The Dreaded App Store Fizzle

While on a long walk one afternoon in the fall of 2010, I thought about how nice it might be to catch up on Twitter while walking. Reading on a screen isn’t particularly practical, but surely someone had created an app that reads tweets aloud. Turns out there were a few in the App Store, but they were either completely broken or poorly implemented.

Elfred Pagan had been doing contract work for App Cubby and mentioned he was interested in doing his own apps at some point. So I pitched him the idea of working together on this Twitter app I was thinking about, Tweet Speaker. The terms: I front the money for art and other expenses, manage the project, and do all those other things I do. He does the coding. We payback the expenses then split the profit 50/50. The app sounded like a fun coding project and seemed like it had some market potential, so he agreed. The only problem was, he had just taken on a new client and ended up completely swamped until Spring 2011.

We set to work in late March with a plan to quickly knock out a simple, but polished app working part-time. The first deadline we set for ourselves was May 23rd, the deadline for apps to be considered for an Apple Design Award. That quickly flew by.

By the time WWDC rolled around in June, we had a decent Alpha version of the app to show off. I got some great feedback from fellow developers and a few friends in the Apple press. Late that week I scheduled a consult in the UI labs. I was pretty confident in the current design, and overall it did look quite good, but the Apple artist I sat down with tore it apart. With a relatively minor reorganization of the various features, he completely changed the feel of the app. Though only a few buttons moved, it meant that the entire UI had to be redesigned.

June passed in a flash, then July and August. By early September I was starting to get worried. If we didn’t get the app out immediately we might end up launching the week iOS 5 was released, or worse — the new iPhone.

Despite our best efforts, we didn’t get a 1.0 build to Apple for review until September 22nd. Then we found a show stopping bug and had to re-submit on the 24th. The app was approved on the 30th, but I decided to hold the release so we could get promo codes to friends in the press and get organized for the launch. So, I set the launch date for October 6th, hoping Apple might feature Tweet Speaker in the App Store that Thursday.

Then Steve Jobs died.

I was stunned. Though I never had the privilege of meeting him, his visionary work changed my life. As a full-time iOS app developer, the money that pays my mortgage and puts food on the family table comes from sales and advertising on those magical little devices designed in California. And as a user, my life is better for the joy and ease of using the various computing devices and accessories he and Apple had the taste and vision to create.

After just a few minutes of mourning and reading tributes on Twitter, it dawned on me that Tweet Speaker had already launched in New Zeland, Australia, and other Asia-Pacific countries. My first instinct was just to pull the app from the App Store and wait for a more appropriate time to launch. But reviews were already scheduled for the next day, and I had a good hunch Apple might be featuring the app. What would happen if I pulled the app? I had no idea. What would happen if I didn’t pull the app? I had no idea.

Somewhat encouraged by tweets like: “If you want to honor Steve, make something. Do something. Ship something. Now. Today. Everyday.” I decided to leave Tweet Speaker in the App Store.

The morning of the 6th I tweeted a few times about Tweet Speaker having launched, but abandoned all other launch plans. As I said the next day on Twitter: “Life does go on, but the sorrow of losing a personal hero doesn’t mix well with the joy of unveiling a new creation. I’ve been a mess.”

Even though I stopped pitching the app, and it got significantly less press than it might have on any other Thursday, sales the first few days were quite strong. Being featured prominently in the App Store all but guarantees a few amazing days. But the speed and severity of the post launch fizzle has been shocking.

After paying Acapela a percentage of sales for the licensing of their text to speech technology, Tweet Speaker has now made $5039.90. Well, not actually “made”, that’s just the gross revenue to App Cubby, we spent $8,600 designing the icon and UI. At the current rate of sales, it will take 630 more days just to break even on the art.

Unfortunately, Elfred and I didn’t track the exact number of hours we spent building Tweet Speaker, but a conservative estimate would be that we each invested $25k worth of time into the app. I’m not even going to bother calculating the number of days it would take to payback our time. Are we stupid for having invested close to $60k on an app that reads tweets aloud? That’s a strong possibility.

But we’re not going to give up without a fight.

A while back I had a rather lively discussion with Daniel Jalkut and a few other developers on Twitter about the potential for an app to make a comeback after a lackluster launch. It was sparked by these Tweets by Mike Rundle: “Laughing at [Hacker News] commenters saying you should [be] embarrassed by your 1.0 release on the App Store. Pro tip: nobody cares about a polished v1.5.” and “If your v1.0 iPhone app isn’t successful, ditch the app and build something new. There’s a fundamental problem that v1.1 can’t fix.” I sided with Mike in the conversation that ensued, but I’m hoping to prove us both wrong.

I’m incredibly proud of what Elfred and I accomplished with Tweet Speaker 1.0 and really think it’s a great app, but we’re facing a huge marketing challenge. When first hearing about the app or reading the description, people seem to react with “meh”, but if they get past that and actually try the app, most find it surprisingly compelling. Matthew Panzarino of The Next Web said it best: “I never thought I would need a Twitter client that spoke my tweets, now I can’t live without it.”

So, how do we overcome that challenge? We’ve got a few ideas and are working on some newsworthy features that should launch before Christmas. And we’d love input/feedback from fellow developers and anyone else who was interested enough to read this entire post. Email, or mention @drbarnard or @elfredpagan on Twitter.