In April last year I started thinking about using Twitter in a smarter way. I wanted to do analytics on my tweets and find out more about people following me on Twitter. What kinds of things do people who follow me like and retweet? I decided to dig into the data and find out.<p>When I went to try and download Twitter data in raw form I found I quickly got bogged down writing API wrangling code and fiddling with API keys. I just wanted to crunch some data but here I was wrangling Twitter's API. This was such a frustrating experience it suddenly looked like an opportunity to me. Was there room for a product here? A product which does one simple thing: help people extract Twitter data painlessly without writing any code.<p>I did some research and discovered some tools that purported to do this, but none of them worked well for my use-case and all were badly designed and/or expensive. I decided to take a shot at it.<p>I worked on this as a side project in a "calm company" fashion. Each week for 26 weeks I would put aside one day to chip away at the features. I tried not to think about how the issue tracker was filling up more and more. Several times I pared back the feature set to try to really focus on the core use-case.<p>When my first user reached out and engaged I knew I might be onto something. I kept posting progress on Twitter and a few more people started to use it each week. Some of them came back, hinting at possible user retention. I hired a writer to write some articles to help with SEO and I kept working towards and MVP that I could use to test the market.<p>Finally the day arrived where all of the critical issues in my issue tracker were closed. That meant it was launch day. That was yesterday. So here I am releasing this on Hacker News to you, dear reader.<p>God speed little micro-SaaS, may the winds of fortune be at your back.
What assurances can you give that you will be around in a year because I think this could get crippled easily regardless of how you are doing the fetching on the backend (scraping, api, nitter). I'd love something like this to not have to wrangle with the api myself but it's giving uneasy vibes for longevity.<p>Edit: I tried it, got the following for "tweets from a search result"<p>Error: Request failed with code 429<p>Request exceeds account’s current package request limits. Please upgrade your package and retry or contact Twitter about enterprise access.
I like the spirit of this, but the monetization strategy rubs me the wrong way. You're basically reselling a UI wrapper for the twitter API.<p>Nevertheless, good luck and I hope they don't shut you down too fast.
Colour me curious, but why do I need to sign in with twitter to use this? And if I do and you get c&d'd do I get blocked by twitter?<p>Edit: spelling
You were banned for either one or both of the following things:
1. Violating II. Restrictions on Use of Licensed Materials of the Developers Policy => A => ) provide use of the Twitter API on a service bureau, rental or managed services basis or permit other individuals or entities to create links to the Twitter API or "frame" or "mirror" the Twitter API on any other server, or wireless or Internet-based device, or otherwise make available to a third party, any token, key, password or other login credentials to the Twitter API;
2. violating the security model of the Twitter API, because with your Service Twitter is no longer in control of who can access their data over the API.
> I just wanted to crunch some data but here I was wrangling Twitter's API.<p>> I worked on this as a side project in a "calm company" fashion. Each week for 26 weeks I would put aside one day to chip away at the features.<p>So you wanted to crunch some data...and instead of spending 3-4 hours setting up the API properly...you instead spent 26 days writing an unofficial hack?<p>It seems like what you really wanted was to spend 26 days on a coding hobby, not actually crunching some data.
Great idea (despite the dubious legality).<p>Perhaps a section on <i>what</i> can be done with the data? Your potential clients are probably the type of users who don't know why they'd need such a tool. And an explanation of the sentiment analysis?
Charging for web scraping feels scummy somehow. Can't quite put my finger on why, since I wouldn't feel guilty about scraping twitter myself.