20,000 Startup Ideas
unawaz.github.io
unawaz.github.io
* Check with customers to determine their usefulness.
* Inoculate animals against various diseases, such as internships.
* Review product promotional materials, and physical fitness.
* Monitor computer-controlled test equipment, using germicides and sterilizers.
* Transpose music to alternate keys, or to towers.
* Operate farm machinery, such as espresso machines.
* Document software defects, using hand and horn signals.
* Prepare bank deposits, or performing arts.
* Run errands, such as animal dehorning or castration.
* Forecast staff, equipment, utensils, and silverware.
I hope this is the next agile
Nice, I always say that customers and testers are source of all bugs!
Seems like this one is destined to fail due to lack of focus.
I generated these with https://github.com/pteichman/cobe - anyone got other markov chain generators to play with? This one seems pretty basic.
> Forecast staff, equipment, utensils, and silverware.
Sounds like something a catering business would need.
> Run errands, such as animal dehorning or castration.
> Operate farm machinery, such as espresso machines.
Both of these sound like things a ranch would need (who says a cattle rancher wouldn't appreciate an espresso while turning a bull into a steer?).
Consider replacing "silverware" with "climate" or "markets" in your deck.
1 - https://news.artnet.com/art-world/taiwan-kinmen-island-sonic...
To make this list more useful, there should be two additional datapoints: 1. How many people does the activity impact 2. How much does the activity impact each person. The product of these would be the total impact. Sorting by the amount of impact would give the most valuable areas for innovation, and there should already be a plenty of businesses focused on these activities.
The idea is when an egg is fertilized. Producing a baby is a lot more work, and producing a successful adult is a lot more beyond that.
The problem is that people confuse having an idea with having a good idea. The latter will be something that has things like a justifiable business model, a business plan (seriously, things like path to market and understanding the consumer market are important), something unique that can hook users etc.
That's true. I see a few things behind the monitor right now.
1) Trees. Let's do something with them, like planting them at scale, maybe mess with their DNA, WCGR?
2) The window has birdshit on it. Let's build window cleaning drones.
3) Ceiling lights. These are really crap make a really smart, well designed one for office environments.
Which is strange to me, since when I hear "drone" the first animal that comes to mind is an ant. I thought it'd just be a catch-all term for any kind of bot.
I can't recall now where, but heard the word drone used for robots. Maybe the boundary is getting blurrier?
What is important to understand is that both types of ideas are useless to the layperson. The former is useless obviously because it hasn't been validated. The latter is also useless however because even if the layperson learns it, they won't have the knowledge and expertise to understand its value, much less know how to execute it.
This is why when someone comes to you and says "I have an idea, but you need to sign an NDA first before I can share it", you should just smile, nod and politely change the subject.
The simple core ideas behind Dropbox and AirBnB were seemingly disregarded by many devs and investors when first pitched, but have proven themselves to be valuable. Even Google wasn't initially considered to be a good idea where Yahoo rejected acquiring them for $1M, but turns out Search Engines displaying contextual relevant ads has become one of the most valuable businesses in history.
Having valuable, proven insights that nobody knows, and are possible to extract value from , that's harder.
but the really hard part is: "and are possible to extract value from"
Why is it so difficult to move us forward?
Where has this list been all my life? I could have been a zillionaire....
Clean, shape, and polish fingernails and toenails, using files and nail polish.
Remove bones, and cut meat into standard cuts in preparation for marketing.
How about: build an augmented reality display for customers who are waiting to get their nails done. It's got a camera, and it will show users what their nails would look like with any of the styles offered by the salon applied.
To this day, I've neither seen nor built such a device. Devices to print patterns on your finger nails do exist, however.
Maybe less actually :)
there aren't even 20k ideas...
"Maybe less" indeed ;-)
If you give the start of this list to transformer
https://talktotransformer.com/
the generated ones look as good as those, example:
5.Evaluate a vehicle's operation and use of the system and record information on the system, such as the number of miles driven, total mileage, speed, and odometer reading. 6.Prepare and administer reports to provide data on vehicle performance, and to support vehicle performance maintenance and repair.
The idea is that if a bot can generate patent applications they aren't very novel.
From this we should backtrack and look at existing patents and throw a large percentage of them out : )
Or do I still not understand patents?
Could this list exist at a stop-patents type website where even the patent office could do a quick search to see if an idea is unique and original, or has been thought of easily already and put on big list?
Okay.
> 11031. Transport the deceased to the funeral home.
I think we've found another use for robots.
> Apply and release hand brakes.
I have started referring to it as, "Our Cubby" and my wife hates it.
