Updated November 2026
The Duolingo owl shows that a mascot can carry a brand's voice for years, yet the public evidence on its commercial effect is thinner than the headlines suggest. This note keeps only figures we could read at the source, and looks at what a village fair, a sports club or a town hall can take from them.
The owl in 2026: audience figures you can check
The most recent count we could open is the tracker CreatorDB, refreshed on 5 November 2026. It lists about 18.4 million followers on TikTok, 5.39 million on Instagram and 7.06 million on YouTube, for a combined 30.9 million. The same page gives a TikTok engagement rate of 14.3 percent and an overall rate of 3.4 percent, against a category median of 1.5 percent.
These are third party estimates. They move every day, and a tracker does not publish how it computes engagement. Read them as an order of magnitude and not as an audit.
Two years of company reporting help set the scale. In its second quarter of 2023, Duolingo reported 21.4 million daily active users, up from 13.2 million a year earlier, in the Duolingo Q2 2023 press release, 2023. For the full year 2025, the company said in its Duolingo full year 2025 results, 2026 that it had passed 50 million daily active users and more than one billion dollars in bookings for the first time. Neither document credits the owl with any of this.
How a costume in an office became a character
The story starts small. According to NBC News, 2021, the idea came when staff returned to the office in September 2021. The mascot sat next to the marketing team, and a 23 year old social media manager, Zaria Parvez, began filming it. Videos such as the owl chasing colleagues who used an online translator, set to a popular soundtrack, passed 5.5 million views in the following weeks.
The numbers climbed fast. Marketplace, 2022 reports that the official account went from about 50,000 followers in September 2021 to more than 2 million in mid January 2022. In the same interview Parvez calls the content absolutely unhinged and says leadership stayed supportive even when much of the company did not understand it.
What was deliberately left out
The NBC report also records the limits the team set: no hateful content, and the owl is never shown with a knife or blood. If your association works with children or families, a written red line of this kind is worth copying before the first costume appears.
What research says about characters on social networks
Academic work is more cautious than the trade press. Kim, Sung and Moon, 2020 ran a longitudinal experiment on Facebook fan pages. Brands given human traits earned more favourable attitudes and more likes, and the effect passed through a sense of social presence. The relationship benefits also held up when news of a brand transgression came out.
That result concerns fan pages and experimental brands, and a mascot at a street festival is a different case. The abstract we read gives no sample size. It supports the idea that a character helps people feel they are dealing with someone, and it does not show that any particular costume will sell anything.
Limits and method
We opened each page cited here. Two widely quoted articles on the owl, from Fast Company and The Drum, returned an access error when we tried them, so we use none of their figures. Several web pages quote a rise of 62 percent in active users or 1.7 billion reactions after the owl campaigns. We found no primary document linking those numbers to the mascot. The 62 percent figure does appear in the Q2 2023 release, but there it is the yearly growth of daily users for the whole company, with no mention of the owl.
Follower counts also differ by source and date. Digiday, 2021 gave 1.9 million TikTok followers in December 2021, which fits the Marketplace figure a month later, but the 2026 figures come from a different tool. None of the studies measures sales for a club, a charity or a municipality.
- A tracker figure is an estimate and not an audit.
- A company growth rate does not show that a mascot caused it.
- An experiment on fan pages is not a field test of a costume.
What this means for a club, an association or a town hall
The useful lesson concerns consistency and place, and going viral has little to do with it. The Duolingo owl sits inside a company with a large digital audience. A French sports club, a parish fair or a municipal event has a different asset: people who already meet in the same square every year.
- Give the character one trait and keep it. A mascot that is the cheerful host of the village fete should not turn sarcastic at the Christmas market.
- Write the red lines first: where the costume goes, who may wear it, what it never does, especially around children.
- Plan the wearer along with the suit. Short shifts, a helper to guide the wearer in crowds and a rest area matter more than the fur.
- Use the character in the printed programme, on the poster and on the club social accounts, so that people recognise it before they see it.
A custom mascot made for your own association also avoids the generic costume that looks like every other one at the regional gathering. If you would like to explore that, ask us for a free quote. We make no claim about attendance or membership, because the sources above do not support one.
Sources
- CreatorDB
- NBC News, 2021
- Marketplace, 2022
- Kim, Sung and Moon, 2020
- Duolingo full year 2025 results, 2026
- Digiday, 2021
Frequently asked questions
How many followers does the Duolingo owl have?
According to the tracker CreatorDB, refreshed on 5 November 2026, about 18.4 million on TikTok, 5.39 million on Instagram and 7.06 million on YouTube. These are estimates and change daily.
Did the owl raise Duolingo's number of users?
The company reports strong user growth, but the documents we read do not attribute it to the owl. We found no primary source for the percentages quoted in some articles.
Can a small association copy this approach?
The principle of a consistent character with written boundaries transfers. The audience size and the results do not, and no source we read measures results for small organisations.
Does a mascot make people like a brand more?
An experiment published in 2020 on Facebook fan pages found more favourable attitudes towards brands with human traits. It was not a test of costumes at live events.
Why are some well known articles not cited here?
Two of them could not be opened when we checked, and we only report figures we have read ourselves.

