Gamification Examples: Duolingo, Starbucks, Strava, Sephora

You've used at least three of these apps without thinking about it. Duolingo. Starbucks. Strava. Sephora. Each one quietly trained you to come back. Not with ads. With a small game baked into the product. This post pulls apart how each one does it, in plain words, so you can borrow the ideas for your own client.

What “good gamification” actually looks like

Most articles treat these features like magic. Add a streak. Add points. Add a leaderboard. Done. But that's not how Duolingo or Starbucks got here. Each brand built one loop that matches their product. The parts you can borrow are the structural ones, not the icons.

Three things show up every single time the loop works. First, the thing being rewarded is the thing that makes the product useful. Duolingo's streak only counts when you finish a lesson. Not when you tap a notification. Not when you open the app. If you reward a fake action, users figure it out in a week and stop caring.

Second, the reward has two uses. A real one and a status one. Stars buy actual drinks. Sephora tiers unlock real perks. But they also make the user feel something. That feeling part keeps the loop alive in dry months. Third, the progress is visible. A streak count. A points balance. A leaderboard rank. A tier badge. If the user has to dig through menus to find it, they forget it. And forgotten progress is dead progress.

The four teardowns below each show a brand solving a different version of the same puzzle. By the end you'll have four patterns to compare against your own product. The question isn't which brand to copy. It's which pattern fits the way your user actually behaves.

Duolingo: how a streak became an identity

Ever caught yourself opening Duolingo at 11:45 PM just to keep the streak alive? That's not random. That's the whole design.

The streak is the most copied feature in consumer gamification. It's also the one most people copy wrong. People see the flame icon, the number next to it, the push notifications, and think they get it. They don't. The flame is the cosmetic part. The reason it works is what's underneath.

The streak only counts days you actually finished a lesson. Not opening the app. Not tapping a notification. A real lesson. So the number going up means you actually practiced Spanish. The thing the product rewards is the same thing the product is for. That's rare.

Most apps that copy streaks count something the brand wants to track. Daily opens. Returns. The user reads that as pestering, because the metric is for the company, not for them. Think about cult.fit's workout streak. It works because finishing a workout is the actual product. Now imagine a banking app counting how many days in a row you opened it. Nobody cares.

The second smart move is that Duolingo makes the streak loud. The number is the first thing you see when you open the app. It sits on your share card. Your friends can see it. A 400-day streak isn't a number anymore. It's a statement. Once the streak gets long, breaking it stops feeling like a small failure. It feels like a public loss. And people hate losing things they own.

The third move is the one most brands skip. The Streak Freeze. Early versions had no safety net. Miss one day, lose everything. People came back, but anyone with a long streak was rage-quitting after one bad day. The freeze is a free skip that absorbs a missed day. On paper, this should reduce engagement. In practice it's now one of the most used features. The moment users know a single miss won't wipe their progress, they stop being scared of the streak. They start defending it.

Three things make the streak land:

1. It rewards a real action. The number goes up because the user actually practiced, not because they tapped a button. So the count means something.

2. It's loud. The number is everywhere. The home screen. The profile. The share card. You can't ignore it.

3. It forgives one bad day. Without the freeze, the streak is stress. With it, the streak is something you protect.

Brands copying Duolingo usually mess up in three ways. They add a streak for something that doesn't deserve a daily habit, like a B2B reporting tool. Daily streaks for weekly behaviour just annoy people. They make the punishment too harsh, with no freeze and no recovery. The streak becomes a stress thing, not a fun thing. Or they hide the count three menus deep. Now there's no identity left to defend.

The lesson is simple. Streaks work when the user is supposed to do this thing daily, when the count goes up only for the real action, when the number is visible, and when the system gives you one free skip a week. Take any of those out and the streak becomes pressure with no point.

Starbucks Rewards: how Stars became a real currency

Most points programs die for one reason. The user has no idea what their points are worth. You earn 412 points. So what? 412 of what? You stop caring within a month. The Cred coin had a similar feel when it launched. Users got coins after every payment, but nobody knew what 50,000 coins could actually do. The early novelty wore off fast.

Starbucks fixed this by making Stars behave like cash. You earn 2 Stars per dollar. The reward menu is published. It doesn't change. The exchange rate is visible on every screen. Stars are not vanity points. They're a parallel currency the user can read at a glance.

