4 ms·

I have a lot of thoughts on this, so please bear with me. I think it's interesting that the model we use to evaluate ideas and fund companies is so terrible.

I have a lot of thoughts on this, so please bear with me. I think it's interesting that the model we use to evaluate ideas and fund companies is so terrible. Starting a company is always dangerous and risky and that's extremely sad. Why is starting a business so hard in 2023? I think we need an "explicit demand signal" and I think it can be technological one and a semantic one. Kickstarter is the nearest to this I can think of but kickstarter is only a primitive marketplace - for example, I don't see services listed. But I want to see vector semantic models used to capture the meaning of business processes and identify weaknesses and match problems to imaginary products and synthesise products. Think automatic market understanding and differentiation. It would be indexed per organisation and across society. There's a carrying capacity for every market. An average city can only support a certain number of coffee shops or X of kind of venue. So we have competition. A lot of people grinding their gears competing needlessly. It's such a waste. One of my ideas was "attention camping", which is waiting for something to match your idea of something you find interesting or want to pay attention to. If I had a commitment to buy from a large number of people, then my business would be a lot safer, except I'm an unknown quantity.

JohnFen 3y ago
> Why is starting a business so hard in 2023? Honestly, I don't think it's any harder now than it has been before (in my adult life, anyway). It's always been very hard. Same with running a business post-startup.
orzig 3y ago
I'm definitely open to more models for funding, but it didn't seem like Kickstarter really worked even for what should have been its sweet spot. Definitely tell me if I'm wrong, but the fact that they had to put out this blog post suggested that the dynamics of its usage never really fit: https://www.kickstarter.com/blog/kickstarter-is-not-a-store https://www.kickstarter.com/blog/kickstarter-is-not-a-store
- JohnFen 3y ago
  Crowdfunding is a terrible way of financing a startup. It can be a reasonable way of financing a product, if done carefully and thoughtfully.
  - imtringued 3y ago
    Crowdinvesting is a form of Crowdfunding in Germany that is very successful. The only problem is that the ECSP license is onerous enough that nobody bothers to apply for it. This means that the crowdinvesting platforms merely act as sales people for conditionally repayable loans. These loans are considered equity and therefore pay higher interest like 8% which in turn makes this form of financing less attractive.
nradov 3y ago
It is not possible to systemize the founding and funding of startups based on disruptive innovations. This can only work from human insight and there will never be explicit demand signals for innovations that don't exist yet. And there's nothing sad about it. It's totally fine for some companies to fail and investors to lose money.
- gunapologist99 3y ago
  People want to remove the 'venture' from 'venture capital'.
  - FloorEgg 3y ago
    Maybe if building a startup wasn't such a risky endeavor we wouldn't need venture capital and a lot more people could do work they feel is meaningful and are passionate about.
    - wnc3141 3y ago
      One interesting implication of AI is the potential decapitalizing of business. I.e, businesses will become more modular and not require things like as much dedicated admin to scale. This will not increase the likelihood of product market fit, (as many point out, that is impossible to model given supply constraints and inability to understand demand side revealed preference without some MVP). - however this may decreased the amount of capital needing to be placed at risk, allow fast failing etc.
    - satvikpendem 3y ago
      Lots of people bootstrap just fine to some tens of thousands of dollars with no VC, ie those on https://IndieHackers.com https://IndieHackers.com. It's not really risky if you don't quit your job until you're making enough money to quit. Then if you want to grow faster, you can take VC, which you're much more likely to be offered and receive due to having derisked the business both for yourself and also the VC, since you're making money. This isn't strictly necessary either since you can grow more slowly and still become big, like Mailchimp did for over 20 years before selling for $12 billion.
FloorEgg 3y ago
I am working on this problem exactly. The way I describe it is that I am making unmet market needs transparent. It involves modelling jobs to be done, importance and satisfaction at scale and then measuring areas that represent bumps in the market. I think to appreciate this being possible one must understand that people are capable of describing their own wants, even if they can't imagine a novel solution that could satisfy them. samsquire, I see you come up with a lot of ideas. How do you decide which ones to work on? Do you also enjoy building/executing? How important is this problem/idea to you?
- samsquire 3y ago
  Thank you for your comment. I was involved in a Business Process Modelling research and I am sure the SAP world models business processes pretty intricately. I like your idea of modelling the "work" that needs to be done. I think it's partly an allocation problem, farming has automated food production that we have a lot of people that need to cover their costs. Recruitment is expensive and tedious and there's a lot of unsolved problems there. Internet advertisements guarantees attention to something for a cost, I am sure businesses would like to guarantee capital and runway. I think your use of the word "wants" is the key word here. If pushed to it, people could probably try work out what they want and someone else could try fulfill what they truly want. Software requirements are not always great, but there is a want behind every product or marketplace. Allocating resources to ideas seems to be something that could transform the world. Identifying the good, profitable ideas is the hard part. One of my ideas is a "Want app" which is that I can insert a demand for a doctor in my local area, or want for a fast food restaurant in this area, or I want a job that uses X, Y, Z technologies. One of my ideas is "wantsfiles" [1] which is a machine readable file you put at the root of your domain and and then people automatically match orders with you based on what you want, as a form of automated business. For example, you could put an order on your wantsfile that you will buy "milk" at this price or you want someone to host "mirror.tgz" for you and you'll give them $X a month for it. There is also the idea of a generic purpose "wants app" where you specify what you want. [2] I wrote about that idea here [3]. It's a machine automatable order matching system. This problem is very important to me. I spend my time on programming language implementation technical projects but I am also thinking of an idea that shall make the world better for everyone and sharing it openly. If you want to talk about ideas, I'll listen. [1]: https://www.halfbakery.com/idea/Wantsfiles#1598878006 https://www.halfbakery.com/idea/Wantsfiles#1598878006 [2]: https://www.halfbakery.com/idea/Wants_20app#1578226067 https://www.halfbakery.com/idea/Wants_20app#1578226067 [3]: https://github.com/samsquire/ideas2#5-open-demand-mapping-and-want-marketplaces https://github.com/samsquire/ideas2#5-open-demand-mapping-an...
SkipperCat 3y ago
How would you stop people from gaming the system? At some point, don't you need an actual human trained in business/markets to make a decision. I wouldn't want capital allocated by something like the YouTube algo or the Facebook feed algo.
- sitkack 3y ago
  The YT algorithm already controls thousands of peoples livelihoods. The shooting at the YT campus was motivated by the shooters loss of income.
paddw 3y ago
I mean specialized venture funds exist, which is somewhat close to what you are describing. I think the way you understand startup formation is wrong though. Most startups are not formed as ways to directly address existing desires or pain points. Most startups are, fundamentally, iterations on existing business models trying to unseat their predecessors through more subtle innovations, riding technology trends, or simply a large war-chest of capital and talent. This isn't a completely accurate view either. You could cite things like biotech which have their own weird dynamics when it comes to startups. Fundamentally, however, I don't think the "market for ideas to solve problems" model works when thinking about startups. Many people have made some variant of this point, and probably argued for other things being the critical factor (e.g "investing in a founder is the most import important thing"). Ultimately, I don't believe there is one single way to think about how to fund companies that will be successful startups. Mostly, I think investors rely on intuition, which is often wrong, but perhaps better than anything else.
imtringued 3y ago
Explicit demand signal? The only way you can do that today is through gift cards. In principle the company could issue gift cards at a discount and if they were tradeable, then "investors" into the gift cards would hold onto them until the gift cards are worth more than they originally paid for. There are also Genussrechte in Germany which are commonly used to pay interest not as money but as gift cards which can only be used at the company you invested in. This is done to fund local bakeries where the future customers essentially pay upfront for their consumption in exchange for receiving a return.

