No one wants to pay for APIs. They want to solve headaches, appease their bosses, build cool things. Figure out a problem that a niche group of people have and you're golden. There are people making 6 figures selling CS:GO skin price feeds.
> There are people making 6 figures selling CS:GO skin price feeds.
Really?
I was working on a project a few years ago that needed this data so I setup a scraper to get all price data from Steam and Bitskins. It was running 24/7 and putting everything into a database that was very clean and neat. Maybe I should boot that back up and just sell access to the API...
Yes. The gambling sites that accept skin deposits invest a lot of effort into guaranteeing that someone can't fleece them by manipulating the price of an obscure item.
On a similar note, there are also folk making 6 figs selling access to steam game hour idlers.
> The gambling sites that accept [CS:Go] skin deposits ...
I had never heard about this before, but it makes (some small amount of) sense. Steam game hour idlers are also something I never imagined btw. We are truly living in the future.
I figured they have their own scrapers and algorithms to determine what the prices are for an item. Because a third-party could subtely scam them by altering the prices
No, that's a whole other business. Idler services just run up the hours on games in your library without the games actually running. Vanity knows no limits.
Ok but you could easily view the question as "What headaches to you want to solve and/or cool things you want to build that could be done using an API that you would be willing to pay for"
So here is an API we would have paid for, and I assume many still will: A product catalog. They do exist, but they are expensive and aren't not that great.
What we where after and gave up on was an API that would accept an EAN-13 barcode and return product details, images, good descriptions, in multiple language (specifically NOT machine translated) and possible alternative EAN-13 barcodes for the same product.
As a side project to this: An API that will take a list of EAN-13 barcode, a price and a quantity (in stock) and transform it into a number of feeds for various price comparison sites. Again, it exists, but it's expensive, not all that good and it certainly doesn't automatically add product name, descriptions, categories and product attributes.
I think that issue here is, and it's almost bound be the same for most other APIs, it's the data that's interesting. So what data do you have access to that most others don't? My guess is that there aren't any large and interesting dataset that a random person on HN can easily sell, that isn't already readily available.
Not exactly what you need but I think it may be close.
I’m running a SaaS turning affiliate platform content into an API which can be looked up by EAN13 : https://datafeedapi.com
Current users are publishers wanting to display affiliate links on their website but I could probably repurpose it for a more generic product search API without the affiliate links and different pricing. If anyone is interested, you can reach me at patrick_[at]_datafeedapi.com
> As a side project to this: An API that will take a list of EAN-13 barcode, a price and a quantity (in stock) and transform it into a number of feeds for various price comparison sites
From this: do the people who have the data have an incentive to make it available, or to keep it unavailable so that price discovery is harder? The more efficient the market, the smaller the margins.
This is a big factor in many things that don't have APIs or actively discourage automated access with anti-scraper tools.
> I think that issue here is, and it's almost bound be the same for most other APIs, it's the data that's interesting.
I love building APIs. The biggest hurdle I see is that you need two active parties on both ends of the API to convince: There is a user, which needs to be convinced to use the API, and a supplier, which also needs to use the API. And there is no incentive to use the API for one or another.
If you want to tackle this, make a Product Information manager or Product Experience manager(PIM/PXM) that is easy to use, very API friendly and sanely priced.
A lot of products in this space are priced crazy for small to medium manufacturers.
May ten years ago, at my former employer, where we would have loved to have this API, we also needed a PIM. At that time the commercially available ones was where all pretty terrible or way to expensive (mostly both) and a colleague was tasked with just building one. He did a much better job than any of the off the shelf solutions we had seen. It still needed data to be manually entered for new product and descriptions done for seven or eight language.
I still see a huge value proposition in an API that can lookup production information based on barcodes alone. Scan the item when it first arrives in the warehouse, or when a PO is made and just suck down the relevant information. You could even tier the offering: Tier 1: you get title and one image, Tier 2: Five images and product attributes. Tier 3: All images, attributes and product descriptions. Then add a fee for every language you need translations for.
If you could do this right, then it is my belief that it could change a number of industries. Technically it's not even that hard, but there's a ton of manual work involved, contract with every manufacturer on the planet and insane QA.
More importantly than "finding an API I am willing to pay for" is the reliability behind that API. Building on top of a third-party API requires trust, and once you've done it a couple of times, being able to trust the interface is maybe 90% of the decision-making on which API provider to pick, where the functionality itself is maybe 10% (unless I can self-host the API, of course).
