category, trying to reach people who would recognize it before the broader market had a name for it, Product Hunt was exactly right.

What we didn't fully anticipate was what the process of launching there would require of us.

The entire office mobilized. Not the marketing team alone, not just leadership: everyone. We spent weeks inside the platform, learning its culture, reaching out to other makers, building genuine connections with founders who had launched there before, understanding what the community valued and how to participate in it authentically rather than transactionally. Those connections had value beyond the launch itself. Founders who understood the gap immediately. Early adopters who had been searching for something in this category and hadn't found it yet. Conversations that opened doors we hadn't known existed.

We also knew that the quality of your hunter mattered. A hunter on Product Hunt is someone with established credibility on the platform who submits your product, lending their reputation to the launch and helping surface it to the right audience. We networked deliberately and ended up connecting with the most respected hunter on the platform. Someone who receives more pitches than most people send emails. Someone who has developed, over years of doing this, one of the sharpest product instincts in the early adopter world.

We showed him Memoket.

He got it immediately. Not after a polished demo. Not after a carefully prepared walk-through. We explained the gap, showed him the product, and watched something shift. The same shift we'd seen on faces at Pepcom in New York, in our Discord threads, in the cramped meeting room in March when everyone had finally used the product enough to speak the same language. The shift that happens when someone stops processing information and starts recognizing something true.

In that moment, whether he hunted us became almost beside the point.

What mattered was that someone with a calibrated, battle-tested instinct for what is real and what isn't had looked at Memoket and known, without needing to be persuaded, that it was real. That kind of validation doesn't show up in a dashboard. It doesn't have a metric. But anyone who has spent time building something genuinely new knows exactly what it feels like when it arrives.

We launched on May 13, 2026. We finished number one product of the day. We wanted 50 founding members from the launch. We received more than 100 applications.

Lesson: choose platforms that are structurally incapable of rewarding anything except the real thing. Then make sure what you bring is real enough to deserve it.

Looking back at the rules we had begun writing during those months, what strikes me is how obvious each one seems in retrospect and how invisible it was before we found it. That's usually how it works with genuine insight. It doesn't feel like a breakthrough when it arrives. It feels like finally seeing something that was always there.

The Shenzhen playbook is a great playbook. We're not arguing against it. We're saying it was written for a different kind of product, and when you're building something that doesn't fit any existing category, the most important thing you can do is resist the temptation to force it into a shape it wasn't made for.

We had to find our own shape. These were the first rules we found along the way.

We were only getting started.