by trevett on 8/5/20, 11:48 PM with 222 comments
1. Start with listing markets that have a low degree of competition, but don't have a mega-monopoly owning them. These will mostly be small markets.
2. Examine the problem space within each and see if new technology (SW / HW) can deliver 10x improvements.
3. Determine whether these markets are a short enough hop away from deeper ones.
It's surprisingly hard to get a "map" of existing markets, but am curious about those the community can readily identify.
by quickthrower2 on 8/6/20, 9:54 PM
1. Start with markets with high competition. E.g. paid alternatives to Google forms.
2. Get a list of 50+ products competing in the space. If < 50 go back to step 1.
3. Google to find what people, who are on the paid tiers of those products, complain about. This is easy as I am now Googling brand names so should get laser targeted results (v.s. googling "problem I had creating a form" -> Stack Overflow user who'll never pay for a form!)
4. Interview them to dig in further. If you can't get any of these people to even spare 5 minutes to talk, then it might be an indicator that you wont get them to buy.
5. Based on this, derive a hypothesis for a MVP that would solve the problem, along with the market it serves and where to find these people.
6. Presell to people in #4. If they say no - dig in further as to why. If they say yes, aim for maybe $1000 monthly revenue presold, then build.
The reason for this approach is it filters for the "are people motivated enough to spend money" which I think is the biggest risk for the ideas I come up with. Since they are (they already use the "competitor" product"), can I carve out a niche where I do something better for a specific group of people? Can I reach them easily without spending crazy money on ads? And am I solving their problem?
Caveat is this is designed as an idea generator for an Indie Hacker style project, not a startup!
by thorwasdfasdf on 8/6/20, 6:47 PM
by Animats on 8/6/20, 7:11 PM
No, it's not. It might take more than a Google search.
You can extract what you want from US business census data. See "data.census.gov". Look for NAICS codes with a small number, but greater than 3, companies, and high dollar amounts for the category.
by jrumbut on 8/6/20, 8:11 PM
The incumbent players are ignoring massive swaths of the addressable market at a time when, judging by my social media feed, there is renewed interest in labor issues.
A big reason for this is their reliance on a high-touch, manual onboarding process and slow, high overhead contract negotiation techniques. Prime for marketing automation, SaaS tools, chatbots, etc.
Additionally, for purely historical reasons, they have segmented the market by trade. There's no particular reason for you to follow this path, may as well help everyone get a better deal from their employer (and a cut for yourself!).
by tcmb on 8/6/20, 8:37 PM
There is software for translation management, which is essentially project management, but because of the intricacies of the business there is specialized software for it. There are maybe 2 or 3 main competitors, all of which do the job but are fairly awful to use. Then there are a handful of products which come from computer-assisted translation (CAT, which is different from machine translation), and try to capture the project management part as well. Because their focus is on CAT, they're also not excellent in the management aspect.
If you want to capture the whole spectrum with all the edge cases, it's going to be a very complex product which will require a lot of user research and take a fairly long time to build. But there might be an opportunity to go to market earlier, with a subset of the functionality, and build from there.
by kirillzubovsky on 8/6/20, 9:52 PM
(original audio segment: https://smashnotes.com/p/y-combinator/e/92-ryan-petersen/s/h...)
by vsskanth on 8/7/20, 12:08 AM
High sample rate (20 Hz min.) sensor telemetry (9 DOF IMU + GPS) over cellular network, as a single encapsulated unit with built in antenna and a battery to handle power interruptions.
Data needs to go a remote server and devices should be remotely manageable if deployed as a fleet.
I searched quite a bit and eventually ended up using a bunch of Samsung Galaxy phones in a hard case, with a datalogging app and an automation that syncs log files to Dropbox. Devices were managed over TeamViewer.
There are some small players in this field each satisfying 70% of these requirements and they're printing money.
All the pieces are out there. Just need someone to make a product so enterprises can throw money at them.
by muzani on 8/7/20, 2:34 AM
This fit in with your other requirements - it would be easy to hop from low carb recipes to say, Indian vegetarian recipes.
The 10x trick is done by looking for crappy apps with a substantial user base. In our case, we were competing with a FB group with 200 thousand users and a constant stream of posts that made it difficult for people to look for recipes. Our app competitor was a HTML5 app which someone did as a technical demo, with 50 thousand users.
by ideals on 8/6/20, 10:24 PM
This is something I've been thinking about since my HOA sent me an email a month ago warning me that the sidewalk outside my property may be damaged and in need of repair because if someone tripped I'd be liable for this. Of course the email had a bunch of red and gold text.
Luckily for me they already had a concrete cutting company come by and survey the area and provide estimates for each plot.
My bill would be about $200 to fix the sidewalk. I looked outside and it was totally level and no edges exposed.
In really small print at the bottom of the email was information telling me it was optional to do, but again I might be liable if anything happens.
