Village Pump: FractalReader’s Operation Diary 2024/4/28
FractalReader’s Operation Diary
Date: 2024/4/28 By: Blu3mo
-
Currently contemplating how much to accelerate from this point.
-
Two major paths are visible:
-
If unable to achieve this, scaling will only result in money being wasted.
-
(If it doesn’t work out, consider closing down)
-
Specifically, once confirmed stable users are present, implement charging features tailored to their usage patterns and attributes.
-
Possible actions:
- Promote in English-speaking communities.
- Self-promotion on LinkedIn.
- Introduce Fractal summaries tailored to various topics in different communities to attract attention.
- For example, introducing a One Piece summary on r/onepiece.
- Try placing it on ProductHunt and similar platforms.
- Consider running advertisements?
- Currently have no clue about it.
- Promote in English-speaking communities.
-
Risks:
- If promotion is successful and traffic increases, money will be spent accordingly.
- Considering $200 as the limit of the risk that can be borne.
- Initially, with only $1000 in the account and no regular income, depletion of the account funds would lead to financial difficulties.
- However, the risk of running out of living expenses can be resolved by borrowing money, so it may not hold much significance in terms of risk-return analysis.
- Considering $200 as the limit of the risk that can be borne.
- Returns:
- Donations through “Buy Me a Coffee” have accumulated around $200.
- Half of it is support from acquaintances, making it difficult to count, while the other half is genuine donations from real users.
- Hence, a certain level of scalability can be expected.
- However, it is uncertain whether this will continue solely based on people who are already familiar with or newly discover Blu3mo.
- (Based on hearing about other cases)
- This interpretation is also considered by Nishio.
- It might be possible to proceed with Blu3mo as a creator’s brand rather than focusing solely on the startup aspect, as suggested by Haruhi and Inajob.
- Building a fan base.
- Blu3mo expresses hesitance about selling themselves beyond the scope of FractalReader.
- Feeling it’s still early for such a step.
- Blu3mo feels not strong enough to maintain themselves amidst external scrutiny, the desire for approval, and financial incentives.
- Starting to sell oneself as a creator may influence one’s actions.
- It seems delicate to let one’s actions in the early twenties be influenced by that.
- Acknowledging the suggestion made by Haruhi.
- However, it is uncertain whether this will continue solely based on people who are already familiar with or newly discover Blu3mo.
- Donations through “Buy Me a Coffee” have accumulated around $200.
- Having a diverse user base would make it easier to find users willing to pay.
- For instance, while a business may not be viable in Japan, it might be in English-speaking regions.
- API costs are universal, but the cost of living in the US is higher, so English content might be more cost-effective.
- English content may also allow handling a larger volume of text for the same cost, as noted by Nishio.
- For instance, while a business may not be viable in Japan, it might be in English-speaking regions.
- While a funded startup would want to quickly iterate on feedback loops, it is a risk that is beyond what an individual like Blu3mo can currently bear.
- If promotion is successful and traffic increases, money will be spent accordingly.
-
-
Considering the implementation of charging features, which historically has been questionable in terms of Visa regulations (refer to the Spring Break Business Development Challenge Records), so it might be wise to wait until returning home in two weeks.
- Monetization could lead to a significant decrease in users, so a certain level of free features is necessary.
- To prevent cheating on free features, consider implementing SMS verification?
- Monetization could lead to a significant decrease in users, so a certain level of free features is necessary.
-
Doing business with online software means competing with people worldwide, so holding back due to reasons like “having only $1000 in hand” or “being questionable in terms of Visa regulations” can be seen as self-handicapping, as pointed out by Nishio.
- Appreciating the objective feedback.- As anyone can assemble and imitate at the same speed, it would be nice to use something that “few people have” instead, and use whatever is available mindset (/villagepump/nishio)
- I see (/villagepump/blu3mo)
- I think that’s true
-
What’s wrong with self-handicap? (/villagepump/blu3mo)
- The issue lies in eliminating options to overcome handicaps in order to create a self-handicap
- It is worth reconsidering if there are ways to overcome the handicap
- “Having only $1000 in hand”
- Options like borrowing money or receiving investments exist
- Even if it’s nice to have $1000 now?
