Issue 98: Launch Day is No Longer the Main Event. How Overseas Indie Developers Have Started Betting on the "Weeks of Waiting Lists"
What struck me most today were the people who abandoned the flashy launch.
Today, while observing the overseas indie development scene, what made me pause wasn't a flashy new feature or a viral hit. It was the opposite: a quiet tectonic shift where "the festival of launch day is no longer the main event."
Until recently, the winning strategy for indie development was the "fireworks of the day," such as hitting #1 on Product Hunt or going viral on X. However, in 2026, the most successful products overseas are shifting their acquisition focus to the "weeks before the launch." They are quietly pouring money and design into the "waiting list" built before the festival, rather than the day of the festival itself.
The reason this is so striking is that it is not a matter of "talent" or "luck," but entirely a matter of "design" and "sequencing." In other words, it is a domain that can be imitated by solo developers, side-hustlers, and small teams like you and me. Today, focusing on this trend of front-loading, I would like to dissect three vivid examples—(1) the revival of waitlists and referral loops, (2) AI-native onboarding, and (3) the mass production of faceless AI-UGC—along with the underlying flow.
I will write all numbers while distinguishing between public data, estimated values, and unverified claims. Since this newsletter calls itself "Time Machine Management," this is the part I want to value the most.
1. "The Visible Waitlist"—Building a queue before the launch day, not on it
Let's start with the most striking story. It is the "2026-style revival" of waitlists and referral loops.
Classic examples are Dropbox and Robinhood. After Dropbox added a referral program, about 60% of new registrations came through referrals, and the number of users grew from 100,000 to 4 million in 15 months (both are public, classic cases). In the case of Robinhood, they accumulated about 1 million pre-registrations just from the waitlist "before" the service was released (public, classic case). This is a story that has been told many times in marketing textbooks.
What I am focusing on again in 2026 is the fact that this old model is not just a "rehash," but is being re-evaluated, backed by the latest data. According to 2026 industry research, customers who come through referrals have about 37% higher retention, are about 4 times more likely to refer others, and the referral rate for top-tier programs reaches 22-25% (cited values from public industry research). In other words, it is at a level where "one in four people brings in the next person." The positioning of referrals has changed from a measure used "because it is cheap" to one used "because high-quality customers come in."
Adaptation to Japan (What is the blank space?)
Although Japan has a weak culture of lifetime deals (LTD), a waitlist that visualizes "rankings, limited slots, and deadlines" is actually a blank space. What you want to be careful about here is reward design. If you make referral rewards an easy "discount," it could violate the Act against Unjustifiable Premiums and Misleading Representations or stealth marketing regulations. Therefore, a smart approach is to replace rewards with non-monetary feelings of progress, such as "rank jumps" or "early access slots." The sharing flow should work well with the dual pillars of X and LINE. The difficulty of reproduction is low (★★☆☆☆), and it can be implemented with tools. The dividing line is just the discipline of the sequence: "Can you set it up before the inflow?"
2. AI-Native Onboarding—Replacing tutorials with "conversation"
The second point is about what happens after acquisition, which is onboarding (the initial experience). To sum up the change that occurred in 2026 in one word: "AI-native onboarding has been upgraded from an 'interesting experiment' to a 'standard initial experience.'"
Looking at the numbers, the importance of this area is clear. The median activation rate (the rate at which users reach their first value) in 2026 is 38% for B2B SaaS, 44% for fintech, 62% for e-commerce, and 35% for vertical SaaS (public benchmark research). And it is said that by leveraging personalization, the activation rate can improve by up to 40% (public commentary).
However, there is another number that I took most seriously. It is the fact that "the gap between the median and the top quartile has widened to about twice that of 2023" (public). To put it simply, the "cost of leaving an average onboarding as it is" has doubled in the last few years. What used to be fine with "well, a normal tutorial is fine" is now a structure where you will relatively lose if it remains normal.
Adaptation to Japan (What is the blank space?)
Domestic SaaS still mainly uses tutorial videos and links to help articles. That is precisely why you can get a step ahead just by adding a conversational concierge "lightly"—that is, with an LLM and just a few branches. You don't need to build out the entire screen. I believe the shortest realistic solution is to start by making just the first screen conversational and replacing attribute input with questions where "the initial screen changes based on the answer."
3. Faceless AI-UGC—"Running" ads without hiring creators
The third point is a new trend in the area of directly connecting short videos with products. The keyword is "faceless AI-UGC (synthetic UGC)."
