Blog
Duolingo's owl, Aldi's carrot, GEICO's gecko: mascot lessons for small product brands
Duolingo’s owl, Aldi’s Kevin the Carrot and GEICO’s gecko have three things in common, and none of them is a big budget: a clear personality, a repeatable situation the character keeps returning to, and a look that never changes. Those three things scale down. A small product brand can’t buy a Christmas TV slot, but it can give its product a face, a running gag and a consistent voice, and use them in every video it posts.
Below is what each campaign actually measured, how strong that evidence is, and the part a one-product shop can copy. Every number links to the page we read it on. Where a figure comes from a pre-test panel or a trade article rather than a sales report, we say so, because those are different kinds of proof.
Duolingo’s Duo: a storyline, not a single ad
In February 2025 Duolingo ran “Dead Duo”, a social-first “whodunit” around the apparent death of its green owl. In its Q1 2025 shareholder letter, the company called it “one of our most successful marketing campaigns ever” and said it “spent practically nothing on it.” The letter says the campaign “generated 1.7 billion organic impressions, and drove a meaningful lift in new and resurrected users.” Daily active users that quarter were 46.6 million, up 49% year on year. A year later, the Q1 2026 letter described that quarter’s growth as “lapping 49% growth in Q1 2025 when we benefitted from our ‘Dead Duo’ campaign.”
How strong is that? It is the company’s own investor letter, so the figures are first-party. But Duolingo doesn’t put a separate number on the campaign’s share of that 49%, and user growth has many drivers. The honest reading: the company credits its mascot storyline with a real lift, and it was confident enough to say so to investors.
The owl kept working after that. In Germany, Duolingo turned Duo into serialized TikTok content, “Mit der Cult Mass”, and boosted the strongest posts with Spark Ads. TikTok’s Ad Awards case study reports 34.83 million impressions, 26.89 million people reached, a 12.63% engagement rate and 4.55 million interactions. That’s a platform case study, so it covers reach and engagement only, with no sales or install figures.
What transfers: Duo works because he has a personality (pushy, slightly unhinged) and because each post continues a story people already know. You don’t need to fake a death. You need a character whose next video people can guess the shape of, and still want to see.
Aldi’s Kevin the Carrot: come back every year
Aldi UK introduced Kevin, a CGI carrot, in its 2016 Christmas campaign. When Aldi brought him back the following year, Marketing Week reported (7 Nov 2017, read via an archive copy) that the 2016 campaign “led to a 15.1% sales uplift and helped the discounter to attract one million new shoppers.”
Treat that number with care. It is Aldi’s figure repeated by trade press, from a campaign that ran on TV, in cinemas, on social and in stores, at a time when Aldi UK was growing anyway. The ad wasn’t the only cause.
What is clear is that Aldi kept bringing Kevin back, and the later ads still test well. Zappi’s review of the 2024 Kevin ad scored it 88 on both sales and brand impact, with Love at 32% against a norm of 19% and Laughter at 14% against a norm of 7%. Those are panel pre-test scores, which predict how an ad is likely to perform. They are not measured sales.
What transfers: a seasonal return. Kevin is an event because he shows up at the same time every year with a new chapter. A small brand can do the same on a smaller calendar: the same character opens every Black Friday, every launch, every “back in stock” post. Each appearance reminds people of the last one.
GEICO’s Gecko: never change the face
The Gecko has fronted GEICO’s ads for over two decades, and that consistency still shows up in testing. System1’s July 2026 Ad of the Week gave the Gecko’s “Drain” spot 3.4 Stars for long-term potential and found that “across all three metrics, GEICO are landing in the top 7% of all US insurance ads.” An earlier System1 analysis of insurance ads found a different Gecko ad scored “3.8-Stars, well above the category average of 2.4-Stars,” and that “96% of viewers correctly identified the brand, largely thanks to the familiar Gecko.”
Again, these are pre-test scores from a research panel, not sales data. But the 96% is the useful number for a small brand. Most ads fail because people enjoy them and then can’t say who they were for. A familiar character solves exactly that problem.
What transfers: lock the look. Same colours, same face, same voice, same catchphrase, in every video. If you redesign your character every few posts, you throw away the recognition you just paid for. We covered the research behind this in why brand characters beat celebrities in ads.
One more: when a mascot gets a plot twist
Planters killed off Mr. Peanut before the 2020 Super Bowl, then brought him back as “Baby Nut.” The agency’s award entry makes big claims about the results, but those are the entrant’s own, unaudited numbers, so we wouldn’t build a plan on them. The lesson is the story structure. Duolingo used the same “the mascot is in trouble” twist five years later. A character people care about gives you plots a product shot alone never could.
What doesn’t scale down
Be honest about the gap. Duolingo has a huge user base that shares its content. Aldi and GEICO have TV budgets and years of airtime behind their characters. A small brand’s first mascot video won’t get those numbers, and no video maker can promise it will.
What does scale down is the mechanism: a character that people recognize, in a situation they already understand, showing up again and again. Recognition builds with repetition, so the payoff comes from the tenth video more than the first.
How to build a mascot for a one-product brand
You don’t need a separate cartoon animal. The easiest mascot for a small shop is often the product itself, with a face.
- Give it one personality trait. Grumpy, smug, overprotective, dramatic. One trait is easier to write and easier to remember than five.
- Tie the trait to the problem you solve. A coffee that’s “not a morning person” fits a coffee. A toothbrush that “keeps receipts” fits a two-minute brushing timer. The joke should point straight at the benefit.
- Find the repeatable situation. Duo nags. Kevin gets into Christmas scrapes. Pick a situation your character can return to in every video, with a new twist each time.
- Lock the look and voice. Write down the colours, the face, the voice and one catchphrase. Use them everywhere: ads, organic posts, your product page.
- Keep the claims true. The character can be ridiculous; the product claims can’t. Everything it says about the product should match your label (more on that in label-true product video scripts).
- Plan a series, not a one-off. Sketch five videos before you make the first: a launch, a “how it works,” a seasonal one, a customer-problem one and a comeback.
Here’s what that looks like in practice: a toothbrush with one trait (it counts how long you really brush) that complains straight at the camera and ends on the product.
Where Productoons fits
Productoons makes short cartoon videos where your product can become the character, in three styles: Funny, Talking Product and 3D Pixar-Style. Your real product photo goes into the video, and claims come from your label. Your finished video arrives within 24 hours of your brief, with one round of revisions included. Using the same character across several videos is how the lessons above add up. See how it works for what to send.
Want a video like this for YOUR product? See the options and order on our pricing page. And follow @productoons.co on Instagram for new examples.