Pokémon Go is a reminder that online gaming infrastructure is hard to get right

On July 6, 2016, Pokémon Go, a location-based augmented reality (AR) mobile game, was released in the US, Australia, and New Zealand. Created by Nintendo subsidiary the Pokémon Company and former Alphabet mobile game developer Niantic, the game has been highly successful so far, topping app store top-grossing and free charts in the US within 24 hours of release. However, it has highlighted the challenges of releasing games that are heavily reliant on online gaming infrastructure.

Nintendo has significantly underestimated consumer demand

Pokémon Go, which allows users to "see" and catch Pokémon in their own neighbourhood via AR, has become the most downloaded app in each territory it has been released in. The game is a breakthrough moment for Nintendo, which released its first mobile game in 2016. As a result, Nintendo has seen its market value increase by more than $7bn. The game is also the top-grossing app in the countries where it has been released, monetised via in-app payments for premium items.

Although the game's popularity has propelled it straight to the top of the mobile-gaming charts, its launch has been blighted by server problems. The game has attracted high levels of demand, causing authentication problems, game-play issues, and error messages asking users to come back later. It seems that Nintendo and Niantic significantly underestimated how popular the game would be at launch and their online gaming infrastructure hasn't been able to cope with the level of interest shown by fans.

It's debatable whether this would have happened if the planned phased roll-out hadn't been disrupted by game websites telling excited Pokémon fans how to access the game from outside the launch territories. However, this kind of behaviour among fans should be expected: Pokémon is a much-loved brand with many followers across the world, and Pokémon Go is arguably one of the best uses of location-based AR in the mobile-gaming space.

Of course, this raises questions about if and how companies can control the roll-out of services when national boundaries mean little in the face of the digitally savvy. Given that cloud-computing services allow for almost infinite flexibility in server demand, is it not better to overestimate consumer demand than underestimate it and generate negative press attention?