Gamification

Gems, coins, XP points, milestones, badges and achievements—gamification is everywhere.

And for good reason: there are some major success stories like Duolingo, who've created sticky habits that appear tantalisingly easy to copy.

But when you try and implementing streaks and XP, they land flat.

This isn't a deep dive into the psychology of 🎮 Gamification.

Instead, it's a cheatsheet on practically building these reward mechanisms, and a breakdown of the mistakes that I see daily.

Notepads at the ready?

Fundamentally, it's easier to encourage (or discourage) a behaviour, if you can increase the reward (or cost) of that action.

For example, I try and eat fruit daily (not to brag), but if I was told that I've been eating fruit for 1,312 days in a row, I'd be considerably more motivated to pick up an apple tomorrow.

The simplest (and most common) implementation of this is a time-based streak, like we see with Duolingo here.

You're simply rewarded for a recurring input.

Or take the news aggregator app Artifact, who would show you your streak visually over a longer period of time.

These are simple to implement, but often aren't that effective.

Let me give you an example.

I send newsletters, but the cadence is less important to me than the quality.

So when an email provider (Beehiiv) displays streaks based on weekly publishing, it doesn't match what I want to do.

This would be better for me, if the streak was how many consecutive emails have had "open rates above 50%" or "unsubscribe rates below 1%".

In other words, my goal is quality, not quantity.

Frustratingly, even once you've found a high-leverage streak to measure, it can backfire over the medium-term.

The way to think about it, is that as the value of a streak increases, you want the cost of losing that streak to increase too.

That's why they work.

But as a result, if they do lose their streak, the user now has to incur or suffer the thing that they were trying to avoid.

i.e., the leverage has gone, and the user has failed.

This is an inherent problem with streaks.

See the problem?

Luckily there are a few counter-churn techniques that I've found to be effective.

For example, the house cleaning app Sweepy allows you to turn on 'vacation mode', pausing all streaks without punishment.

This is really important for streaks that will have natural pauses due to predictable events, like a holiday (e.g., newsletters, playing the guitar or jogging).

In a similar vein, Duolingo offer 'Streak freezes' to give you an automatic small buffer. 

i.e., "you messed up yesterday, but don't worry, you've not been punished yet".

This highlights the importance of really understanding what your users are trying to do.

For example, if you're building accounting software, having a streak for reconciling transactions kind of makes sense, as an idea.

But what happens when one transaction needs approval? Or requires speaking to a merchant, which takes 10 days? 

You'd lose the streak, and from that point on would be reluctant to invest time into it again.

How do you get someone to care about saving 59p? 

Well, the utility provider Octopus Energy simply renamed pennies as 'Octopoints' and provides ways for the user to earn them in the thousands. 

The flow goes like this: Octopus Energy will incentivise customers to submit their meter readings, by letting them spin a wheel of fortune.

This then rewards the user in Octopoints.

Most people will win 8 points, which is literally worth one penny. 

But 8 points feels better than one penny (i.e., it increases the potential leverage).

To be clear, points (or XP) doesn't need to have a monetary value.

The household chore app I mentioned earlier, Sweepy, will give you "coins" for completing daily tasks.

These coins objectively do nothing, but being able to measure and compare your input over a longer period of time is the value.

In the same way that streaks allow users to log and track habitual inputs, coins and XP allow you to track a wider range of repeat behaviours.

In other words, you can then incentivise specific or routine tasks, to build habits—people forget this step.

For instance, Duolingo has a hook to unlock a period of "double XP", by connecting with a friend.

By getting you to care about XP, they're now able to leverage something that they want you to do (connect with friends), via an action that benefits long term habit creation (using the app for another hour).

When done well, it feels like a cheat code.

Going from 10XP to 20XP feels meaningful, but the jump from 10,000 XP to 10,010 XP, much less so.

The initial delight of watching coins shower down the screen might help make those first few sessions feel more satisfying.

But they have diminishing returns—you can't just collect XP forever.

This is why video games reward users on a non-linear scale. 

As you get better, you're rewarded with more XP, and need more XP to level up, so you fight harder bosses, and get more XP for doing so.

Some services lean heavily into this with increasingly more difficult social leaderboards, 'double or nothing' wagers, XP boosts and challenges.

So let's talk about the alternatives.

Hyper-gamified XP isn't suitable for most businesses. 

For example, I don't want my accountants dishing out 'Accounto-Coins' every time I send an invoice.

For example, Google will show new users a 'to do list', which unlock badges for submitting reviews and community information.

There are a few nuances here.

Firstly, I might have 3 places that I'm happy reviewing, but I don't want to do 3 a week, forever.

Nor is earning say, 30XP, meaningful to me in the longer term, if I don't plan on doing this routinely.

Take Apple Fitness for example.

You might go for a run, and unlock a badge you didn't previously know about—this makes you feel good.

Or if you were really proactive, you might look-up a badge and work towards unlocking it.

Some achievements can be both, but usually they encourage different user journeys.

The difference is subtle, so here are a few examples:

The proactive goals (i.e., badges you want to unlock) are based on predictable inputs and outputs.

For example, Headway offer achievements for key milestones. 

I doubt many people know (or care) what these moments were in advance, but they're creating a sense of accomplishment.

This "I did the very first step" isn't a badge I'm going to want to look back on—the value is only in this very brief moment.

But, once someone has reached a milestone that they were aiming for, the big question is how you leverage that feeling to last more than a few seconds.

Let's cover that.

Badges (to record reaching a milestone) are a useful technique because they allow you to leverage the reward of an achievement, beyond the event.

They're usually more effective if:

And if you're going to use these as a way to work towards a specific behaviour, then you'll need:

An example of this would be how Skillshare dynamically suggests your next milestone, based on activities you've started.

This is effective because it keeps your next milestone within reach, and reminds you what you need to do.

Let's end on something a bit more fun, and unusual.

After completing lessons on Ahead, you'll trigger an 'emotional tap' mini game. You need to mash the screen quickly, hitting as many emotions and emojis as you can within a limited time frame.

This then compares your personal best score to the all time high score.

You complete an action (a lesson), and get rewarded with an attempt at the arcade machine.

Now compare that to other "mini games", like how Spotify allows you to play Snake with your playlists.

Or Instagram's hidden 'pong' mini game.

Spotify and Instagram's games are nice easter eggs, implemented well—but they're not rewarding.

Ahead's challenge is likely more effective because the input is controlled (to have an attempt you need to complete a lesson), and the competitive incentive is to match or beat the global score.

If any banks want to now implement Frogger, I'm in.