- Even if it works out, I can’t imagine spending more than that
- Well, if 200 users use it every day, it will melt 3000 in a month
- “Visa-related suspicions”
- Is the benefit of advancing actions by two weeks worth the risk of Visa-related issues?
- Honestly, I don’t really understand what the risk is
- Would creating a company in Japan solve it?
- Seems quite cumbersome to do it from abroad
- And even if you do it as a sole proprietor, don’t you need to file for opening a business?
- Due to the vague “Visa-related suspicions” of not understanding what the risks are, it’s like stepping on the brake instead of the accelerator (/villagepump/nishio)
- It’s not a thought process towards accelerating, given the constraints of the visa, to see what can be done at maximum
- For example, even with visa constraints, there are many ways to move forward, like implementing a charging function (without releasing it) and conducting closed beta tests with a few acquaintances to investigate the best charging style
- Is the benefit of advancing actions by two weeks worth the risk of Visa-related issues?
- “Final exams in 5 days”
- Is the benefit of advancing actions by one week worth the risk of a significant drop in GPA?
- Study? (/villagepump/はるひ)(/villagepump/nishio)
- It’s funny how each of us is becoming a self-handicap in our own way (/villagepump/inajob)(/villagepump/nishio)
- “I didn’t know there was an exam in five days” lol (nishio)(nishio)(nishio)
- “Having only $1000 in hand”
Deep down, I think, “There’s no way competitors will appear at such a high speed” (/villagepump/blu3mo)
-
It’s clear that it’s not a service with high operational cost, so in that aspect, it’s somewhat reassuring
- It’s unclear what it means by “somewhat reassuring” compared to what and how
- Being able to think more clearly about these aspects could make one stronger (/villagepump/blu3mo)
- It’s unclear what it means by “somewhat reassuring” compared to what and how
-
Honestly, it might be better to consider the possibility of someone creating a clone (laughs) (/villagepump/nishio)
- It seems unlikely to have a direct confrontation as a service
- Reasons could be wanting to customize the UI freely, wanting to input data that would be troublesome to see there, or wanting to input data of a different format using the structure of that format
- They seem to intend to share their knowledge lol
- lol (/villagepump/blu3mo)
- I felt that saying “It’s not a service with high operational cost” implicitly assumes the current business model (/villagepump/nishio)
- Certainly (/villagepump/blu3mo)
- If someone with genius business model design skills sees it, they might come up with a better model (/villagepump/blu3mo)
- I don’t think the key point is geniusness (/villagepump/nishio)
- By expressing it as “geniuses would think of it,” it gives the impression of drawing a line between oneself and others
- Instead of a division like genius or not, I was using it in a casual way like “Business Model Design Skill Level 100” (/villagepump/blu3mo)- Even though I am only at level 1, I believed that with effort and determination, I could reach my goals (/villagepump/blu3mo).
- I don’t think the key point is geniusness (/villagepump/nishio)
- Mr. N’s company has already acquired customers, each of whom holds undisclosed information on the internet. Therefore, functions that are beneficial for utilizing this “undisclosed information” contribute to the increase in customer value (/villagepump/nishio).
- I see now (/villagepump/blu3mo).
- I often hear about ”non-public information held by companies” in the context of LLM’s fine-tuning, but honestly, I don’t have a very concrete image of it (/villagepump/blu3mo).
- Furthermore, since Mr. N belongs to the research department of that corporate group, there is a personal incentive for the accumulation of insights on its utilization. Also, from the company’s perspective, it would be positive for Mr. N to spend time and budget on these activities (/villagepump/nishio).
- One could easily deduce up to this point from Mr. N’s public information.
- I see now (/villagepump/blu3mo).
-
At the same time, I was considering whether I could sell this mechanism to companies (/villagepump/blu3mo).
- However, it seems that tweeting about it might prevent obtaining a patent.
- I didn’t know that.
- Too late~
- Is there something called grace period?
- Since I haven’t tweeted about the internal structure, it’s irrelevant. The issue is more with Scrapbox. Whether obtaining a patent is beneficial is a separate matter. Patent costs are not cheap, and having a patent doesn’t necessarily attract buyers (/villagepump/nishio).
- I see (blu3mo)(blu3mo).
- Therefore, it seems challenging to exclusively sell the mechanism.
