Feed Wrangler, The First Month

One month ago I launched Feed Wrangler. Feed Wrangler is an RSS aggregation platform that I’ve been working on since the start of this year. Google killing off Reader forced me to get my act together and actually ship the product. Here are some notes from the first month of it out in the wild.

Before I launched the aspect of it that I was most worried about was the price I charged. Charging $19/year is a level that I simply didn’t have much experience with before this launch. I am delighted to report that I’ve had little push back on the price. It seems like people generally accept that providing a stable, sustainable platform costs money and are willing to pay to support that. It is a welcome break from the cut-throat world of the App Store.

When I released Feed Wrangler the response I received was far stronger than I was expecting or hoping for. It was picked up by a number of sites in its first week and got a solid adoption right out of the gate. I’ve never built anything that was so difficult to scale out. Most iOS apps scale themselves to some degree. There is often some kind of backend you need to build out but previously for me this was never the core functionality.

Feed Wrangler is its backend. Keeping the feed scrapers scraping, the website loading and the app syncing turned out to be a much more nuanced problem than I had anticipated. The first week or two were full of late night server administration and frantic performance optimization to keep up with the demand. Though I can’t really complain, it is a delightful problem to have.

While completely exhausting this was incredibly interesting work. The most interesting part is the inter-disciplinary nature of the problems you have to solve. I’m not just sitting in Xcode writing Objective-C. I have to jump between low-level Linux kernel optimization, database tuning, application programming and visual design. This stretches you as a developer and can be painful at times but that is how you get better too.

It looks like Feed Wrangler’s next chapter will be focused on helping 3rd Party developers integrate with the API. The range of developers that are working on adding Feed Wrangler support to their apps is stunning. It looks like there will be great coverage across most platforms. While I can’t talk about specific apps quite yet, I don’t think anyone will be disappointed by the integrations coming soon.

Feed Wrangler has grown in its first month of life into something I’m rather proud of. I’ve been able to tame the technical growing pains and now find myself with a solid platform from which to grow. I look toward the July 1 Google Reader funeral feeling comfortable about my preparation for it and hopeful that I’ll be able to provide a smooth transition for my fellow news junkies going forward.

Some interesting facts about Feed Wrangler after a month of usage:

  • 67,457 RSS feeds
  • Which generate an average of 9,844 new articles each hour
  • The scrapers routinely sustain a combined 40Mb/s while pulling feeds
  • A typical user subscribes to around 65 feeds
  • The user with the most subscriptions watches 1,140 feeds
  • The most popular sites are Daring Fireball, Marco.org, TUAW, MacRumors, and MacStories.
David Smith