Third-party risk is the biggest problem with APIs. The amount of calls I've been in where a third-party provider tried to sell their API and told me they don't need versioning right now, they will "try to" pre-announce breaking changes, they don't have an SLA or uptime commitment, etc. etc. is way too many. It's so hard to find an API provider that seems to be trustworthy enough to build a product on top of it.
I feel like there's a magic number of salaried positions that need to be tied to an API provider, to ensure it's both dynamic enough to react to generational changes in networking/security/etc. but also slow enough to not keep throwing everything overboard whenever a new framework or middle-manager comes along.
I would pay a small amount (maybe $5 AUD per month) for access to an Open Banking API that lets me get the transactions, balances, etc from my accounts. Unfortunately under the current regime for Open Banking in Australia (part of the Consumer Data Right) you need to be an "Accredited Data Receiver" which I completely understand if I am accessing and storing financial data on the behalf of others.
However for my own personal accounts... just let me call the API!
I too would like this for the UK, and I believe the situation is similar. I've just tried Nordigen and it looks like it's the solution! ""Free forever"", as well.
I didn't know about Nordigen, but how are they able to support the absurd amount of 1222 banks in Germany? Looks like a good indicator how broken the digital world of banks here really is.
True, from a technical point of view, this should actually be quite simple. But I'm referring more to the administration overhead caused by the large amount of small banks that are considered their own credit institution. In a lecture, we were once shown a chart comparing the costs of financial operations in the European countries. Germany had by far the highest costs. This mess will not survive the next 10 years.
I am curious as to your phrasing of "besides by doing market research" — market research is the whole point!
Most successful API products in the sense that you're describing share two common characteristics:
1. They are a thin wrapper around some dynamic dataset.
2. The cost to an organization maintaining that dataset is prohibitively large and/or difficult relative to the value they receive from it.
If I were in your shoes, I'd find a niche that I was relatively well-positioned to address (say, Shopify stores, or programmatic SEO — something small enough to have a uniform opinion but large enough such that there are multiple cohorts of individuals willing to pay) and spend time in their forums looking for people complaining about datasets or information they wish they had. The API part is an implementation detail.
I would like an API which provides consumer electricity cost data globally. There is a huge amount of potential for internet connected energy consumers to optimise their usage against carbon emissions and energy costs. Carbon emissions are available for a lot of countries through APIs like https://app.electricitymaps.com/map and https://www.watttime.org/, but consumer energy costs is much harder to find. There are some APIs that give some pricing data for the purposes of comparing suppliers in competitive markets, but few of these give accurate time-of-use costs for energy. https://www.genability.com/signal/ covers the US and a few other markets quite well, but doesn't have complete and global coverage.
A geo places API with the actual international ids of those places. There are a lot of geo something APIs but I still cannot find one where I can:
- Ask for all admin1 regions of a country and have all their actual data (like: all states in the US, or all regions in Italy)
- Ask for all admin2 regions in a given admin1 region (all counties in the state of TX, all provinces in the region of Lombardy). As you can already see "US.TX" is an identifier but "IT.Lombardy" is not, it is language dependent and I'm sure there are international identifiers for those but they are lacking in almost all geo something APIs
- Ask for all admin3 in some admin2
- Ask for all admin-whatever on a given ZIP code
- Ask for all ZIP codes in a given admin-something
- Bonus point if it supports i18n and l10n
If anyone knows of one please tell me, I would pay for it right now.
You can use Wikidata and its Sparql query language for this. Although it's not straight forward for all countries, e.g. some have especial properties for regional breakdowns that are not defined as ISO standard subregions.
Thanks for the suggestion, but I already tried it in the past and it is almost impossible to have correct results, it just takes a missing category on a country/state/province/city to make it all fall apart.
For example it is common to receive historical entities if they are mislabeled, entities that no longer exist because they were merged... it is good for a side project but not suitable for production.
I'll try to say it in the least words I can so that it might be easier to explain:
Suppose I want to let my users input their customer's addresses. First of all I need some auto completion: the "state" select must suggest only states of the country selected previously and then so on with province, then city. The ZIP code must be one of the ZIP codes of the admin area selected.
Then they should be able to make this data accessible to other entities, for example a shipping company, so the identifiers of those address components must be "universal". But each of those address components also need localization and internationalization: if shared with an Italian I want the region to be called "Lombardia", if it is shared with an American it should be called "Lombardy", etc...
Crime data by location. Many real-estate sites removed their crime maps due to racism, but I think it’s still a valid metric to consider when moving to a new location.
> Many real-estate sites removed their crime maps due to racism
Really? Forgive me for wanting to know whether a neighborhood is fraught with crime before forking over hundreds of thousands, possibly millions, of dollars into a home.