*in a very specific way https://www.cnn.com/2019/05/22/us/washington-human-compostin...
Some are oddly specific
Very specific, and somewhat disturbing.
I've thought that inserting IVs is something that nurses struggle with a lot and maybe could be automated. It would be an absolute nightmare to develop that robot though, as you'd have to go through all of the medical device certification on top of dealing with all of the edge cases of human anatomy.
7024. Collect payment upon job completion.
> 1600. Collect payments from customers.
I've cracked the code.
Was this an ML project? This looks like the output of a very basic machine learning project.
The commentary here is the best part of this. As (un)valuable as this list may be, it definitely gets the wheels turning in your mind.
Seems to check out
Sounds profitable.
> Nebula Brain - Virtual Offices like Regus Offices
> Galaxy Brain - Coworking spaces
> Exploding Galaxy Brain - WeWork
> Super Exploding Galaxy Brain - OurWork a per second billed coworking/living space
More like Amazon Workplace Services, if Amazon does it.
Also:
> 20. Carry messages or documents between departments.
"I have people skills! I am good at dealing with people! Can't you understand that? What the hell is wrong with you people?"
#11755: Separate scrap waste and related materials for reuse, recycling, or disposal.
#11784: Separate scrap or waste materials for recycling, reuse, or environmentally sound disposal.
#11997: Identify or separate waste products or materials for recycling or reuse.
2948. Deliver babies.Succinctly named "Apparatus for facilitating the birth of a child by centrifugal force"
Just not sure if it's just a satirical thing -- like, "Hah, look at the silly word jumbles my script generated!" -- or if it's meant as a tool for something.
Literally scale.ai
>51. Conduct raids and order detention of witnesses and suspects for questioning.
But somehow that didn't do anything for me.
Some even would offer to connect me to people with the problem, but I had the feeling even with people who would say "I have problem X and I'd pay for a solution" the whole process was still too abstract for me.
I think this has more to do with the business side of things than with talking to potential customers and implementing a solution.
How do I start? What are good first steps if you don't know anyone, don't have a network etc.
It’s not necessarily quick but it works. Amy Hoy & Alex Hillman were the ones who taught me this strategy and I heartily recommend reading everything they’ve written at https://stackingthebricks.com :)
The best link is: https://nugget.one/ideas
We have 4,000+ useful ideas manually sourced from people with problems to solve.
Yeah, well, not sure about this one.
> 874. Carve meat.
> 5221. Dig trenches.
> 8361. Graft plants.
> 10799. Align wheels.
> 16441. Shampoo hair.
And the five longest ones (291, 305, 308, 317 and 332 characters long):
> 6231. Keep abreast of government regulations and emerging Web technology to ensure regulatory compliance by reviewing current literature, talking with colleagues, participating in educational programs, attending meetings or workshops, or participating in professional organizations or conferences.
> 12863. Design, develop, select, test, implement, and evaluate new or modified informatics solutions, data structures, and decision-support mechanisms to support patients, health care professionals, and their information management and human-computer and human-technology interactions within health care contexts.
> 7530. Keep abreast of game design technology and techniques, industry trends, or audience interests, reactions, and needs by reviewing current literature, talking with colleagues, participating in educational programs, attending meetings or workshops, or participating in professional organizations or conferences.
> 4806. Direct environmental programs, such as air or water compliance, aboveground or underground storage tanks, spill prevention or control, hazardous waste or materials management, solid waste recycling, medical waste management, indoor air quality, integrated pest management, employee training, or disaster preparedness.
> 5895. Compute, retrace, or adjust existing surveys of features such as highway alignments, property boundaries, utilities, control and other surveys to match the ground elevation-dependent grids, geodetic grids, or property boundaries and to ensure accuracy and continuity of data used in engineering, surveying, or construction projects.
---
My method for determining these things (which may be of interest to some), using just my favourite text editor, rather than slurping it into a REPL for some programming language and manipulating it thus:
1. Copy the whole document to the clipboard;
2. Paste in Vim (and note at this point that lists lose their markers in favour of just a tab character, so the numbers disappear);
3. Manually remove everything that’s not the list;
4. Manually remove the tab at the start of each line with `:%s/^\t//` or block selection (this could also be done in the next step with slight modification, matching `^\t\(.\+)` and using `submatch(1)`);
5. Prefix each line with its length and the line number using `:%s/.\+/\=len(submatch(0)) . "\t" . line(".") . ". " . submatch(0)` (I actually used asterisk instead of \+, but HN formatting hates unmatched asterisks; see `:help sub-replace-\=`; another approach for the line number part alone would have been to block-insert `1.\t` on every line, then block-select all but the first, and use `g<CTRL-A>` to increment them, see `:help v_g_CTRL-A`);
6. Sort by line length with `:sort n` (since I had put it at the start of the line for convenience);
7. Optionally strip the character counts out again with `:%s/^\d\+\t//` or similar.
sub-replace-\= and v_g_CTRL-A are among my favourite not-so-well-known Vim features. They’re not useful very often, but when they are, they’re really great.