Today I saw this person outside doing the concrete repairs. It was a guy with a small trailer and grinding down the sidewalk with an angle grinder and a vacuum to suck up the dust.
This basic concrete cutting job is pulling money in if he can go around and tell all the HOAs about the urgent need to fix sidewalks and HOAs pass that along to all the residents.
How many people paid the $200 "to be on the safe side"?
Now you repeat for all the other services you can think about for home maintenance and repair and contact HOAs with estimates they can pass along.
by claudiulodro on 8/6/20, 7:07 PM
This doesn't necessarily help you come up with new software, but it is something I've spent a lot of time thinking about since I'm also trying to start a business.
by jmchuster on 8/6/20, 8:11 PM
It might be worth looking for evidence of attempts at building such businesses, since often such domain expertise doesn't pair with business-building expertise. So if you're trying to build such a map, then maybe look at all the businesses (including all the failed ones) that served each market.
by bobosha on 8/6/20, 8:52 PM
by mamcx on 8/6/20, 8:53 PM
I work in the small-bussines and before in government. have done some stuff for some of the biggest companies in my country (Colombia).
To say ALL of them are like 20-10 years behind is not say enough.
Today, I'm integrating with cobol and other stuff that only have text-based files as interface.
A problem is that not many investment are in this long tail, so when talking about solutions for this market is possible to NOT get excitement for it.
I'm building on the side a relational language that eventually could be an Access+Excel tool, but everyone is interesting in the markets that reach billons :).
by rsweeney21 on 8/7/20, 12:29 AM
I picked tech recruiting for my most recent startup. :)
We've bootstrapped to over $2M in revenue in 18 months. People seem to like the service much better than traditional recruiting.
(Shameless plug: https://www.facet.net)
by pieno on 8/6/20, 7:12 PM
There may be an expectation that there are higher profit margins in non-competitive markets, so you may have an advantage there that any innovation will allow you to capture more excess profits as it will take longer for those excess products to be eaten by increased competition you triggered.
On the other hand, there’s an expectation that the incumbent(s) in a non-competitive market has a lot of funds at its disposal (due to years of capturing producers excess profits in a non-competitive market) to hamper innovation.
To me, this illustrates that whether the market is competitive or not is not very relevant. The question is whether you are able to outcompete the market by reducing costs and/or improving your offering compared to your market, whether that’s a single monopolist or 1,000 highly competitive companies.
by tedmcory77 on 8/6/20, 8:06 PM
If you're not, it's worth it's weight in gold when evaluating markets and looking at how do build your business model.
by jariel on 8/7/20, 4:03 AM
A different way to approach this problem, is to consider culture: 'people like us, are attracted to certain things'.
Those fields are overwhelmed.
So, where are the 'completely unsexy' areas?
1) Human Waste. Port-o-potties are a huge business, and it's a nasty problem, pun intended.
2) Agriculture. Farms are not hip.
3) A lot of goods and services targeting low-income people who are extremely price sensitive and don't care about how cool things look.
4) Anything to do with a business that is mundane: cleaning, forensic accounting, physical security, garbage, etc..
Literally walk down the street and look at what people do, it's funny how you can get an instinct for how very 'specific' YC pattern types are, and how we're a little bit out of touch.
by dzonga on 8/6/20, 11:05 PM
by adrr on 8/6/20, 7:56 PM
Licensed Banking: a there hasn’t been a new chartered banks in a decade
Also Clearing Brokerage. Credit Bureaus. US Core banking software, KYC providers(Eg:lexis nexus), bill pay providers
by PeterCorless on 8/6/20, 11:23 PM
X. Market size (actual values, divisible into competitors) Y. Market valence (positive/negative consumer/user/customer/investor perceptions of each of the competitors; e.g., some sort of objectified, normalized product/service/company rating) Z. Market momentum (vectors of growth/shrinking/stagnation of the market overall, and each of the competitors within it, broken into actual past vs future projected)
Many times a market may not grow not because the market "isn't there," but because all of the existing offerings frankly stink. Imagine movies. If you have 12 terrible science fiction movies in a row, does that mean there's no market for science fiction movies, or does that mean someone finally needs to make a good science fiction movie?
You might also have to define more granular domains. Maybe the market might be for a geographic region, or a vertical or horizontal market sector.
by austincheney on 8/7/20, 1:32 AM
2. The cool thing about this space is that it isn’t new technology that delivers 10x improvements. The opposite actually. 10x improvements are found by doing the least possible effort in a tiny library, scaling to support a wider problem space, and doing it substantial better than large products present in the marketplace.
The few competitors are easy to search but most of the commercial products are so expensive that it’s hard to really gauge the precise performance capabilities are the smaller players without a large financial investment.
