5 ms·

A Quick and Dirty Analysis of Apply HN

18 points by danieltillett 11y ago

qu1mby 11y ago
Thanks for doing this! Are we able to see which startups fell into which categories?
- danieltillett 11y ago
  Since this is just my subjective analysis based on limited data I don't think it is fair to single out individual applications. The value I hope is in making people think more about their ideas and asking objectively themselves what category their startup falls into.
smikhanov 11y ago
The author should really have provided not only the groups and how many ideas (he thinks) land in that group, but also what are those ideas. Not only this would've make an entertaining read in 2026, but also give more ground for reasonable discussion now.
- danieltillett 11y ago
  I have the list so I will try to remember to comeback and do a postmortem in a few years time. The likelihood is that every single application on the list will fail to become a unicorn and very few will ever make a profit. This is just the way the numbers fall out on early stage startups. The reason I didn't actually list each of the companies in each category is that the purpose was not to critic individual applications, but to look at them as a whole and try to draw lessons from what I saw. There is a real disconnect between what people are proposing and the questions investors want answered. If you want someone's money you need to have good answers for these questions.
david927 11y ago
I'm afraid you weren't objective enough to do this kind of analysis. As one of the Apply HN candidates (Brodlist), the only category I can think you sorted us into is 'existing competition'. While we did have some comments saying that what we have wasn't interesting (some from those competing in the same space), we also had comments hinting that it wasn't even possible. Someone said we're trying to do the Tesla of databases. I'm also subjective, so I see it as the latter. You would need TDD by an objective third-party to know better, or just wait to see what happens after we launch. Second, unicorn embryos don't show themselves that way or everyone would just do it if it's that obvious.
- david927 11y ago
  What would happen if someone had the Tesla for databases? What if it had a way to query data that was also vastly better and simpler? (This wasn't mentioned in passing in the comments on our application since we're not ready to show it, but it exists and is reflection of the language.) If you squint you could imagine that taking off and eating a large portion of the market. Now let's take a random situation: You have a business that makes widgets. You have customers, suppliers, parts, etc. Do you store that information in Word as text? Of course not. You either have a spreadsheet or database, right? Text is great for a novel or a poem... but not for data. That would be crazy. Currently, the web is still a big collection of text documents. How optimal is that? Deeply sub-optimal. The number of people going to the Google homepage is declining because people are now just going directly to the database that represents what they're looking for. Looking to buy a home? You go to Zillow or Trulia and query their database. People want data, not text. 95% of the web wants to be data. Now imagine if that company mentioned before had significant market share. The authentication is common across all of these databases, they all have the same API (the language), and so instead of appearing as many databases, it appears as one large database of everything. Suddenly a network effect kicks in and if you have something you want to be discovered (which is what the web consists of) it makes sense to join in. The next Google won't be better text search but instead something completely different -- it will be data. And the web will become essentially a massive database. I don't know who will get there, no one can know, or how it will happen, but it will happen; it's waiting to happen.
  - p4wnc6 11y ago
    I can't comment on the technical implementation, but just from visiting your website I think you are doing yourself a disservice by having the up-front, immediate form of communication be a negative comment about traditional SQL. Traditional SQL stands up extremely well to modern use cases, and tons of use cases never require the magnitude of data at which other solutions become pragmatic. As a result, it's important to consider traditional RDBMSs with respect, acknowledging that they are extremely resilient to tons of use cases, schema or schema-less situations, distributed access, etc. etc., and active development is making them better and better all the time. By basically saying, "put away your boring old SQL database grandpa, it's time for something better" you're giving sort of an immature and technologically hasty vibe that's going to alienate people. Your technology may be amazing, but people often don't respond to things that present in this "your whole way of thinking is totally wrong" sort of way.
david927 11y ago
AirBnB at this stage (Apply HN is prototype-stage) would have fallen into Network Effect and Lifestyle Business (the size of market for sharing airbeds at conferences is small), and not into Unicorn Embryo.
- p4wnc6 11y ago
  I don't see how AirBnB would ever be classified as Lifestyle Business. The intent was never to make a small amount of money at a fixed scale. It was always about huge scale, which is almost always at odds with a Lifestyle Business. Network Effect yes, Lifestyle Business no. Further, any given investor would be wise to look at businesses like AirBnB with skepticism from a "can they actually achieve critical mass of users" point of view. Effectively 0% of these types of businesses are unicorns. Unless you personally have special knowledge or are willing to take an extreme risk, it's foolish to invest in these ideas and it amounts to literally buying a lottery ticket.
- danieltillett 11y ago
  I would not put the original AirBnB into the lifestyle business category, but the non-profit category. If you look at how many people use AirBnB to share an air mattress in their living room to strangers it was an idea that would never have made money. It was only when they pivoted to their current approach that they became a huge success. Actually AirBnB is a good example of how you can solve the chicken and the egg problem with networks. Because real estate rentals are local, it is possible to bootstrap a strong network in a geographically small location and then scale out. All the successful network businesses out there have used a similar approach (Uber, Facebook, etc).