The other thing they got right? Stars are earned on top of something the user was already doing. They were going to buy coffee anyway. Stars don't ask them to change behaviour. They just reward what's already happening. That's the lowest-friction onboarding any loyalty program can have. No new habit to build.

The tier multiplier is the second smart bit. Gold members keep that 2x rate as their minimum and pick up bonus offers regularly. Moving up the tier isn't just a name change. It's better economics. Gold users see their balance grow faster on their actual receipts. That's a benefit people notice.

The third piece is the spending side. Starbucks publishes five clear unlocks. The user always knows what their balance is worth. 25 Stars gets a syrup upgrade. 100 gets a brewed coffee. 200 gets a handcrafted drink. The first reward is close enough that even casual buyers can reach it. So the program feels alive, not theoretical.

The takeaway: a points currency only works if three things are true. The user can convert points to value in their head. They earn at the speed they naturally buy. And the spending menu is public, with rewards they can actually reach. Most brands fail the first test. They launch points that aren't tied to anything specific. They change the exchange rate later, which feels like theft. They set harsh expiry policies that make people stop trusting the whole thing.

Starbucks does use expiry too, but they do it gently. Stars expire after a stretch of inactivity. The user gets warnings well before. The framing is “come use your reward” not “we're taking your reward back.” That tone difference matters.

Strava: how leaderboards turn solo runs into something social

Most leaderboards kill motivation, even when the team building them doesn't realise it. The default move is to show the global top 10 or all-time top 50. For most users, that view is basically a sign saying “you'll never catch up.” They look once, decide the contest isn't for them, and stop playing. The leaderboard does the opposite of its job.

Strava's breakthrough was splitting the leaderboard into segments. Every street, every climb, every running loop has its own leaderboard for the people who tracked it. You're not competing with elite athletes in Kenya. You're competing with the people who ran your own street last weekend. The leaderboard becomes local. The ranks feel reachable. The contest is one a normal person can win. The insight is that competition only motivates people who think they actually have a shot.

On top of the segment leaderboard, Strava added two more social bits. Kudos work like Instagram likes for runs and rides. Every workout gets a small social pat on the back, even when you didn't finish anywhere near the top. Clubs are mini-communities for a city, a brand, or a friend group. They give users their own crowd to post into and race against. The annual year-in-review shows your total distance, time, longest ride, biggest climb, all packaged into one shareable card. Put together, these turn a solo workout into a social event. That's why people keep tracking on days they almost weren't going to run.

Brands that copy Strava usually mess up on one decision. They launch a single global leaderboard, no cohorts, no reset, no segments. The top 10 ends up locked by a few power users who never move. Everyone else looks at it once and leaves.

The fix is structural, not pretty UI. Reset the leaderboard often. Weekly or monthly works for most brands. Split users into cohorts by city, region, or tier so they compete with their peers. Show the user's own bracket, the few above and below, not just the absolute top. None of these are fancy features. They're the difference between a leaderboard that motivates the middle 80% and one that only motivates the top 5%.

The lesson is that competition only works when the contest feels fair, reachable, and resets often enough that new users can win. Strava is famous for leaderboards, but the actual brilliance is the framing. Local first. Global second.

Sephora Beauty Insider: how tiers built a moat

Sephora's Beauty Insider is the cleanest tier example in retail. Three named tiers with clean thresholds. Insider is free. VIB unlocks at $350 of yearly spend. Rouge unlocks at $1,000. Each tier carries its own identity and a different set of perks. Not just a slightly bigger discount. Think about Nykaa Beauty Insider or Myntra Insider, both pulling the same trick. The tier name itself becomes part of the brand experience.

The thresholds matter. They're high enough that the user feels they earned the tier. Low enough that a meaningful chunk of regular customers can reach VIB. A smaller, visible group reach Rouge. That balance is the whole game. If everyone reaches the top tier, it loses meaning. If nobody does, the ladder feels fake.

The perks design is where most retailers go wrong and Sephora gets it right. A weak tier program just gives the top tier a slightly bigger discount and calls it done. The user gets a 2% edge and doesn't feel special. A strong tier program gives the top tier things that change what the user can do, not just what they pay. Rouge members get free Beauty Studio sessions, early access to launches, free same-day shipping. The perks are about status, not just savings.

