In a seeming nanosecond, the mobile game Pokémon Go has shown us how:
- Demand and “going viral” are unpredictable
- Technology is not an objective; it continues to be an enabler
- Rapid technology scalability facilitates improved user experience (or lack thereof)
- Testing and automation need to be given due attention
On 6 July 2016, Pokémon Go – an augmented reality game launched by Nintendo and Niantic, a Google spinoff – was released in the Australia, New Zealand, and the U.S., and has already taken the Internet by storm. Its fan following and welcome is largely unprecedented, not only in the technology world, but also in the equity market. Nintendo, the part owner of the hugely popular Japanese franchise Pokémon, has already witnessed a sharp uptick in its stock price, soaring 23 percent in a single day, the company’s best one-day jump since the 1980s.
So what’s driving this popularity? First, people’s associations with Pokémon, which has a global audience across a large age group, and high merchandise sales. Second, the immersive experience of the game, which brings augmented reality into the mobile phone and engages users to take steps (literally) to interact with the Pokémon that appears on their smartphones.
Without putting you to sleep with talk of how smartphones are pervading the user environment and how they are redefining the digital landscape, here are the important take-aways:
Demand arises out of nowhere
When Niantic began developing the mobile game, it had no idea that it would be wildly downloaded on both iOS and Android devices. In fact, in terms of app downloads, it has already overtaken Tinder, a popular dating app, and is well on its way to beating Twitter! That’s commendable, given that Tinder was launched in 2012, Twitter in 2006, and Pokémon Go in July 2016!
If that’s not enough, Pokémon Go’s daily active users, a metric used to measure audience engagement, is touted to be much higher than even Twitter. The Internet is flooding with data points showing how the game is beating every standard metric used to measure success. In fact, its usage time is higher than that for the communication app Whatsapp, photo sharing app Instagram, and the rainbow filters of Snapchat.
Demand for software can spring up from anywhere without warning or notice. Even a simple game can become an overnight hit, grabbing as much media attention as Brexit. No amount of planning will help you be completely prepared for unexpected demand. When demand is so unpredictable, what can you, as technology practitioners and service providers, do?
Test, release, scale … repeat
Time and again, our conversations with enterprises and service providers have been limited to the context of technology without it being leveraged to achieve core business objectives. This overemphasis on technology, without understanding the true nature of its ability and purpose as an enabler, is misleading and, in a business environment, dangerous.
The developers of Pokémon Go had to learn this the hard way. They clearly had not anticipated such massive demand in only three countries. It reached such proportions that their international rollout plans had to be put on hold until they got their house back in order.
Users have reported cases of the game failing to load, servers going down, the game crashing mid-play, account integration with Google failing, identity management not working, camera function for the augmented reality piece backfiring, and the data management function erring. As a whole, the only thing consistent, other than the dramatic popularity of the app, is the failure of technology to scale up to meet the demands of its Poke-hungry consumers.
This highlights the importance of testing in a mobile context. When apps and games are swarming the app stores on a daily basis, there is a dire need to perform stringent levels of load testing to ensure that demand spikes can be handled. This obviously requires the application and the underlying infrastructure to be highly scalable in a short timeframe, especially when popularity (and hence the consumer experience) gets defined overnight. Testing and operations, in the context of application services, need to work in tandem to pre-empt and prevent such instances. Further, these incidents bring to the fore the need for infrastructure automation and rapid provisioning. Cloud-based models that lack rapid scalability (both up and down) simply do not serve the purpose.
Pokémon aficionados might forgive the app developers for their shortcomings for a while. But keep in mind … user forgiveness is fickle and short-lived.