Sponge

High Level Concept:

The ultimate text-to-speech app

 

Job To Be Done:

Listen to text

 

Growth Channel:

Prelaunch: Reddit

Reality: Reddit + content marketing

Analysis: We were able to get a few hundred downloads from making interesting blog posts about different books that were available in the public domain, then posting the blog to Reddit and then advertising Sponge on the blog.

This was a difficult channel to sustain due to the effort it took to create the blog and the relatively small amount of traffic it drove, particularly when we did not have any way to monetize our users.
 

Business Model:

In-app Purchases

 

Technical Difficulty:

4/10

It was quite a bit of work, but it was nothing that was particularly challenging.

 

Closing Thoughts:

We had a pretty clear plan on what we wanted to do with Sponge and were executing on our product roadmap well. However, two things started to become clear. The first was that even though we were solving a problem for many people, it was not a huge problem for a large portion of people. This meant that for the people who it was not a huge problem for they did not care to download a new app to solve it and for the people who this was a big problem for they already had a solution (even if it maybe be slightly inferior or more expensive) and our solution was not compelling enough for them to switch.

 

Here is an interesting twitter thread about TTS apps.

 

 

This is why we decided to stop executing on our product roadmap, pivot and launch a product (in a similar space) that we hope will be so compelling that people will want to download it regardless of any other similar products they might be using.

 

View Screens

 

 
 
 

Advent Calendar

High Level Concept:

Advent Calendar app

 

Job To Be Done:

Get into the Christmas spirit

 

Growth Channel:

Prelaunch: Search ads + user invites

Reality: users did not send other users invites

Analysis: It is incredibly basic advice that I have failed to follow way too many times – you must provide users with a clear reason and appropriate incentive as to why they should invite someone to download your app.

We did not have a clear reason or appropriate incentive for users to invite anyone to the app and therefore they did not invite anyone and we wound up with no growth.


 

Business Model:

Ads + affiliate + In-app Purchases

Analysis:  FOCUS ON ONE GODDAM THING AND MAKE THAT THING EXUDE EXCELLENCE.

 

Technical Difficulty:

2.5/10

The app was quite simple, but we had some backend stuff that was not complicated, but just made for a little more work.

 

Closing Thoughts:

Did a lot of things wrong with this one, but even if it did work and exceeded our wildest expectations it would only be useful for 25/365 days. My irrational love for Christmas got the best of me here.

 

View Screens

 

 
 
 

Truth or Dare

High Level Concept:

Drinking game

 

Job To Be Done:

A game that will help to break the ice at a party

 

Growth Channel:

Same as kings cup

 

Business Model:

Freemium – same as kings cup

Technical Difficulty:

1/10 this app will get a 1/10 rating because it was basically a skin of our kings cup app.

 

Closing Thoughts:

To took less than a day to develop this app since it was an almost an exact skin of Kings Cup and we figured it would just be another ticket in the app store optimization lottery… It was not a winning ticket.

 

One other side note, even though the app only took a day to develop, it did not go live on the app store for quite some time. Apple rejected the app on a few occasions, saying that there were enough truth or dare apps and that the app store didn’t need anymore. I’ve gotten quite a few bizarre rejections from the app store, but this has to be up there. We wound up just changing the name and resubmitting a bunch of times until it got approved.

View Screens

 

 
 
 

Kings Cup

High Level Concept:

Play kings cup without needing a deck of cards

 

Job To Be Done:

Play kings cup without needing a deck of cards

 

Growth Channel:

Prelaunch assumption: App Store Search ads + App Store Optimization

Reality: Could not find a sustainable growth channel

Analysis: Coming off of the app store optimization success of Electric and seeing the absolute killing that Picolo and others have been making in the drinking game space we thought it would be straightforward to make a simple and highly profitable app. Our thinking was that if we made a good product (which we did), got good app store optimization and had a business model that enabled us profitably acquire users via search ads we would have a winner. While the approach makes sense, we failed to get good app store optimization and we could not acquire users profitably via search ads.
 

Business Model:

Freemium. We would offer some game types for free and others that you would have to purchase.

Technical Difficulty:

2/10

 

This is about as simple as an app can get, but just in case I am wrong I won’t give it a 1/10 on the technical difficulty scale.

 

Closing Thoughts:

App Store optimization is black magic. I have no idea how it works and would not recommend that anyone plans for it to be their primary growth channel prelaunch. If you stumble into good ASO like we did with Electric then power to you, but it is not something you can rely on.

 

View Screens

 
 
 

Hydrate NOW

High Level Concept:

Get reminded to drink water when you are drunk

 

Job To Be Done:

Minimize hangover

 

Growth Channel:

Prelaunch assumption: Press

Reality: Press! (love it when a plan works)

Analysis: Snapchat discover really can drive app downloads

 

 

 

Article:

Here is a link to a Cosmo article about the app
 

Business Model:

Use the app as an advertising vehicle for other apps that monetize.