JohnFen 3y ago

> Why is starting a business so hard in 2023? Honestly, I don't think it's any harder now than it has been before (in my adult life, anyway). It's always been very hard. Same with running a business post-startup.

orzig 3y ago

I'm definitely open to more models for funding, but it didn't seem like Kickstarter really worked even for what should have been its sweet spot. Definitely tell me if I'm wrong, but the fact that they had to put out this blog post suggested that the dynamics of its usage never really fit: https://www.kickstarter.com/blog/kickstarter-is-not-a-store https://www.kickstarter.com/blog/kickstarter-is-not-a-store

JohnFen 3y ago
Crowdfunding is a terrible way of financing a startup. It can be a reasonable way of financing a product, if done carefully and thoughtfully.
- imtringued 3y ago
  Crowdinvesting is a form of Crowdfunding in Germany that is very successful. The only problem is that the ECSP license is onerous enough that nobody bothers to apply for it. This means that the crowdinvesting platforms merely act as sales people for conditionally repayable loans. These loans are considered equity and therefore pay higher interest like 8% which in turn makes this form of financing less attractive.

nradov 3y ago

It is not possible to systemize the founding and funding of startups based on disruptive innovations. This can only work from human insight and there will never be explicit demand signals for innovations that don't exist yet. And there's nothing sad about it. It's totally fine for some companies to fail and investors to lose money.

gunapologist99 3y ago
People want to remove the 'venture' from 'venture capital'.
- FloorEgg 3y ago
  Maybe if building a startup wasn't such a risky endeavor we wouldn't need venture capital and a lot more people could do work they feel is meaningful and are passionate about.
  - wnc3141 3y ago
    One interesting implication of AI is the potential decapitalizing of business. I.e, businesses will become more modular and not require things like as much dedicated admin to scale. This will not increase the likelihood of product market fit, (as many point out, that is impossible to model given supply constraints and inability to understand demand side revealed preference without some MVP). - however this may decreased the amount of capital needing to be placed at risk, allow fast failing etc.
  - satvikpendem 3y ago
    Lots of people bootstrap just fine to some tens of thousands of dollars with no VC, ie those on https://IndieHackers.com https://IndieHackers.com. It's not really risky if you don't quit your job until you're making enough money to quit. Then if you want to grow faster, you can take VC, which you're much more likely to be offered and receive due to having derisked the business both for yourself and also the VC, since you're making money. This isn't strictly necessary either since you can grow more slowly and still become big, like Mailchimp did for over 20 years before selling for $12 billion.