The third decision is showing the badge. The user's tier shows up in the app, on the website, on their profile, on receipts. Rouge members can see they're Rouge wherever they go. The badge is part of the user's identity inside the brand, not buried in a benefits PDF. That visibility is what makes the tier feel real. And feeling real is what makes losing it sting.

A user who's been Rouge for a year and is about to drop back to VIB feels that loss long before it happens. That fear is what drives one last spending push in December. It's also why Rouge members stick with Sephora even when a competitor offers a better price on a single product. The tier is doing more work than any individual discount could.

Tier programs usually fail in three ways. The perks feel weak, so the top tier reads as a marketing label. The thresholds are too high, so nobody bothers chasing them. Or there's no badge, so the tier just sits in a database and never shows up on a screen. Fix all three and you've got a moat your competitors can't easily copy. Skip any one and you've just built a discount scheme with fancy names.

The lesson: tiers create durable behaviour because losing a tier hurts more than gaining one ever felt good. The badge matters as much as the perks. A tier the user can't see is a tier the user doesn't care about.

What these four have in common

Pull the four loops apart and the same five traits show up every time. The mechanic rewards a real behaviour the product depends on, not a vanity action. There's a way to recover after a bad week, so users don't fall off a cliff. Progress is visible to the user, often to their friends too. The system builds up over time, so a year-three user has way more invested than a week-one user. And none of them work on rewards alone. There's always an identity layer that turns the system from a benefits page into something the user describes themselves with.

The five traits feed each other. A loop tied to the real product naturally builds up, because the user is doing the thing that matters. A building loop creates the conditions for an identity layer, because long-term users start defining themselves by their position in it. An identity layer makes visibility valuable, because now the user wants the badge to show. Visibility creates the fear-of-loss that forgiveness has to absorb. Take one piece out and the rest weaken. That's why slapping a flame icon on a B2B SaaS doesn't give you Duolingo's retention. The icon was the last decision, not the first.

These four brands also share something quieter. Each one treats the gamification layer like a product, not a campaign. The streak, the Stars, the segments, the tiers aren't seasonal pushes. They are systems the team tunes over years. Earn rates shift. Thresholds get rebalanced. Freeze logic gets refined. The teams running these aren't running a launch and walking away. They run a maintenance schedule, watch for abuse, listen to complaints, and adjust the numbers. The return on the loop is mostly a return on that discipline.

What to steal, what to leave alone

Steal the patterns. A streak shape for daily behaviour. A clear-rate currency for transactional brands. A neighbourhood leaderboard for audiences that enjoy a bit of friendly competition. A status tier for retention plays. These four cover almost every situation you'll face. Picking the right one for the context is the biggest decision you'll make. It has nothing to do with copying the surface of any one brand.

Don't steal the surface mechanic one-to-one. The most common failure is taking a famous feature and dropping it into a product where the user behaviour doesn't earn it. A daily streak on a tool people use weekly. A leaderboard for an audience that finds competition stressful. A points currency with no published menu. A tier program with no real perks at the top. Users sense the mismatch within days.

Here's the short version. Look at your user's most valuable behaviour. If it should happen daily, build a streak with a freeze. If it's transactional, build a currency with a published menu. If it's competitive and the audience enjoys a contest, build a segmented leaderboard with the user's own bracket on screen. If it's long-term and status matters more than savings, build named tiers with real perks and a visible badge. Pick one. Ship it well. Tune it for a year. Then think about adding another. The four brands above didn't launch all of this on day one. They built one loop until it worked, then layered the next.

Where Bricqs fits

Each of the four patterns above (streaks, currencies, leaderboards, tiers) is a ready piece inside the Bricqs engagement and retention platform. Brands plug Bricqs in to build their own loop without writing all the plumbing from scratch. You get the points ledger, the tier engine, the streak counter, and the segmented leaderboard as parts you can assemble. The work in front of you is the work that actually matters. Picking the right pattern for your behaviour. Tuning the numbers. Writing the maintenance plan. The mechanics are the easy part. The fit is the hard part. To go deeper, the points systems guide, the reward systems guide, and the sports engagement guide walk through the design choices for specific situations.

Related reading: the points systems guide, the streak psychology and design deep dive, and the designing progression systems walkthrough cover the next layer of detail behind each of these patterns.