- It’s something I could create in a day, so there’s no need to buy it. Companies can simply imitate the idea/UI and create it in-house.
- I don’t really understand the benefits of obtaining a patent in the first place (/villagepump/blu3mo).
- I occasionally hear that it’s difficult to protect ideas with patents in software.
- https://www.slideshare.net/slideshow/it-8465452/8465452
- I wonder how much novelty can be claimed in interface design.
- Expensive
- I thought it would cost at most 20-30,000.
- If you hire a professional, it could cost up to 300,000 including success fees.
- Probably if it’s just a patent in Japan, it would be around 1 million for an international patent, right? (/villagepump/nishio).
- Cost of obtaining a patent
- I see~~ (/villagepump/blu3mo).
- As an intern at a certain newspaper’s research institute, I am encouraged to continue development and obtain a patent as a job invention (/villagepump/blu3mo).
- What is job invention? (/villagepump/blu3mo)
- By default, inventions produced as part of one’s job belong to the individual, and through a transaction between the individual and the company, it can be transferred to the company (unless explicitly stated in the employment contract).
- That’s how I understand it (/villagepump/blu3mo).
- For now, I am more interested in turning fractal summary into a personal development service and venture into business from that angle (/villagepump/blu3mo).
- While the invention belongs to the individual, I imagined a scenario where both the company and the inventor could create competing services using it without any objections (/villagepump/inajob).
- In this scenario, wouldn’t it be beneficial to create a personal service?
- I’m curious about whether the person suggesting this has the authority to make decisions on “intern hiring” or “obtaining a patent as a job invention” (/villagepump/nishio).
- If someone with decision-making authority is saying, “Why don’t you come work for us?” then it indicates a need.
- If it’s someone without authority saying it, it’s probably just a casual suggestion.- It’s over here (/villagepump/blu3mo).
- However, it seems that tweeting about it might prevent obtaining a patent.
I’m curious about how experts in business view this. (Is it said to be impossible to make it work because the cost is too high?) @(/villagepump/blu3mo) April 28, 2024
- They are also doing self-handicapping here (/villagepump/blu3mo).
-
If you can’t input large texts without paying, or if you can only generate summaries up to three levels, wouldn’t that be a good approach?
-
@(/villagepump/shokai) [April 28, 2024](https://twitter.com/shokai/status/1784587696654143509?ref_src=twsrc%5Etfw)- As an example, DeepL’s free version has a system where you can’t input large texts without paying.
(/villagepump/blu3mo)
- I’d like to mimic this display (/villagepump/blu3mo).
- As an example, DeepL’s free version has a system where you can’t input large texts without paying.
-
Personally, I might just go for a fully paid subscription (with a 7-day free trial) because it seems like it could sell. For cases where users input keys like in Cursor Editor, it could be free, but for others, a subscription might work. Also, assuming users input keys themselves, providing it as a one-time purchase tool could be an option.
-
@(/villagepump/nztm_tw) [April 28, 2024](https://twitter.com/nztm_tw/status/1784585418891227239?ref_src=twsrc%5Etfw)- If it’s a few hundred or a few thousand yen for a one-time purchase, I’d buy it immediately (/villagepump/はるひ)(/villagepump/nishio).
- But it seems like API costs would be a separate burden (/villagepump/who).
- Well, with a one-time purchase, the incentive to improve the service might be lost (/villagepump/nishio).
- It seems like a decision whether to target users with the literacy to input keys or not could be crucial (/villagepump/sta).
- Obtaining an API key like OpenAI (or was it claude?) isn’t something an average person can do.
- I can’t think of a specific reason to focus on one group (/villagepump/blu3mo).
- Regarding API key input, if done, could it be offered as a side option? But it seems like a hassle to implement and there might be few users who would appreciate it, so it could be postponed (/villagepump/blu3mo).
- It seems good for those who want to pay based on usage, but is a metered plan sufficient…
- Regarding API key input, if done, could it be offered as a side option? But it seems like a hassle to implement and there might be few users who would appreciate it, so it could be postponed (/villagepump/blu3mo).
- If it’s a few hundred or a few thousand yen for a one-time purchase, I’d buy it immediately (/villagepump/はるひ)(/villagepump/nishio).
- It was helpful to hear the sentiments of people currently working as product owners.