Until now, in this series, UGC (user-generated content) has been treated as "having people shoot raw but authentic material with a high sense of involvement." However, in 2026, that premise is wavering. The movement to mass-produce UGC-style ads "without showing faces" using AI avatars, lip-syncing, and voice cloning is spreading rapidly.
The numbers speak for themselves. The number of advertisers using Meta's generative AI ad tools has increased about fourfold in six months, from about 1 million to over 4 million (public). One faceless UGC tool (VIDEOAI.ME) has announced that about 80% of its paid users are solo or micro-teams (public). Looking at the reference pricing for tools, Videotok can be used from about $31.20 per month (public). By the way, the market rate for hiring someone for UGC is said to be around $50 per video for beginners, over $2,000 per project for experts, and top creators are building businesses on the scale of $5,000 to $15,000 per month (public/industry commentary). The essence of this story is that solo developers have started to internalize this unit price structure using tools.
Adaptation to Japan (What is the blank space?)
In Japan, many people want to avoid showing their faces, and in that sense, there is an aspect that is compatible with synthetic UGC. However, at the same time, consideration for the Act against Unjustifiable Premiums and Misleading Representations, stealth marketing regulations (enforced in October 2023), and 'disclosure that it is AI-generated' becomes essential. Therefore, the first step is to safely try 'functional demos that do not require human performance'—such as screen recording or showing before-and-afters. Here, the issue of authenticity is relatively small, yet you can still receive the benefits of mass production.
Themes that emerge across measure types
When you line up the three cases, although they seem disjointed, you notice that there is actually one backbone running through them. It is the flow where 'the moment of the protagonist is moving from back to front, and from human to AI'.
Acquisition has been moved forward from 'launch day' to the 'weeks of pre-waitlist'. Onboarding has moved from 'tutorials to read' to 'concierges that change the situation on the spot through conversation'. And creative has changed its form from 'one video shot by a person' to 'hundreds of videos mass-produced by AI'. What they have in common is the attitude of stopping reliance on 'one-shot fireworks' and trying to rotate continuously with a 'designed mechanism'.
And there is one more hidden theme that should not be overlooked. It is 'discipline' and 'regulation'. The discipline of the order of 'setting up before inflow' for waitlists determines victory or defeat. Synthetic UGC will lose trust in one go if it lacks consideration for regulations such as stealth marketing regulations and AI disclosure. Rather than flashiness, plain order and etiquette are effective—this is the texture of selling in 2026, I feel.
This concludes the educational part on 'why it grew, market proof, and where the blank space in Japan is'. From here on, we enter the 'backside' for those who actually move their hands. I will write in order about the concrete flow design for each of the three measures (what to show on the screen immediately after registration / questions to leave in conversational onboarding / funnels to rotate synthetic UGC as consumables), the tools to use, and the initial actions you can take starting tomorrow.
The backside of ①—Technology x Gritty Marketing: The battle is decided on the 'screen immediately after registration'
This is the core to ensure this doesn't end as just a success story. The real weapon of this measure lies in the UI/UX of the screen immediately after registration (post-signup).
An LP that just collects email addresses can no longer be called a waiting list. The design that is growing shows the following four-piece set the moment you register. First, your current ranking number. Second, a unique referral link with a copy button. Third, one-tap share buttons for X, LinkedIn, WhatsApp, and email. Fourth, the next milestone you can aim for (how many more people to refer to get what). This visualization of 'ranking' and 'a little more for a reward' becomes the fuel that moves people.
The key points in operation are also gritty and concrete. First, rewards are paid only for 'confirmed referrals'. If you pay for raw registration numbers, the leaderboard gets dirty, and even your email sending reputation is damaged. Second, the leaderboard is not a private dashboard, but is intentionally made 'public'. The visibility of competition itself becomes social proof, which pushes up the volume of invitations. Third—this is the most overlooked—the referral program is set up 'before' flowing traffic. If you add it later, the first 200 people who come will pass through without ever having an 'opportunity to share'.
Robinhood's initial screen embodied this philosophy to the extreme. One input field, one line of promise ('commission-free trading'), and after registration, all you see is your ranking and a share link to move the line forward. It is thorough subtraction (public).
The backside of ②—Technology x Gritty Marketing: Only leave questions where 'the next screen changes when you answer'
So, what exactly is the growing initial experience doing? It is shifting from document-centered self-service (making them read help articles, showing videos) to an 'activation curve' where conversational AI accompanies them until their first success. No-code Webflow is cited as a leading example of remaking the initial experience into a conversational concierge (public/explanatory article).