3. I am just now learning about second order effects from this thing as I extend it to support execution across networks/VMs. There are all manner of second order surprises that emerge from better software automation.
by disambiguation on 8/6/20, 7:09 PM
A sample scenario is one of his senior colleagues takes a PDF and writes in some comments -- almost like a teacher grading an essay with a red pen, change this x to y, etc. This goes to my friend, who reviews, edits, and returns the draft by email. And there will be a lot of back and forth, plus other people making edits in real time.
watching this i can't help but notice how remarkably similar it is to programming, minus all the convenience of programming tools.
i don't know what the product or barriers are here, but these people could easily 10x if they knew about version control and markup languages.
by atarian on 8/6/20, 7:58 PM
- home services: lawn mowing, pest control, firewood delivery
- vehicle services: mobile oil change, mobile tire sales, locksmith
- event/seasonal services: catering, event management, tent rentals
- entertainment: bike tours, party rentals, hunting guides
- personal care: mobile haircuts, mobile makeup, mobile massages
- training/coaching/consulting: pet training, college application consulting, credit repair consulting
- trades/construction: electrical, plumbing, carpenting
- business services: videography, bookkeeping, junk removal
- real estate transactions: moving services, realtor, vacation rental management
by kraig911 on 8/6/20, 7:48 PM
by bingobongo1 on 8/7/20, 9:23 PM
You should look for markets where a reconfiguration of leading technology could serve a new market that existing leaders in the space are unable to service due to structural necessities in their own established business models.
Trying to compete against incumbents by improving what they are already making progress on (sustaining innovations) is nearly impossible because they've set themselves up to be leaders in incremental innovation already.
Disruptive markets are so because they emerge at profit margins and market sizes much too small for large existing organizations to justify servicing, and also the new products do not meet the requirements of the established customer-base. Eventually, there is a chance that the disruptive technology will intersect with mainstream customer requirements and this is where customers begin to see the 10x benefits you're looking into, much further down the road.
by skmurphy on 8/6/20, 7:15 PM
by keiferski on 8/6/20, 7:17 PM
by whalesalad on 8/6/20, 11:56 PM
Startup idea. Automate this analysis and sell the aggregate. No one wants to do this work, but if you had a trustworthy source to exhaust all info for a niche or corner of the market, that would be cool.
by eaandkw on 8/8/20, 12:30 AM
I've lost count of the things that I have used in the past that I have stopped using because the product sucks now and is of lower quality or they have blessed me with a subscription model that only cost $5 or $10 a month. Instead of sucking it up and paying for something once I have to die by a thousand cuts and pay for something the rest of my life if I want to use it. Sure it's great for the profits of that company but not for me.
So maybe that's the underserved market. A product that respects privacy and leaves you the heck alone.
by apapli on 8/7/20, 12:21 AM
You are better off finding something you are passionate about in a competitive market, then working out how you will differentiate and market it with the budgets you have to work with.
by rsecora on 8/7/20, 6:25 AM
Current messaging solutions have different issues.
1. Not encrypted.
2. Transmission is encrypted but they are centralized, the message is open in the server.
3. End to end encryption but susceptible to block because some signaling/presence is towards a shard of central servers.
4. The infrastructure relays on well know properties of IP/TCP (proxies, ports...).
5. the packet has some well know properties (checksums, structure...) that allows to be identified and blocked.
5. Authentication and peer identity discovery.
There was a time (early 2000s) where Skype was a headache for system administrators, because it was able to pass any firewall and the conversations were able to pass. That is missed in the current world.
None of the current solutions, telegram, line, whatsapp, skype, email... can pass the previous checklist.
by missedthecue on 8/6/20, 11:31 PM
by this2shallPass on 8/6/20, 7:12 PM
What's the value of a good "map" of an existing market? What are you willing to pay for one?
by lukaszkups on 8/7/20, 9:11 AM
I've heard multiple times already that price margins are huge at that one, and nobody are determined enough to offer lower prices for them.
But that's only what I've heard from here and there, I'm not into that topic "that much" to confirm that - OP asked for some suggestions, so I gave one ;)
by Keyframe on 8/6/20, 8:23 PM
by flaque on 8/5/20, 11:57 PM
by maxk42 on 8/6/20, 7:55 PM
by Spooky23 on 8/6/20, 11:05 PM
Sometimes you can buy some little company that does something minor like format print output for some ancient mainframe and the extract more money or find complimentary sales.
by landemva on 8/7/20, 2:36 AM
Unfortunately, States have medical device approval hurdles which often explicitly consider competition, and prohibit the economic competition.
by slantaclaus on 8/6/20, 6:47 PM
by mNovak on 8/7/20, 4:15 AM
by nosmokewhereiam on 8/7/20, 3:18 AM
by m463 on 8/7/20, 6:13 AM
by kyawzazaw on 8/6/20, 11:18 PM
by m3kw9 on 8/6/20, 11:06 PM
by JTbane on 8/7/20, 12:47 AM
by faangFar on 8/6/20, 8:05 PM
You want medicine, you need to pay someone who's part of the Physician cartel, a pharmacy, a pharmacist, and whoever has the patent.
With computers and science you often need 0 of these except the manufacturer of the drug.