“It’s remarkable that even these short ads confirm the inequality we have seen in almost all our previous analyses – that even if women are seen on screen, it’s the men who get to have a voice.” – Matilda Kong, CEO.  

Read the full blog post here (German).

The post YouTube leaderboard: Diversity analysis of most watched ads during COVID-19<span class="badge-status" style="background:#0cbdb4">Collaboration with: Pinkstinks Germany</span>  appeared first on Ceretai.

Ceretai is the story of two people with a passion for social change, thoughtful use of artificial intelligence, and an act of courage to step outside of the box, who created the first diversity tech start-up and commenced on the journey to transform the media industry.

Matilda Kong and Lisa Hamberg had to meet twice in life to realise what they were capable of doing together. As kids, they attended the same school and played ice hockey together in the six-month-winter country of Sweden, but at some point, their ways had parted.

Before Ceretai was born, Matilda had a career in business development, while Lisa was leading projects as a software developer. However, in fall 2017, things started to shift for them.

As the #metoo campaign was unfolding in Sweden Matilda found herself deeply affected by the stories and wanting to act, but she wasn’t able to do so within the scope of her then employment. So she decided to quit her job.

“I strongly felt that our unawareness of bias and our extreme prejudice towards women and equality were the root problems behind the traumatising events that led up to the campaign,” Matilda says. “And I realised that there must be an arena where my passion for gender equality, my interest in technology, and my education – a combination of technology and business development – could be put together to solve this.”

Matilda had already found it frustratingly hard to find media content that didn’t contain sexism or stereotypical portrayal to some extent for years. As a consumer and audience she was starving for content that applied to her values. Matilda started to investigate if this was a common observation and came to the conclusion that people around her were also craving content that was aligned with their values.

Empowered by these learnings she started to search for people to help her take on the challenge. And that’s when Lisa and Matilda reconnected around the same idea by coincidence. With Matilda’s business sense and Lisa’s technical know-how, they were ready to turn the idea into reality.

Lisa, how did you become a co-founder of Ceretai?

Lisa: By that time I was working as an IT consultant, which I had been doing pretty much since I graduated from high school. I had been quite a career-focused person. But at one point when I was working with yet another consulting project, I realised that whatever I was doing there was not making any difference for anybody. And it didn’t make me any happier how much money I put in my pocket.

As a developer I was moving buttons around on web pages just to increase sales, I was involved in the materialisation and monetisation of a big company. I felt that doing a job that didn’t apply to my values was one of the reasons why I was not happy at that point in life.

When I found myself with this new mindset I went looking for something more value-based, and that’s when I saw a Facebook post from Matilda saying that she was looking for people: maybe co-founders, maybe part-time workers, freelancers… she seemed to be very open to anything. She explained her idea pretty briefly in this post, with only a couple of sentences: “I want to do something by using technology to improve diversity within media content”. It was a short mission line, but I immediately fell for it.

I consume a lot of media content. However, I hadn’t been one of those people who put too much thought into what they watch, as Matilda had, and it was an eye-opener. I realised I could use my knowledge of technology for this great cause.

How long did it take before you had the first product in place?

Matilda: We started working together at the beginning of 2018, and we had the first prototype that summer. Then we pivoted, and instead developed a prototype of a product that is pretty close to what we have today. It took us almost a year to release.

During this process, we came to the conclusion that we need to support the media industry in this transformation towards more diverse content with fewer stereotypes. We realised that even if we had the support of the audience, it was difficult for the media companies to adapt.

Lisa: In the beginning, we had a product that was more directed to consumers, but we realised that if we want to fulfill this mission we have to go directly to the media houses. And even though the actual product changed many times, what stayed untouched is the most important part of our work – our mission to change the media industry.

And even though the actual product changed many times, what stayed untouched is the most important part of our work – our mission to change the media industry.

You released a very successful campaign with BBC last year around gender diversity in the Game of Thrones series. How would you describe this experience?

Lisa: It was extremely encouraging and I still don’t think we know exactly how big of an impact it had. We still hear about it today, and it showed just how much power this type of data has in the news media. We might be aware that women are under-represented, but when we see exactly how little they actually get to speak (in Game of Thrones 75% of the speech was male) this does something with us. I feel that the audience also realised that they need to actively think about what they consume.

Why is the time right now?

Matilda: A window of opportunity has opened. There is more focus on gender equality and other types of diversity than ever, in part thanks to the Me too campaign. However, in terms of technology and how we can use this, the approach is very new and we are truly pioneering the field. We know we are first movers. We know we are doing the right things, but because we are early we need to take a big responsibility in educating the industry.

We know we are first movers. We know we are doing the right things, but because we are early we need to take a big responsibility in educating the industry.

Lisa: It has been proven to us along the way that this is the right time. We have financial support from Hamburg City in Germany and the Swedish innovation agency Vinnova, support that says “this is needed”. It is only a matter of time before initiatives in diversity tech accelerate and legislation on the topic comes.

What is the short term goal that the team strives to achieve right now?

Matilda: Bringing awareness of the necessity to measure! Everyone knows that we have to promote sustainability by measuring CO2 emissions, but we don’t measure diversity. So that is the number one priority: making people and the industry understand the need. And secondly, it’s extremely important that we establish a standard of what exactly to measure in order to break our unconscious biases. It is not enough to simply count how many men or women are represented – we need to go much deeper in what and how to measure to make a genuine impact.

