How Duolingo pushes users from freemium to premium

5 monetisation lessons from the world’s most downloaded education app

Duolingo has a huge 15.2 million reviews on the Play Store and 2 million on the App Store, with 4.5 and 4.7 stars respectively. It is by far the biggest learning app in the world.

But the story wasn’t always so dreamy.

Back in 2018, Duolingo was seeing single-digit daily active user growth year-over-year, according to ex-Chief Product Officer, Jorge Mazal. Growth was slow, people were getting pissed at losing their streaks, and it wasn’t looking good…

Fast forward to Duo’s 2021 IPO, and DAU had skyrocketed 4.5X since 2017. Shares closed 36% up on their first day of trading — valuing the company at a huge $5 billion.

Now in 2023, Duo’s growth spurt continues. The Co-Founder and CEO reported a 47% increase in revenue year over year, with $100.3 million in subscription purchases.

**Dollar signs in eyes**

A few weeks after reading Duo’s case study in Lenny’s newsletter, I was sitting next to my sister when she got a push notification from the app itself. She’s been learning French for 3 weeks (having not touched the language since school, desolée).

I wanted to see Duo’s monetisation and gamification in action — to untangle the web of monetisation meets gamification meets education. So, I stole her phone (with her permission), and had a play around.

What I found was 5 monetisation insights from the freemium user experience. We’ll go through them one by one, and see how Duolingo drives users to pay through UX flows, reverse trials, multiple entry points and leaning into the learner psychology of Self Determination Theory.

On y va 🥖

Lesson 1: Don’t just ask once. Ask many times.

Pretty much everything you do in the app links to premium. The app architecture is full: both in the header and main navigation.

This is possible the biggest, worst-kept secret of monetisation UX. Ask, ask and ask again.

If you only ask users to pay once, you’re leaving money on the table.

In Duo’s case, users are prompted to upgrade to SUPER:

• In the shop (twice)
• When viewing hearts
• On the homepage characters
• In the review tab (twice)
• After a lesson when watching an advertisement (more on this later)

There were so many entry points to the subscription it was hard to fit into one visual….

That’s at least seven touch points in one user session if someone is playing around for the first time. And if they come back to do another lesson? They’ll see a few more.

You don’t get to 4.2 million paying subscribers by asking once.

Key takeaway: ask users multiple times if they’d like to upgrade. But don’t litter the app with it, think about where it makes sense.

Next: make it contextual.

Lesson 2: Make your paywalls relevant

So there I was, tapping around in Duolingo. I’d seen the same paywall countless times. But then I noticed something… it wasn’t exactly the same paywall.

I feel like this purple colour is singed onto my eyeballs at this point.

Duolingo actually changes the first feature bullet point on the paywall to match where you entered the paywall from.

This is a nice touch, and relatively easy in terms of effort.

However, I didn’t notice it…

For me, this would be stronger if the massive purple paywall changed significantly from each area to cut through. I get they have strong premium branding, but users scan, so you need to avoid banner blindness by changing it up sometimes, like Second Nature do below.

Love this app. 10/10

You don’t necessarily have to add more friction to get users to a contextual paywall, but if you add enough motivation on there, it could be worth it.

Lesson 3: Reverse free trial for early power users

Now this was new to me in a B2C app.

3 days into my sister’s new user experience with Duo she’d smashed through a fair few lessons and been on the app each day. She was in the higher percentiles of users in terms of engagement (a power user I reckon).

Then, she was automatically given access to premium for three days. Without opting in. The normal Duolingo trial is 14 days long, and she still has access to it. So this was more like a taster for the unlikely payers.

This was an example of what’s called a reverse trial.

A reverse trial is when you automatically bump users into the full premium experience, then downgrade them when the time runs out. There’s no opt-in or opt-out.

Some users will think ‘I don’t need to pay, this should be a free service'. It’s those users who you may catch with a reverse trial. When I asked my sister, Izzy, how she felt on the free trial, she said:

I enjoyed being bumped up —I felt like I’d put some good work in and earned something. It gave me a full idea of all the app’s capabilities.

I found doing the mistakes super useful. And I did a lot of the challenges and levelling up bits that usually require you to spend your gems for them. Once I was back in freemium, I missed ad-free the most.

Seeing this gated feature almost made me want to join in…

Duo has gated arguably the most valuable learning tool in their arsenal: the mistake review. Via the free trial, users can try it out to see what they’re missing.

What’s interesting is that not all users are put on a reverse free trial. You have to be:

• Using the app regularly (so that you don’t miss it)
• Hitting engagement milestones that unlocks the trial (e.g. certain streak lengths).

It’s unclear the exact rules for Duolingo users, but what is clear is that you can continue to unlock this over time.