18 points by danieltillett 11y ago

qu1mby 11y ago

Thanks for doing this! Are we able to see which startups fell into which categories?

danieltillett 11y ago
Since this is just my subjective analysis based on limited data I don't think it is fair to single out individual applications. The value I hope is in making people think more about their ideas and asking objectively themselves what category their startup falls into.

smikhanov 11y ago

The author should really have provided not only the groups and how many ideas (he thinks) land in that group, but also what are those ideas. Not only this would've make an entertaining read in 2026, but also give more ground for reasonable discussion now.

danieltillett 11y ago
I have the list so I will try to remember to comeback and do a postmortem in a few years time. The likelihood is that every single application on the list will fail to become a unicorn and very few will ever make a profit. This is just the way the numbers fall out on early stage startups. The reason I didn't actually list each of the companies in each category is that the purpose was not to critic individual applications, but to look at them as a whole and try to draw lessons from what I saw. There is a real disconnect between what people are proposing and the questions investors want answered. If you want someone's money you need to have good answers for these questions.

david927 11y ago

I'm afraid you weren't objective enough to do this kind of analysis. As one of the Apply HN candidates (Brodlist), the only category I can think you sorted us into is 'existing competition'. While we did have some comments saying that what we have wasn't interesting (some from those competing in the same space), we also had comments hinting that it wasn't even possible. Someone said we're trying to do the Tesla of databases. I'm also subjective, so I see it as the latter. You would need TDD by an objective third-party to know better, or just wait to see what happens after we launch. Second, unicorn embryos don't show themselves that way or everyone would just do it if it's that obvious.

david927 11y ago
What would happen if someone had the Tesla for databases? What if it had a way to query data that was also vastly better and simpler? (This wasn't mentioned in passing in the comments on our application since we're not ready to show it, but it exists and is reflection of the language.) If you squint you could imagine that taking off and eating a large portion of the market. Now let's take a random situation: You have a business that makes widgets. You have customers, suppliers, parts, etc. Do you store that information in Word as text? Of course not. You either have a spreadsheet or database, right? Text is great for a novel or a poem... but not for data. That would be crazy. Currently, the web is still a big collection of text documents. How optimal is that? Deeply sub-optimal. The number of people going to the Google homepage is declining because people are now just going directly to the database that represents what they're looking for. Looking to buy a home? You go to Zillow or Trulia and query their database. People want data, not text. 95% of the web wants to be data. Now imagine if that company mentioned before had significant market share. The authentication is common across all of these databases, they all have the same API (the language), and so instead of appearing as many databases, it appears as one large database of everything. Suddenly a network effect kicks in and if you have something you want to be discovered (which is what the web consists of) it makes sense to join in. The next Google won't be better text search but instead something completely different -- it will be data. And the web will become essentially a massive database. I don't know who will get there, no one can know, or how it will happen, but it will happen; it's waiting to happen.
- p4wnc6 11y ago
  I can't comment on the technical implementation, but just from visiting your website I think you are doing yourself a disservice by having the up-front, immediate form of communication be a negative comment about traditional SQL. Traditional SQL stands up extremely well to modern use cases, and tons of use cases never require the magnitude of data at which other solutions become pragmatic. As a result, it's important to consider traditional RDBMSs with respect, acknowledging that they are extremely resilient to tons of use cases, schema or schema-less situations, distributed access, etc. etc., and active development is making them better and better all the time. By basically saying, "put away your boring old SQL database grandpa, it's time for something better" you're giving sort of an immature and technologically hasty vibe that's going to alienate people. Your technology may be amazing, but people often don't respond to things that present in this "your whole way of thinking is totally wrong" sort of way.

david927 11y ago

AirBnB at this stage (Apply HN is prototype-stage) would have fallen into Network Effect and Lifestyle Business (the size of market for sharing airbeds at conferences is small), and not into Unicorn Embryo.

p4wnc6 11y ago
I don't see how AirBnB would ever be classified as Lifestyle Business. The intent was never to make a small amount of money at a fixed scale. It was always about huge scale, which is almost always at odds with a Lifestyle Business. Network Effect yes, Lifestyle Business no. Further, any given investor would be wise to look at businesses like AirBnB with skepticism from a "can they actually achieve critical mass of users" point of view. Effectively 0% of these types of businesses are unicorns. Unless you personally have special knowledge or are willing to take an extreme risk, it's foolish to invest in these ideas and it amounts to literally buying a lottery ticket.
danieltillett 11y ago
I would not put the original AirBnB into the lifestyle business category, but the non-profit category. If you look at how many people use AirBnB to share an air mattress in their living room to strangers it was an idea that would never have made money. It was only when they pivoted to their current approach that they became a huge success. Actually AirBnB is a good example of how you can solve the chicken and the egg problem with networks. Because real estate rentals are local, it is possible to bootstrap a strong network in a geographically small location and then scale out. All the successful network businesses out there have used a similar approach (Uber, Facebook, etc).