- I took it as a message like “It may not be impossible (though I don’t bear responsibility).”
- It’s encouraging.
- Should we also have a metered plan for users who don’t use it much? (/villagepump/who)
- It could attract people back with a free tier.
By the way, is Japan entering the Golden Week period now? (/villagepump/blu3mo)
April 28, 2024
-
The “two big paths” I mentioned above were not accurately perceived (/villagepump/blu3mo). - Oh, it’s the ”False dilemma” (I named it, but it’s hard to find a good concrete example haha) (/villagepump/nishio).
- Planning to further promote and expand in English-speaking regions.
- The point of contention is when to do it.
- 1: Do it now
- In this case, if we expand in the current free-for-all state, money will just drain away.
- Based on past examples, donations are not something to count on much.
- It’s just one example, though.
- 2: Implement the billing system first and then do it
- In this case, it will either be 1) no one pays and no one uses it, or 2) people pay and use it, so there is no risk of money draining away.
- 3: Implement the billing system, validate that the business can thrive in the Japanese market, and then do it.
- This would provide further reassurance.
-
Implementing a billing system is surprisingly a lot of work.
- Research and clear legal requirements.
- Implement Google account login.
- Integrate Stripe.
- Decide on the monetization model.
- Ideally, want to be in Japan for Visa reasons (5/11~).- On the other hand, it might be surprisingly easy to take a step forward in verifying that business can thrive in the Japanese-speaking community.
- Set up a button labeled “Unlimited Plan” to guide users to a temporary Waitlist form.
- Content of the Waitlist form:
- Just an email address for now?
- How about adding market research-like questions such as “How much would you be willing to pay?”
- Maybe not. Since the sample size is not that large, it would be troublesome if people find it bothersome and withdraw.
-
There seems to be no significant risk in spending the next week focusing on studying for exams and leaving FractalReader idle.
- Concerned about someone else creating the same thing and starting to sell it?
- In that case, whether there is a lead in a week or not, it doesn’t seem like we would win in a business competition based on business strength, so it doesn’t seem like a problem.
- People who are willing to copy a service that was lightly buzzing on Twitter and release it within a week are likely entering the business game with the mindset of competing (/villagepump/blu3mo).
- I don’t want to compete with such a person.
- Concerned about someone else creating the same thing and starting to sell it?
-
The plan for the next week might look something like this:
- Promotion
- Regularly tweet on Japanese-speaking Twitter to remind users
- Have it used in an English Plurality Book reading group
- Not doing anything else
- Implementation
- Not implementing a billing system
- If time is spent on implementation, it could jeopardize GPA and affect graduate school applications
- It is possible to spend 1-2 hours implementing the Waitlist form, so we will do it.
- Not implementing a billing system
- Other
- Take care of things like Microsoft for Startup API Credit, Claude’s upper limit upgrade, which take a few days for approval.
- Measure how much people continue to use it and how many are interested in billing.
- Is it the end of the turn now TurnEnd?
- The way of thinking is still fun like a management game like Coffee Inc 2 (/villagepump/blu3mo).
- Promotion
-
Added a Pro PlanButton (/villagepump/blu3mo).
- Guide users to the Waitlist form for now.
- Is it bad to pretend to have a Pro Plan when it’s not actually ready yet?
- I think it’s a common MVP approach (/villagepump/nishio).
- Since the link destination of the button is a strange shortened link, it would be less suspicious to aggregate it through an internal redirector (/villagepump/bsahd).
- (If the connection is slow, the address during redirection will be displayed)
- I can’t spend too much time on implementation, so I’ll do it later if I feel like it (/villagepump/blu3mo).
-
Job of replying to people who reacted on Twitter
- Throw a reply after about a day to remind them
- It seems that there are cases where a reply leads to another action (/villagepump/blu3mo).
- That’s great (/villagepump/blu3mo)
-
I will also introduce it to my friends!!
-
@(/villagepump/shitsudo_K) April 29, 2024
-
Thank you. I tried it right away! I haven’t tried it in actual cases, so I can’t give detailed feedback, but it seems useful for finding related information from long descriptions in patent documents. https://t.co/wVJYCFjEwB
-
@(/villagepump/PatAndMarketing) April 29, 2024
