App design best practices: What (not) to learn from Pokémon Go

November 6, 20175 Minute Read

Select article text to share directly to Twitter!


Remember Pokémon Go? You know, the game almost everyone was obsessed with for approximately two weeks?

If you haven’t thought about your Snorlax in a hot minute—or at all—you’re in the vast majority of users who downloaded within a week or two of release. Whether or not you think the app’s days are numbered, Pokémon Go can serve as a case study in app design best—and worst—practices.

Pokémon Go’s ongoing decline

By the end of 2016, a little less than six months after the July release date, daily unique players had dropped from nearly 45 million to 5 million. Bloomberg, amusingly, declared by late August 2016 that a one-month drop-off of 15 million users signaled a “number of important metrics have Pikachu’d.” Today, daily unique log-ins are still around 5 million.

While there are still a few million diehards playing around the globe, 40 million of us are over it. Stick a Moofork in it, we’re done. In the approximately 2-3 weeks before it began to decline, Pokémon Go was the most popular app of all time. By late July 2016, millions of us realized we were fully grown adults who were chasing Dratini through parking lots.

The interest faded, like it did for 23-year-old Brenda Wong of London, who said, “Maybe it’s sad that I now prioritize saving my battery over hatching an Ekans.”

5 reasons Pokémon Go may be destined for app death

Truth be told, there’s more to the app’s rapid crash-and-burn than its battery-draining tendencies. Here are a few other factors behind the sharp user decline:

1. Bugginess

Pokémon Go was unusable for extended periods of time in July 2016, maybe because user demand was 50 times higher than Niantic anticipated. To be fair, even Google Cloud director Luke Stone was shocked by the experience, stating he’d never seen anything close to the transactions-per-second and meteoric user adoption.

2. More bugginess

While Niantic and Google Cloud’s hard work to decrease overload-related bugs helped for a bit, Niantic continued to struggle even as usage rates declined sharply. In July 2017, the app didn’t support the demands of 20,000 attendees at Pokémon Go fest in Chicago. App performance was so poor, Niantic gave full refunds to attendees and offered $100 worth of in-game incentives.

3. Poor reward design

“Let’s get real for a minute” begins a year-old post by Reddit user Human-Chickenpede in the Pokémon Go sub. Among Chickenpede’s many complaints? The oft-cited “sharp curve in XP points” that makes it near-impossible for anyone but the most dedicated power users to level up at a point. The number of XP required to progress from level 21 to 22 is 50,000; twice what’s required to go from 20-21.

Poor reward design likely contributed significantly to the mass die-off phenomenon. No one wants to work harder for the same reward after weeks of dedication. Maybe that’s because it’s directly in contrast to the payoff principle of game theory.

4. Too little, too late

On December 23, 2016, Niantic announced their official holiday releases. Users gained access to free stuff daily, like single-use incubators and increased chances of hatching Gen 2 Pokémon, like Togepi, Pichu, and Igglybuff from eggs. Limited-edition Pikachu hung around a little longer, wearing some festive hats, and numbers of Gen 1 Pokémon increased during the first week of January 2017, giving you the chance to score a Charmander or Charizard (and more). And if you were worried about missing them, Niantic increased Lure Module time to a full hour—but the change only lasted one week.

Forbes Video Game contributor Paul Tassi was “pretty amused” by the event announcement. Like many others, he felt that incentives for play should have been released along with the massive Gen 2 update from weeks prior.

5. No meaningful player-to-player interaction

Despite the fact that playing Pokémon Go was, for many users, an inherently social experience, it lacked player-to-player interaction. Analysis of app store reviews from late July 2016 revealed a massive demand for the ability to battle players and other social features, including trading Pokémon and avatar customization.

In June 2016, Niantic finally added raid battles, but it wasn’t what users wanted, and people were unimpressed. When coupled with the fact that they were really hard—so hard that no one on Reddit managed to completely beat them—people were ticked. The raids “are a mess of bugs, and players deserve compensation” wrote author Daniel Friedman.

App design lessons from Niantic’s blunders

Industry-wide, the average amount of app users who’re still engaged 90 days after download is just 4 percent. So Pokémon Go’s crash-and-burn is perfectly average. Truth is, since the media’s still talking about Niantic, it’s just a really visible case study in app design best practices.

Here’s what IT pros can learn from the rise-and-fall of this app:

  1. Listen to your users: Pokémon Go players knew what they wanted and made it clear in many ways via social media and app store reviews. If Niantic had rolled out player-to-player interactions in August 2016 versus June 2017, they may not have lost so many players.

  2. Understand human behavior: Humans are motivated by reward and will be understandably ticked if it’s literally impossible to beat your new raid feature. Understanding behavior psychology and the specific needs of your user base is just good design thinking.

  3. Functionality is everything: Pokémon Go’s initial bugs were pretty understandable, considering the flood of downloads. The bugginess of the Gen 2 and 3 releases or at an event a year later with 20,000 attendees? They’re far less understandable to dedicated players.

In reality, Niantic didn’t really fail. The company succeeded at releasing the most popular app in history—it just had a hard time keeping its user base engaged. There’s a lot to learn from this: If you want to keep your users coming back for more, you need to keep giving them a reason to. And with app design best practices, that can be done.

Squash the app bugs before they get out of control, and listen to what your users want. They’ll guide you in the right direction.

  • Recommended for you
  • Recommended for You