Analysis: Even though we did get a decent amount of downloads (12K) only a very small fraction (<100) downloaded Electric, which was the app we advertised in Hydrate NOW. Another reminder why you must provide users with a clear reason and appropriate incentive if you want them to take action.

 

Technical Difficulty:

1/10

 

Closing Thoughts:

More than anything this app was a confidence boost for the team and I. It was a reminder than plans can work out when you are thoughtful about them. Unfortunately, we were only thoughtful about the growth of the app and not the conversions.

 

View Screens

 
 
 

Electric

High level concept:

Random video chat

 

Job to be done:

Assumption: Meet new people

Reality: The thrill of having a wild (most often sexual) conversation

 

Distribution channel:

Prelaunch assumption: Influencers (Instagram) + word of mouth.

Reality: App store optimization.
 

Analysis: In retrospect, our lack of strategy around how we were going to grow the app was absurd – we were more or less just hoping people would tell their friends about the app without having any built-in mechanisms that would help enable sharing. However, despite our weak strategy and execution around distribution, we got lucky and wound up with relatively good app store optimization. The app has consistently been in the top three search results for “random video chat” which has led to a consistent flow of 300-700 new users a day.

 

Screen Shot 2018-06-11 at 10.17.11 AM.png

 

 

Business Model:
 

Prelaunch: None

Post launch: Sell coins that you could use to specify which gender you’d like to chat with
 

Analysis: I am going off memory, but I believe about 3-4% of users would make a purchase with an average selling price of $8, meaning we were able to make some money. I think we could have done many things that could have increased both of those numbers but ultimately, we knew the only sustainable way to increase revenue would be to work towards having an equal number of males and females on the app (it was 85% male, 15% female). We were unable to do this and that was the main reason we ultimately decided to sell the app.

 

Pivoting to a gay male dating app would not have been a terrible idea.

 

Technical difficulty:

5/10

 

Building the app was pretty straightforward, however, we did anticipate increased technical difficulty if we reached a greater scale.

Closing thoughts:

I think random video chat is a well-established niche where fun products can be made. That being said, I believe that growing the product beyond random video chat will be incredibly challenging. If anyone can do this my bet is on Holla/Monkey.

View Screens

 
 
 

Quvoke

High Level Concept:

Video question and answers

 

Job To Be Done:

???? (should have been a big sign)

 

Growth Channel:

Prelaunch assumption: influencers + users sharing videos outside of the app

Reality: People would not share their videos outside of the app and the app saw virtually no growth outside of the initial users we got from a few influencer.

 

Business Model:

We had none.

 

Technical Difficulty:

2.5/10

There was nothing particularly challenging about developing Quvoke.

 

Closing Thoughts:

With Quvoke, I think that we built a beautiful app, however, that was irrelevant because the aesthetics, functionally and features are not where the app derived its value from – its value came from the other people that were actively using the app. We failed to get people to actively use the app and that is why the app failed.

If I were to do it over again I would focus the Q&A around one vertical and then try and expand from there.

View Screens

 
 
 

Mustknow

High-Level Concept:

Anonymous video question and answer

Job To Be Done:

Learn what your friends think about you

Growth Channel:

Prelaunch: text message invites

Reality: text message invites

Analysis: I am such a scummy asshole for doing this. Our onboarding flow made it so that it was very difficult to not end up texting invites to your entire contact list (I did it twice… by accident). When all was said and done, we wound up sending >1,000,000 text, costing us >$10,000.

Tuition at my university was $9,000 a year – Mustknow was slightly more expensive, but I wound up learning infinitely more. We got >40,000 app download in the first 3 days and it’s safe to say that we did not have 40,000 DAU’s the three days following.

Business Model:

We were going full silicon valley on this one - a business model was the furthest thing from our mind.

Technical Difficulty:

5/10

Some of the face tracking and voice changing stuff a bit tough.

Closing Thoughts:

I cringe every time I think about Mustknow.

We were reckless children in a space that requires a company to be the exact opposite; thoughtful adults. On top of the scummy behavior, it cost us a bunch of money and we failed to execute in many other important areas of the app.

View Screens

 
 
 

Headswitch

High-Level Concept:

Insert any head into any video

Job To Be Done:

Put a head overtop someone else's head in a video without requiring any video editing expertise. 

Growth Channel:

Prelaunch: Watermarked videos shared on other platforms

Reality: could never get the product to work and thus could not really tell

Business Model:

In-app purchase

Technical Difficulty:

9/10

When building this app in the fall of 2016 this app was very challenging to build. Perhaps advances in face tracking tech will make this a more attainable app to build.

Closing Thoughts:

If someone is actually able to build an app the allows people to select any video and then overlay a new face on the faces that are in the video I am certain they will have a big time winner on their hands. Writing this I am tempted to go back and take another stab at creating Headswitch.

View Screens