The content of the trick is a combination of role-based adaptive learning paths and learning summaries or personal quizzes generated by AI. The biggest key point here is to stop attribute surveys (age, gender, company size). What you leave instead are only questions where 'the next screen actually changes when you answer'. The conversational UI asks only 1 to 3 questions like 'What do you want to do?', and replaces the initial dashboard or template according to the answer. The goal is only one: 'to reach at least one success experience within the first session'.
Surveys that only ask for attributes are just 'work' for the user and do not change the experience by even a millimeter. On the other hand, questions where 'the screen changes when you answer' provide a reward the moment you answer, so they become the driving force for onboarding. This difference may look small, but it affects the number called activation rate.
The backside of ③—Technology x Gritty Marketing: Rotating creative as 'consumables'
The change in thinking here is the transition from 'making one masterpiece' to 'mass-producing consumables that keep rotating'. Conventionally, one good video was made and used to death. However, SNS ads quickly cause 'creative fatigue' (a phenomenon where the same material gets boring and numbers drop). Synthetic UGC counters this wear and tear with 'quantity'. You generate hundreds of videos for the production cost of one, replace the appeal for each stage of the funnel—awareness, comparison, decision—and look for winning patterns through AB testing.
However, I do not intend to praise it unconditionally here. Synthetic UGC has a fundamental weakness called 'authenticity'. It is not the voice of a real user. And the issues of regulation and stealth marketing disclosure will definitely be the next point of contention from now on. Being able to mass-produce and being allowed to use it are different problems.
Today's Featured Products: Those Who Sell the 'Template'
Among the products mentioned today, the ones I found symbolic are 'waitlist support tools' like LaunchList and Waitlister. What they did was take the 'waitlist with visible rankings + referral loop' template—which we dissected in (1)—and turn it into a product that indie developers can use just by 'configuring it.' In other words, they aren't just using a winning pattern themselves; they are lightly commoditizing and selling the winning pattern itself.
This is a direct extension of our 'Time Machine Management' philosophy. When bringing a proven overseas model to the domestic market, you can adopt a two-pronged approach: not only 'using it yourself' but also 'productizing the model for the domestic market.' Today's tools provided a great example of this.
Initial Actions Starting Tomorrow
Finally, so that you—the reader—and I myself can actually take action starting tomorrow, I have laid out specific initial steps for the selling side.
For short-form videos, try using a faceless UGC tool to generate just three patterns of videos with the same appeal, then swap them out by funnel for A/B testing. It is especially important in Japan to remember to include a disclosure that the content is AI-generated.
For onboarding, be bold and eliminate the attribute survey during the first launch, replacing it with a conversational branch of 1–3 questions where 'answering actually changes the next screen.' Then, be sure to measure the percentage of users who reach their 'first success' within the initial session. Adding features without looking at this is like pouring water into a bucket with a hole in it.
For AI utilization, try adding just one lightweight LLM feature to an existing function and rephrase the value proposition to 'AI does X for you.' However, avoid exaggerated claims. The honest approach is to demonstrate the time or man-hours saved in a verifiable way.
And for pre-launch, before your next launch, implement a waitlist + ranking counter + public leaderboard + 'rewards for confirmed referrals only.' I repeat: it is crucial to set this up *before* driving traffic and to give the first 200 people the opportunity to share.
Summary: Stop the Fireworks, Build a System
The idea that struck me most today—'taking launch day off its pedestal'—boils down to the philosophy of 'stopping the one-shot gamble of luck and talent, and running things continuously through design and sequence.' Waitlist ranking visualization, conversational onboarding, and mass-producing synthetic UGC all exist on the same map.
A single firework is beautiful the moment it goes off, but it disappears by the next day. What we want to build is a system that quietly continues to bring people in the next morning and the month after. Tomorrow, I will be back to dissect another new 'system' cooked up by someone overseas. That is all for today; thank you for joining me.
I run a community that studies how to take profitable overseas businesses, adapt them for the Japanese market, and grow them without missing the mark. We cover everything from candidate selection, numerical verification, and Japanese localization to manual testing, MVP, pricing, and scaling. Register for LINE here: https://lin.ee/8Yyw7wk
Note: Numerical values distinguish between public and estimated figures (in particular, '$2K → $50K / Fortune 500 saves $200K' is a claim from the aggregator article and is treated as unverified). Please check individual rules such as the Act against Unjustifiable Premiums and Misleading Representations, stealth marketing regulations, and AI generation disclosures when implementing these measures.
#GrowthHack #IndieDev #SNSMarketing #buildinpublic #TimeMachineManagement #Waitlist #Onboarding #AIMarketing