Gus on the Go: Year Two

This is part two in a series of three blog posts giving a background on what my parter and I have done so far for our language apps, collectively titled Gus on the Go. You can find part one here. Go ahead and give it a read. I’ll wait for you.

Year two started out with Alice and I treading brave new waters… at least for us. I had been working on converting the code base for Gus from using the cocos2d-iphone framework to using cocos2d-x. Why go from Objective-C to C++? Portability. We had a bold plan to double our sales. We were going to release Gus on Android!

Ok maybe we didn’t realistically expect to be able to double our sales, but at the time, Spring 2013, Android was aggressively growing its market share. Android phones were about 50% of the market. Although Android users spent significantly less than their iOS counterparts, based on our research (read googled), one exception was the children’s education market. As best as we could tell, for every dollar a user spent on educational children’s apps on iOS, an Android user spent 65 cents.

Expected Android/iOS split

Sorry, I couldn’t actually find the research we read at the time. The important thing is that we believed it. Mistake #3. For mistakes numbers 1 and 2, see the first blog post. To date, we routinely see sales on Android (from Google Play and Amazon combined) as about 17% of our total monthly revenue – which is a little less than half of what we expected.

Actual Android/iOS split

We don’t necessarily regret porting for Android, but there are some headaches involved. For starters many more devices to support[1], which required carefully optimizing certain aspects of the code to ensure compatibility with slower devices. Additionally, the switch to C++ made crash reporting a pain at the time (things have gotten much better since then with Fabric[2]).

But the really big change from the business side that came with the Android port was the decision to hire an app marketing agency. Alice and I decided that marketing, after all, was our weak point. We were essentially running out of marketing ideas and avenues. We had a wonderful product that elicited some fantastic, heart warming emails from our users. The problem, we felt, was that not enough people knew about it. A professional was going to be our silver bullet. Mistake… well, was it really a mistake?

A Marketing Agency to the Rescue

In May 2013, I emailed 5 different app marketing agencies inquiring about their services. I immediately heard back from all but one. After scheduling Skype[3] calls with the 4 candidate agencies, I had a clear favorite. Appency. We would not work with them, then, however.[4] Aaron at Appency, flat out told me that his company was too expensive for us. At the time we were doing $1,796 in monthly revenue. He did, however, speak to me for about 30 minutes on app marketing and gave me some tips that really positively affected our bottom line![5]

After further deliberation, Alice and I chose Appular to help market our Android launch. Something interesting to note here. Appular convinced us to do a launch-style marketing campaign for our Android release as opposed to the long game of marketing. We agreed as they were the experts. Mistake #4 It turns out that an Android launch is not interesting enough of a story to justify a launch-style campaign.

Appular had a list of 166 sites they wanted to target (narrowed down from their full pool). From that list we got 23 write ups, of which 12 were sponsored blog posts). While we did get quite a few write ups, the overall bump in sales was essentially zero.

30 day moving moving average around android launch 30 days of sales before and after android launch

During the campaign, I kept a spread sheet that correlated a site’s write up with their analytics to the traffic brought to our site.

Site Unique Visitors[6] Social Week 1 hits on our site 1.3 M ??? 43
angelasanalysis 1,335 FB: 2,216; TW: 3,342; PIN: 2,335 6
sensiblysara 2,008 FB: 1,560 2
cyber-kap ??? TW: 3,904; PIN: 1,190 27

This is just a small sample of my analysis, to give you an idea of what I was tracking.

We were disappointed. And justifiably so. We were sold on this idea and the results weren’t there. At least not immediately. It would take another 6–7 months before we would realize what we truly got out of this 1.5 month marketing campaign.

One benefit we immediately got from our contract with Appular was a new and improved app description and list of keywords. We took these and immediately plugged them back into the iOS side along with an important tip from Aaron of Appency.

Metadata Localization

One of the pieces of advice Aaron gave me during our chat was to localize our app description and keywords. His logic followed, since our app was localization agnostic (the only language in the app was the target language – immersive learning), we could easily localize our app’s metadata and attract users from around the world to download our app. No change was needed to the app itself.

I discovered fiverr and set to work hiring translators from around the globe fairly cheaply. Fiverr is a micro job marketplace, where people provide services, such as translations, in multiples of $5 (also known as a gig). One piece of information that fiverr does provide is the country of the seller. I decided early on to only hire translators that came from the country where the target language was spoken. My reasoning being, I would more likely get a native speaker.