Anyone else been given 3 days of Duolingo plus? A Reddit thread

Does this feel dark, sneaky and sinister? Or does it feel rewarding, congratulatory and like Duo are just proud of their cool features?

I’m on the fence.

Lesson 4: Lean into the human psychology of learners

I love checking off lists. I love learning. And most of all, I love binge-learning. Give me a 6-week course? I’ll do it in a day. This is the sort of psychology Duolingo leans into.

When we complete tasks, we get a flood of the happy hormone dopamine. Ticking off a list feels good, feels satisfying (and its incredibly addictive for some).

Duolingo’s home screen acts not only as a course overview, but as a list to check off. Crucially, there’s two ‘complete’ modes:

1. Complete: character colour UI [e.g. purple] → feels like ‘well done you’
2. Legendary: gold, shiny UI → feels like ‘YOU’RE THE BEST’

What this screenshot doesn’t show is how nice the UI changes when you scroll between sections. Like melted butter.

The list format is a clear path to follow, reducing decision fatigue and motivating users to complete the list to get that dopamine hit as well as avoid the misery of unfinished tasks. The extra legendary level allows users to choose if they want to do an extra test to prove they’re the best.

These gamification tactics lean into Self Determination Theory. According to Tech Outcasts Founder, Ben Davies-Romano:

Competence is one of the three basic psychological needs required to be well and build intrinsic motivation. An extra hard challenge is perfect for fulfilling this for those who have already mastered more basic competences.

How is this relevant to monetisation? Well, you can upgrade to speed up the process of getting to legendary status. If you don't want to pay, you have to use your gems to do each lesson (and you soon run out).

Duolingo has worked out what drives their users. They’ve dug into their psychology. What drives their users is getting up to speed with a language as fast as possible.

So they’ve made SUPER for the impatient learners. They’re competitive, they want to learn fast and they can ultimately pay to do so.

Veeery interesting when you consider 17 of the top 20 universities in the USA (incl. Stanford) accept Duo’s online test as proof of English language proficiency for international students. Target market maybe?

Lesson 5: Freemium ad model

Ads have been around in Duolingo since before the subscription. Its a common way to monetise your freemium users: sell their attention to the highest bidder.

It also has the added effect of pushing people towards premium to escape the ads.

Duolingo show full screen ads after lessons for freemium users. There is often the CTA to ‘Go Ad Free’ prompting a contextual upsell screen on the first ad, followed by no CTA on subsequent ads.

The least Pret could have done was optimise their ad placement for story dimensions **facepalm**

The ads start with no exit, then a subtle ‘play’ icon appears top left. Sneaky, as we’d expect an X top right, following the mental model of exiting screens.

No word of a lie, I actually watched this add fully. Almost downloaded the game. Cuff me.

I didn’t initially tap the play icon, as I thought I’d be tapping the ad but then found it’s actually the exit. Doh.

I imagine this purposefully dark UX to increase the ad watch times.

Diversification of revenue streams (whether through brand partnerships or ads to any buyer) is a good idea to reduce business risk. There are ways to do this ethically and in the interest of the user — but you need to balance results for advertisers (and therefore the business) as well as users.

What is jarring about this is that it inhibits users to continuing to the 2nd, 3rd, 4th lesson as soon as possible. According to one Duolingo learner I spoke to :

When I was on my 3-day trial having no adverts was a relief, it made me do more lessons.

However, if the ad model allows Duo to “develop the best education in the world and make it universally available” [Duo’s mission] then I guess its OK…

To conclude

There’s a reason Duolingo has almost 5 million paying subscribers and continues to see this double-digit revenue growth year-over-year. They’re not afraid to push the boat out with their wider monetisation strategy and on-the-ground tactics.

Here’s a recap of our 5 key lessons on how Duolingo monetises freemium users:

1. Asks users multiple times: use a range of places across the architecture to trigger the paywall. You’d be silly to ask once.
2. Make the paywall relevant: Contextual messaging on the paywall will ensure the UX isn’t jarring, but make sure your screens are not too similar to each other to avoid banner blindness.
3. Reverse trial: Try bumping users up on a reverse trial to see if you can enlighten them with how great your paid experience is (for Duo, it’s the mistake review feature).
4. Work out what makes your users tick: Lean into user psychology. Do people want to learn fast? Well, ensure your premium service makes them feel like they’re jumping the queue.
5. The-feature-that-shall-not-be-named **ads**: If it keeps your business afloat so you can keep providing a free service for millions around the world, why not. Just ensure you think about ad UX to avoid blocking users from getting to the core value.

Anything else? Did you notice something I didn’t? Comment and let me know.

À bientôt!