FloorEgg 3y ago

I am working on this problem exactly. The way I describe it is that I am making unmet market needs transparent. It involves modelling jobs to be done, importance and satisfaction at scale and then measuring areas that represent bumps in the market. I think to appreciate this being possible one must understand that people are capable of describing their own wants, even if they can't imagine a novel solution that could satisfy them. samsquire, I see you come up with a lot of ideas. How do you decide which ones to work on? Do you also enjoy building/executing? How important is this problem/idea to you?

samsquire 3y ago
Thank you for your comment. I was involved in a Business Process Modelling research and I am sure the SAP world models business processes pretty intricately. I like your idea of modelling the "work" that needs to be done. I think it's partly an allocation problem, farming has automated food production that we have a lot of people that need to cover their costs. Recruitment is expensive and tedious and there's a lot of unsolved problems there. Internet advertisements guarantees attention to something for a cost, I am sure businesses would like to guarantee capital and runway. I think your use of the word "wants" is the key word here. If pushed to it, people could probably try work out what they want and someone else could try fulfill what they truly want. Software requirements are not always great, but there is a want behind every product or marketplace. Allocating resources to ideas seems to be something that could transform the world. Identifying the good, profitable ideas is the hard part. One of my ideas is a "Want app" which is that I can insert a demand for a doctor in my local area, or want for a fast food restaurant in this area, or I want a job that uses X, Y, Z technologies. One of my ideas is "wantsfiles" [1] which is a machine readable file you put at the root of your domain and and then people automatically match orders with you based on what you want, as a form of automated business. For example, you could put an order on your wantsfile that you will buy "milk" at this price or you want someone to host "mirror.tgz" for you and you'll give them $X a month for it. There is also the idea of a generic purpose "wants app" where you specify what you want. [2] I wrote about that idea here [3]. It's a machine automatable order matching system. This problem is very important to me. I spend my time on programming language implementation technical projects but I am also thinking of an idea that shall make the world better for everyone and sharing it openly. If you want to talk about ideas, I'll listen. [1]: https://www.halfbakery.com/idea/Wantsfiles#1598878006 https://www.halfbakery.com/idea/Wantsfiles#1598878006 [2]: https://www.halfbakery.com/idea/Wants_20app#1578226067 https://www.halfbakery.com/idea/Wants_20app#1578226067 [3]: https://github.com/samsquire/ideas2#5-open-demand-mapping-and-want-marketplaces https://github.com/samsquire/ideas2#5-open-demand-mapping-an...

SkipperCat 3y ago

How would you stop people from gaming the system? At some point, don't you need an actual human trained in business/markets to make a decision. I wouldn't want capital allocated by something like the YouTube algo or the Facebook feed algo.

sitkack 3y ago
The YT algorithm already controls thousands of peoples livelihoods. The shooting at the YT campus was motivated by the shooters loss of income.

paddw 3y ago

I mean specialized venture funds exist, which is somewhat close to what you are describing. I think the way you understand startup formation is wrong though. Most startups are not formed as ways to directly address existing desires or pain points. Most startups are, fundamentally, iterations on existing business models trying to unseat their predecessors through more subtle innovations, riding technology trends, or simply a large war-chest of capital and talent. This isn't a completely accurate view either. You could cite things like biotech which have their own weird dynamics when it comes to startups. Fundamentally, however, I don't think the "market for ideas to solve problems" model works when thinking about startups. Many people have made some variant of this point, and probably argued for other things being the critical factor (e.g "investing in a founder is the most import important thing"). Ultimately, I don't believe there is one single way to think about how to fund companies that will be successful startups. Mostly, I think investors rely on intuition, which is often wrong, but perhaps better than anything else.

imtringued 3y ago

Explicit demand signal? The only way you can do that today is through gift cards. In principle the company could issue gift cards at a discount and if they were tradeable, then "investors" into the gift cards would hold onto them until the gift cards are worth more than they originally paid for. There are also Genussrechte in Germany which are commonly used to pay interest not as money but as gift cards which can only be used at the company you invested in. This is done to fund local bakeries where the future customers essentially pay upfront for their consumption in exchange for receiving a return.