My solution:
curl "https://unawaz.github.io/stochastic-hill-climbing/tasks/" \
| awk '{ print length, $0 }' \
| sort -n -s \
| sed 's/<li>//' \
| (tail -n +11 | head; tail)
curl [address] > /tmp/foo
cat /tmp/foo |
1. copy the text 2. split by \n 3. sort by length 4. you got sorted list
^^ I built this more or less on a bet about a year ago. Give it a dollar and it will give you a startup business idea.
(No Markov chains here, it's a couple of arrays of industries and keywords and a JS function that uses Math.random()... )
> Adjust felt hammers on pianos to increase tonal mellowness or brilliance, using sanding paddles, lacquer, or needles.
> Review filed tax returns to determine whether claimed tax credits and deductions are allowed by law.
Three adjacent ones there, could be rewritten as:
* become a cobbler
* become a piano technician
* become an accountant
In general the IT-oriented items seem like variations of "fix my goofed up spreadsheet". Did they scrape the tech support logs?
I wasn't able to find cloudflare in there
Validation is the first step to product market fit.
(function () {
var els = document.getElementsByTagName("li");
var startups = ["A web-based platform", "An AI-powered platform", "Uber for",
"Subscription-based service", "Blockchain", "A folder that syncs"]
for(var i = 4 ; i < els.length ; ++i) {
els[i].innerText = startups[Math.floor(Math.random() * startups.length)] + " for people who " + els[i].innerText;
}})()
The code is GPLv3, and can be trivially modified to generate further interesting startup ideas.
EDIT: Modified to add a greater variety of instant startup ideas!
Watch out guys we've got a unicorn over here
Fair enough
Genius!
What is rare is the hard work needed to make an idea a reality.
When other people say "hey I'd love an app that let me do X, you should build it, you'd make so much money" it usually turns out there are already 20 in the app store, and they "love" the idea so much that they'd never bothered to look for one. This is true whether it's for themselves, or for their job.
No...no no don't do that you will stand a good chance of being electrocuted.
Watch out you guys we've got a unicorn
what?
i have a backend tech stack that i believe is scalable and efficient (REST or websockets). i'd like to prove this out with a demo or app. i can do a pure-js SPA frontend (eg fitno.us) but i'm terrible at UI so it's not gonna be pretty
is there a free service that i could provide that people would use at high scale ? there doesn't need to be a way to eventually monetize it
required: store data to database optional: rest calls to 3rd parties
anyone have an idea they're willing to share ?
Instead of just a 2-way chat program, there is a 3rd participant "the FBI guy" who also joins the chat. Every message sent must get approved by the FBI guy before it goes through to the recipient.
It should also be possible to send messages to the FBI guy directly. e.g. "Does she/he like me?"
what's the motivation for a user to use this app vs un-moderated chat ? is it driven by the moderator, eg a parent wanting to limit what their kids can say or see ? or is it one of the 2 chatting parties that prefers to have someone else sanitize content before they have to see it, eg to ensure that they don't get d*ck pics ?
in the latter case, what motivates the moderator to participate ?
Although there probably is a way to pitch it as a parental control, I think other demographics could enjoy it too. As you say, the moderation basically ensures no creepers or d*ck picks. That will probably encourage more female users to trust the platform, which would certainly help to boost overall numbers.
It could be a means of chatting to strangers, where the moderator chooses 2 people to pair up. That could get interesting. The idea is still pretty half-baked, but I'd like to explore it more, and maybe use it as a testing ground for decentralised chat technology.
For people who are unsure about social norms of chatting to people, such as those with Aspergers, autism spectrum, or who are just plain shy, I think being a moderator would be a great way to learn how to have an ordinary conversation. I'd also want to make the website fully accessible, with text readers for the blind, and automatic language translation so that people can make friends in other countries.
this sounds fun :)
I guess some money could be made by using the chat logs to train AI chat bots, and selling those bots. I have no idea how that would work in practice though.
* Accept uploads.
* Push them to S3 / similar.
* Marvel as you have to filter out nasty images.
* Struggling under the load of hotlinked images.
* Start performing rewrites to serve ads.
* Add chat/comments to try to build a community.
* Get more and more slow and ad-heavy over time.
The process repeats every few years.