Some already do, depends on what you have bought.
I would start by reading Home Assistant documentation, that tends to help figure out if it has local API.
And you can almost never go wrong with devices that use Zigbee.
So my lessons here after trying to tackle this market, it's hard. You can look at articles by Rapid and others that list "Top APIs". You can do some customer research by going to places like HN or reddit where people say, "I need an API for this". You can look at existing single purpose APIs in the market for niche use cases like currency conversion, IP data, etc. Or you can try build something and anticipate user needs based on your own. I built M3O.com for this purpose. Ultimately it was supposed to be a two sided market, where backend devs build for frontend devs, but never materialised that way.
I think either catering to your own first need is better than trying to build a thing you might not understand well. The alternative is to just do a poll on a large channel of distribution like a nocode platform where they're dying for a certain dataset.
I want you to provide me an API with sane limits so that a) I don't need to spend time scraping your site b) you don't need to serve me superfluous HTML just because I want a bit of data from your site.
In general I might pay for APIs that correlate data from multiple sources into one endpoint.
As an European, with the current energy situation, I might pay for an API that combines weather data, electricity market prices and other factors into a single API that estimates electricity costs.
Or maybe an API that combines ratings from IMDB, Rotten Tomatoes, Metacritic, Letterboxd etc into one call.
Even without buy/pay get good at offering something people want to use at all. It's harder than it seems. It took many tries for me to get the hang of it, most things people just don't want.
I might be off base but I’ve been thinking about building something that addresses “Automatic Personal file Archives”.
What I’m noticing is that my mom has 1000’s of embroidery files that she’s collected over the years but they are everywhere. They are in zip files, in several different formats, and stored all over her various drives. I think her strategy for storing them changes over the years, ending in a mess.
I’ve been thinking about building an app that she can drop files on to collect them. It would capture some meta data, store the binary files somewhere, and make them searchable and browsable, like a custom file explorer.
I recently read about the advantages of storing small binary files in SQLite, and I thought that sounded like a good solution. The “archive” in your example. The SQLite file could then be stored on Dropbox, Google Drive, or backed up to my own hosting service.
Once she starts to collect them, I can add features like the ability to export to a local directory or in a specific format. The machines themselves take proprietary data cards, 1.44 floppy disks or USB sticks and support files in various mfg formats.
These machines cost thousands of dollars and each set of designs costs a few dollars, so people with this problem (if there are others) are pretty invested in their collection.
Wow if you could do this this would be absolutely amazing. Thank you for looking at the page and the idea and thinking of building it. I appreciate it.
I would hopefully be capable of using it too if you choose to share it.
I have a external M2 SSD enclosure and SSD which I use for backups and I use freefilesync to synchronize the files from the OS SSD to the enclosure SSD. It's manual.
I think synchronization is a difficult problem. But if your using OneDrive, Google drive you could simply have a frontend that copies the archive file there.
I pay for the Seeking Alpha API (as hosted by Rapid API) for finclout.io.
The problem that I am solving for the app is that if I only get a stub then I can augmnent this stub with additional information including earnings call recordings.
Generative AI models. E.g. Midjourney which is trained on specific image styles. The only problem is you risk getting sued by copyright holders. No idea where that one is going though, we'll have to see how things play out in court.
Usually good money in scarping price data and selling to traders. HOWEVER, they will expect a very robust API that goes beyond a hobby project / side-hustle level of polish and robustness.
I'd gladly pay for an eSignature API that has pay-as-you-go pricing or a good entry-level plan. Products like HelloSign or DocuSign have high monthly costs associated with API usage.
I was going to suggest Signable [1] which I have used in the past, but it looks like their API is only available on a monthly plan (I'm sure it was available on PAYG before) - but it is cheaper than the 2 you mentioned.
Side note about rapid API, I originally built https://text-generator.io on rapid API but I got 0 customers there, you would think you could just look on rapid API to see the gaps, there wasn't and still isn't a good alternative to text-generator.io on there but that doesn't guarantee anything, market research takes a lot of thought and the product needs constant marketing just like all businesses, I had to launch it as its own website for it to get users, so there's a lot more that goes into it than just building the right API.
I've also had customers scared off or investors not want to invest because it is a plain API platform and not B2C...
To answer the question market research is kind of the only way to know, Google adwords keyword planner and Google trends are quite good, as well as researching the companies in the same space well and what gaps there APIs have.
As an example https://text-generator.io distinguishes itself from OpenAI by including automatic analysis for input links, linked images, documents etc and doing speech to text/self hosting/embedding images, text and code in the same space, so isn't just an API clone