Why Mobogenie Failed — Part 3: The Failure of Operation

This is the last article of this series analysis. I will be focusing on Mobogenie’s operation and examining what went wrong. [Note1]

Part 3: The Failure of Operation

Being Lean

“Lean” is a buzzword right now in Silicon Valley. It means having a framework and trying fast, learning fast and adapting fast.

One of the key problems I found Mobogenite teams(both in Headquarter and In US) had is : They don’t have a framework to think through problems, which is : coming up with hypothesis, building and measuring things to test those hypothesis, and learning from success and failures.

The teams approached to problems in a “you do what you just think of” way. Usually when they face a problem, they don’t breakdown it into small sub-set problems. Instead, they would think some solutions that they thought would solve the problem. After implementing those solutions, they don’t measure the resutls. As a result, they don’t learn meaningful things.

As explained previously, for your product to work, a set of hypothesis must be true. Without clearly stating the hypothesis they want to test, the team lost focus on why and what problem they were trying to solve. Instead, the goal simply became doing things.

Knowing Your Stage

At different stage, your business or product require different goals. Knowing the stage will help you focus on the most important things, help you layout your hypothesis in the right order, help you build things to test and measure those hypothesis, and in the end, help you achieve goals.

There are five stages from building a product to achieving market success. They are empathy, stickiness, virality , revenue, scale[1].

Looking at the growth curve( in part 1), I believe Mobogenie was still in between empathy and stickiness stages. Management didn’t think that way and push the product to virality and later on to revenue stage. As a result, the team should have focused on solving users problems and making the first adopters really love the product, but was pushed to grow the product when it didn’t achieve product/market fit, and pushed to ( also by Wall Street) generate revenue when the product hadn’t demonstrated value for users.

One key analysis we missed when operating mobogenie was to measure product/market fit in terms of retention. The ideas is brought from Alex Schultz [2]

Try to flatten the retention curve before driving growth. If the retention curve flatten, that means users would stay with this product. If not, they simply left.

Not understanding business stage, I think, is the biggest mistake Mobogenie team made.

Driving Growth

Finding metrics that matter

After launch, the team was stretched to drive growth. Three goals were set: daily content downloads, daily active users, and retention.

Daily content downloads and daily active users are not good indicators of your store health. Few people use app store every day. That number will fluctuate. However, it’s very likely that people user store once a week or several times a month. So a weekly acitve user or monthly active users are more relevant metrics. And the team should measure engagement rate = DAU/MAU.

The key to growth is retention. That number should definitely be monitored.

Growth driver

Content team was put in front to drive growth. It’s intuitive: if you have good content, people will download it. The more downlaods you have, the closer you get to your goals.

But we missed the more fundamental question -what on earth drives growth? why people comeback, why they choose us versus google play, why they stop using mobogenie after they login? what’s the magic moment for users when they open mobogenie? anything caught their attention and interests? what features people use most? who are power users? Do they download from searching or from exploring?

We knew if we can have them made the first download, they are more like to download again. but we don’t how we can make them stay longer. These questions never be answered. some were never be asked.

Content management

A lot of efforts are put into selecting what content to put, monitoring its performance and adjusting accordingly. If they see a game performs poorly, they move it to lower rank/poistion in store; if they see an app performs well, they move it up, and then move it down until they find a better performing app.

There are two key pieces in managing content: the content you choose, and the way you promote it. For Mobogenie, the problem is more about distribution than content selection. Put in an analogy, it’s like you have a small restaurant with a few tables in a big city like San Francisco or Beijing. You’re less well-known than those big restaurants. To grow your business, you have to come up with good food and let people know about it. Simply changing food menu everyday based on a small sample won’t work.

The right starting point for managing content is to think about “driving users in and back”. Who might like my stuff? Where to find them? Why they come and comeback? As an editor, why do you choose a specific content? what hypothesis you have regarding to users?

Test your hypothesis with users and build channel to reach them. Along with the content management, other marketing activities should be involved. SEO, PR, social media etc. One you have the content, help to educate the users, get awareness from users, make sure when users search, they find you. Ask users to spread the word, etc. This should form a virtuous circle to driving users.

Live operation

In July, Brazil World Cup became the hottest event around the world. Mobogenie made a right move: allow users to watch games through mobogenie and have a section for game stats.Retention went up dramatically.

This was the first time Mobogenie team realized that events, operations are important to drive user engagement. As a results, later on, more efforts were direct to design live operations. The team brainstormed NFL streaming, NBA streaming, or event live concerts in various city, such as San Jose.

One insights from the world cup event was: live broadcasting might work in US. You could pivot to that idea. But the team never explore that idea. We are so stick to the app store business.

Another mistake Mobogenie made : contrary to usually rule that make your product simple, intuitive and user friendly, Mobogenie did the opposite. One example is doing e-commerce, partnering with Newegg to promote deals in order to drive users and make revenue. As explained at the benning, we are not at a stage of generating revenue becuase we haven’t acheived important growth on product side. The focus should be on growing retention or monthly active users, not revenue! Second, people who come to Mobogenie for Newegg deals won’t be loyal to Mobogenie.We won’t have those “everyday low price” strategy to attract users in long term. Those users are easy to get, but easy to lose.

After e-commerce failed, the team’s next attempt was to design all kinds of events to retain users. This made the product more unintuitive to use and more complex. The team leveraged their strong gaming background and use a lot of gamification mechanics to design events, such as lucky draw, spin to win, an account system which asks you to do several tasks each day and get points, the more points you have, the more level you get. You can then uses the points to do other stuff. Honestly speaking, I got lost.

At this point, the product has more features than users could actually need. The product becomes more complex than users could intuitively understand. It’s a confusing app store with a lot of game features. As Jeff Bezos paraphrased Warren Buffett’s memorable saying: “You can hold a rock concert, and that’s OK. And you can hold a ballet, and that’s OK. Just don’t hold a rock concert and advertise it as a ballet.”

Monetization

When you don’t have the user base to generate revenue, you just don’t generate good revenue, no matter how hard you try. To meet the revenue goal, Ads were put in Mobogenie all over the place. Banner, Native Ads, etc. This caused serious user experience problems.

Data driven

Until recently, I am a big fan of data-driven approach.But as I think more about previous experience and listen more good talks, I now think the right approach should be “Hypothesis driven and data backed” .

Reid Hoffman has a great saying: “data only exists in the framework of a vision that you are building, a hypothesis of where you are moving to”

I think this is one of the weakness of whole teams. Without any vision or hypothesis driving them, look at the data all over the place or just make decision based on intuition, not data.One analyst in headquarter once released an analysis of Mobogenie’s performance in several countries. He put notes in the report that data of some countries are not accurate. Why would you put there if they are not accurate?

The back-end data tracking system doesn't work very well. Some data are missing, some data is not accurate. And it takes forever to fix that. If they could more be disciplined about running business, they can do better.

Data should exist in the framework of vision we are building, a hypothesis where we are moving to! We shouldn’t expect perfect data, but get accurate data to test our hypothesis and to make us honest and trasparant about whether or not we are growing the business.

In summary: Mobogenie totally failed on the operation side. A big big mistake is it does’t understand the business state it’s at. Without a good product, you cannot grow. Without growth, you cannot generate revenue. All the operation should make user experience simpler, more intuitive, not the more complex and hard to understand.

Appendix

Note 1: “Team” are mentioned a lot in this article. When I say team, I mean I am one of the members.

[1] Lean Analytics

[2]How to start a starup — growth. http://startupclass.samaltman.com/courses/lec06/