In total, I spent $75 to get translations in 15 languages. As the prices are cheap, it made sense to at least double check the results using Google Translate to ensure that it roughly means what you expect. I had no problems I could detect this way, but have no way of knowing if the nuances of the translations are correct. For the rest of the languages, I got help from friends, who are native speakers of those languages.

The results?

Pre-localization of metadata Post-localization of metadata

Prior to localization, 74% of sales of Gus on the Go were from the US. Afterward, it was closer to 56%. This is amazing – effectively a 32% increase in total sales solely due to markets outside the US!

In recent times, sales in the US have crept back up to 74%. I need to investigate if this is due to a disproportionate increase in US sales or if international sales have since fallen comparatively.

New Languages

While ordering translation services on fiverr, I also noticed many people offering voice over services in various languages. I realized this could be a great opportunity to expand the number of languages in the Gus on the Go series.

Over the course of several months, I spent $190 to record 114 words in 12 languages. Out of that came 7 new apps on 3 different platforms (so, 21 new apps). That is cheap. Sometimes quality is great and other times it’s so-so.

In addition to those, we also recorded friends for 4 further languages.

Percent of revenue from apps released in year two

These new languages brought in a total of $8,752 in year two. Not too shabby. We also still have 4 recorded languages we could theoretically release. Anyone interested in Bulgarian, Dutch, Indonesian, or Urdu?


I mentioned earlier that we wouldn’t understand the effects our Android launch write ups would have for 6–7 months. It turns out that those articles and blog posts had a big net positive effect on our search rankings. One of the original write ups was on[7], which was recycled by many other sites.

We didn’t really understand this effect until February 2014, when we received emails from Fuhu and Fingerprint. Fuhu was a company that made kid-friendly tablets and ran a curated App Store[8]. Fingerprint works with companies to create platforms with kids apps. Both were interested in bringing our language apps to their platforms.

When meeting with them, possibly the most important question we asked was, “how did you hear about us?” Each time, the answer was the same. They did a Google search for best language apps for kids and Gus on the Go ranked fairly highly. This, most certainly, was because of the media coverage Appular helped us receive.

Cyclical Sales?

After two years of Gus, we started noticing a distinct cyclical nature to our sales. At least we thought we did. I originally wrote this section completely differently. I think our mistake might have been recalibrating revenue expectations based on current months. This is shortsighted because we lost perspective of how much the current months had improved over the previous year.

Instead of showing you the graph I had in my head, here’s the actual graph of year one vs two revenue.

Overlay year 1 and year 2 revenue

Not a hockey stick, but still nice growth.

Email List Update

Our email list also grew quite a bit this second year.

Email list growth

While writing this blog post, this was the first time I had seen or generated this graph. I did not expect the 3 major jumps in size. It’s most likely due to a popular person or site linking to our printables page, but I should probably double check and report back. Writing these blog posts has also given me insights as to what kinds of graphs I should generate regularly and keep tabs on!

Year Two: By the Numbers

  • Sales: 17,224
  • Revenue: $44,883
  • 1* Reviews: 6
  • 2* Reviews: 2
  • 3* Reviews: 3
  • 4* Reviews: 3
  • 5* Reviews: 71
  • Languages Available: 26
  • Total Apps: 78
  • Email List Size: 2,375

Where Do We Go From Here?

This post has gotten really long. I think I’ve pigeonholed myself a little in trying to do one post per year of business… Next up, in Year Three we ran some experiments that would influence our business model for Gus 2. Additionally, we developed some business relationships and looked at some startup accelerators. Coming to a blog post near you!

In the meantime, you can find me on Twitter @yonomitt. I’m sure I’ll tweet when future posts are up.

Have a nice day,

  1. Although Apple is working hard to bridge this gap!  ↩

  2. That’s a blog post for another time.  ↩

  3. It’s amazing how many companies use Skype to conduct client meetings.  ↩

  4. We are currently working with them for the release of Gus 2, though!  ↩

  5. Aaron’s honesty and helpfulness is why we came back to them when we were ready with Gus 2.  ↩

  6. Unique monthly visitors was estimated by a 3rd party service and therefore accuracy is questionable.  ↩

  7. I had to get the page from the WayBack Machine, as the site is now 404’d  ↩

  8. Fuhu recently filed for bankruptcy.  ↩