Building api[dot]market. Anyone Interested?
I am building the "API market" of the Internet. Anyone Interested in collaborating?
Trending on Indie Hackers
I am building the "API market" of the Internet. Anyone Interested in collaborating?
There is a lot to be desired from a UX standpoint. For example the large amount of very bold text. The individual API names are very long and the actual content of the title is not well suited for quick scanning, etc.
There shouldn't be a hero-style CTA on the landing page, the first thing visitors should be greeted with is a well-organised preview of the APIs you have available. Don't think RapidAPIs landing page is designed the way it is by accident, a lot of tuning has gone into it. You can also get inspired by Zapier's App directory https://zapier.com/apps.
The search bar should be available by default, as I might be visiting with a clear goal in mind and want to get to the search immediately. The goal should be to get me to an API I am interested in testing as that is where I will convert.
The individual API page is severely lacking. It's neither SEO-optimised nor UX-friendly. There is an optimal text width to retain readability which is about 70 characters.
I think not having a three-column API explorer is unacceptable nowadays for this sort of service.
Don't worry it's very common for highly technical SWEs to create poor UX, but you will have to fix it in order to have a chance at competing.
Marketplaces are notoriously difficult to succeed in without outside investment. Is this something you are preparing for and in that case what is your USP? You mentioned a better API market, but so far I am not seeing anything that can compete with RapidAPI. Maybe it's something deeply technical or it's a price compression play?
Have you considered how you can seed your marketplace on your provider side? e.g. bringing over RapidAPI providers to your platform yourself, so you will be paying RapidAPI for the usage and increase the latency but have a larger API offering for the developers.
Or maybe you can specialise in a niche of APIs or offer bindings for AI engineers to use in their agents.
Feel free to reach out if you want to chat more.
Thanks a lot for this detailed feedback. I totally agree the UX is really bad and I am actively working on improving it.
Regarding what's the USP? A better experience for both API Sellers & API Buyers, it's a bit detailed and based on feedback from actual API Sellers & Buyers. There is no single feature but a mix of various features such as Stripe billing, searchable logs, exportable user list, and faster load times but above all it's mostly better customer service.
This is a very good point, thanks for sharing
"The search bar should be available by default, as I might be visiting with a clear goal in mind and want to get to the search immediately. The goal should be to get me to an API I am interested in testing as that is where I will convert."
I think the problem is so big that most API Sellers & Buyers totally ignore this horrible UX and are actually buying & selling already on the platform.
Regarding seeding the APIs, we are building APIs in-house also reaching out to RapidAPI sellers to list their APIs here as well and reaching out to companies to list their APIs here. We have received very positive feedback from all API Sellers so far, we are going to use all three approaches.
Again, thank you so much for this feedback. This is very valuable.
What’s does it do ? I mean is there some innovation behind ?
No new technology but the market is mostly an improvement on the API buying and selling experience. I have been selling APIs for last 3 years now on RapidAPI and when I talked to over 50 API sellers and users they all wanted a better API market. At first, I built MagicAPI.com (Shopify for API Sellers i.e. separate store for each API seller for example: pipfeed.magicapi.com, checkiday.magicapi.com, trueway.magicapi.com and around 12+ stores. ) But the problem was that each new user that signed up only signed up for one store at a time. High CAC.
Hence decided to combine all separate API stores into a big API market.
Somehow I got the domain "api.market" as well.
What's unique about it?
Replied to this question in the comment, above please see that.
Unique would be our customer service and the platform itself. We are ex-AWS engineers, so we built the entire backend to be auto scalable. It's serving already 250K APIs per month now. We have tested it to be scalable for up to 36 Billion API calls per month.
Our API Proxy Infrastructure has an uptime of over 99.3% (last 3 months ) -> https://status.magicapi.com/.
We have optimized the platform to be able to do authentication, API analytics, usage calculation, and other operations within 20ms on average. The 99th percentile (p99) of this OverheadLatency(i.e. the total latency added by api.market's systems) is under 40ms.
Lastly we are building the platform to drastically reduce the overall cost/time it takes from discovering an API to actually integrating it into your software.
I'd suggest you work on your messaging
Thanks for the suggestion. Any recommendation? What should I change?
But mainly it's a two sided marketplace and you should have messaging differently for sellers and buyers
Each want the other in a classic chicken and egg, the most important for the sellers to hear about buye numbers, volumes and search/discovery
And buyer actually want price and ease I guess unless your competitors systems are really bad and you can prove improved performance for both and than you just need to hunt these bigger players, make it easy for them to a/b test you against current and help them convert
Thanks a lot for your advice/
-> Yeah working on stabilising the systems. 2 9's is pretty good for most of our API Sellers, working on improving this. Almost no other API Market reports this.
-"By developers, for developers" is nice, but what about it? seems like you want to focus on operational excellence.
-> You sound like an Amazonian :-p Are you?
Anyway, it means that this platform it built for technical people for technical people.
"scalable for up to.." today sounds more of a question about why can't you autoscale? Why are we talking about limits? Are you at the tier 1 enterprise edge where limits are that real?
-> Yes, this is what we have tested so far i.e. up to 15K TPS. Yes, we are a tier-1 API Gateway.
Your adding 20/40ms latency, why? Why not 5ms? As opposed to what others do? Do you have to be a man in the middle for actual calls? (Maybe that's targetting people with more specific experience that I don't have, just as an outsider dev, IDK what's the story) I assume that depends on being in the same dc as you?...
Is your discovery different? How? Is it better than just google searching?
-> Yes, we are man-in-the-middle. We do authentication, usage, API path validation, proxy etc in under 20ms usually and under 40ms (p99). This is much faster than any of our competitors. Most API markets don't really report this number as well.
I totally agree on the messaging part, will definitely work on improving that in the coming weeks.
Thanks for your valuable feedback. Our competitors' systems are actually really bad. I sold APIs on RapidAPi for the last 3 years now and was told by at least 25 API Sellers that they would want a better API Market platform.
If you want your USP to be a comparisons, do a comparison, the number randomly aren't meaningful unless your target audience is exactly you, someone who identifies these issues exactly as their pain point from bad experience and losing meaningful money/sleep due to it
That's a very good point. Will definitely add this this week.