When would you consider your mission completed?

Matilda: When every media company has a diversity partner. When they know that they have to report on diversity, just like they do sustainability reporting. When they know they have to monitor and follow up on the issue. And finally, when all audiences have access to a diverse media landscape and can decide on what to consume based on their personal values.

The post “We want to change the media industry” – Meet the founders of Ceretai appeared first on Ceretai.

Read more about the project, its goals, expected results and plan for implementation.

The post Ceretai receives 2m SEK (180k euros) from Vinnova<span class="badge-status" style="background:#0cbdb4">Granted by: Vinnova</span>  appeared first on Ceretai.

The Oscar’s last month once again reminded us about the fact that only one woman (Kathryn Bigelow) has ever won an Oscar for best director in the Academy Awards’ 92-year history, and that this year there was only one single black actor nominated for an award. But Swedish startup Ceretai, who define themselves as a “diversity tech startup”, think it’s time for a new perspective on the film industry; namely the perspective of the audience – focussing on what is actually shown to us and which immense effect this industry has on our beliefs and behaviours.

“In the bigger perspective, the lack of diversity in media and popular culture hinders the development towards a more equal society”, explains Matilda Kong, CEO and co-founder of Ceretai. “In the Western world we consume around 10 hours of media per day, of course it is going to affect us!”

Ceretai have built a software that can run through any kind of video and uncover some uncomfortable truths about what we are watching. In this study they have complemented their automated machine learning analysis with looking at Bechdel Test scores**, comparing plot summaries on IMDb, and studying movies’ financial success – all with the aim of defining what is actually “the norm”, and if this norm is worth breaking.

“WIFT International are happy to support the work of Ceretai. The data they produce tells us the inconvenient truth which is necessary to create the change we want,” says Helene Granqvist, President of WIFT International.

The first result from the study gives a thorn in the side to the truths about profit maximisation:

• Even though movies are still heavily male dominated both behind and in front of the camera**, no financial benefit of having a male lead actor can be seen when studying the 100 movies with most cinema visits in Sweden 2012-2018. If anything, movies including a female lead outperform those with a male lead in this category of highly successful movies.

“Although this specific dataset is too small to deliver statistical significance, the overall results follow the general trend: lack of equality in the film industry that has no feasible explanation”, says Angnis Schmidt-May, Head of Insights and data scientist at Ceretai.

A result that was however significant, and also historically significant since it has never been done before, was found when analysing another set of 100 movies produced in Sweden between 1970 and 2018. It is something Ceretai call “the smile factor”:

• Women smile on average 2.2 times more than men in movies (referred to as the smile factor).
• There are movies that have a smile factor of up to 7, but only 8% of movies have a smile factor below 1, meaning that men smile more than women.
• A movie has the highest probability of financial success if its smile factor is slightly below the average of 2.2.

“What this last point actually tells us is very interesting; why do we as audiences prefer women to smile twice as much as men?” asks Matilda Kong. “It’s a hen-and-egg question – is it because we are taught to like this from the movies we see, or do we make movies that way because audiences like it? I think it’s both, and the smile factor says something about our unconscious biases as well as our movie-making. We all need to ask ourselves if we are okay with these different expectations on men and women.”

And this also ties to the most interesting result of the study – a never-before revealed correlation between breaking norms and gaining financial success. By looking at the content of these 100 movies and labeling them as normative or norm-breaking according to criteria that can be read in this blog post, Ceretai could determine that in the past five years, the financial success of movies that break norms has increased dramatically. Norm-breaking movies now make up more than half of the top performing movies in Sweden every year.

Schmidt-May: “Again, this is not statistically proven, but it indicates that we are witnessing a great upswing in popularity of movies that break norms and defy stereotypes – something that the film industry has long proclaimed that the audiences do not care about. Also, as a side note, it’s pretty funny that almost a third of the movies from 1970 onward that were labeled as ‘norm-breaking’ are adaptations of Astrid Lindgren books. Imagine what it would look like without her.”

In order to enlarge the dataset and determine the norm more accurately, Ceretai also used IMDb and Bechdel Test data. The complete findings can be seen on their web page, but what could be concluded was that even in movies with a majority of women gender equality is seriously lacking, and that passing the Bechdel test is no guarantee for non-stereotypical portrayal. For example, when analysing 8000 plot summaries on IMDb, the most common keyword used to describe movies where the four main roles are female is “mother”.

Matilda Kong summarises: “The most important finding is that the audiences are ready for a change. They are hungry for more movies questioning our normative world, the traditional narratives and stereotypical portrayal. Even though this is of course not the only factor determining the success of a movie, mapping out and understanding what we are watching is immensely important, and Ceretai is determined to keep spreading this awareness – to the film and TV industries as well as to the audiences”.

Sources:

*Bechdel Test

** Gender distribution in international movies shown in Sweden 2012-2018 (statistics from the Swedish Film Institute)

The post Turning the spotlight on norms – and the success story of movies that break them<span class="badge-status" style="background:#0cbdb4">Press Release by: Ceretai</span>  appeared first on Ceretai.

International Women’s Day on March 8 often comes with a whole lot of “congratulations!” from people of all genders (although, in my very subjective opinion, it seems that among non-binary people this Freudian slip occurs to a lesser extent than with other genders). 

Among those who do not express a happy “congratulations!”, there is also a small but growing group who don’t consider gender inequalities to exist any more (at least not to the disadvantage of women), particularly in Scandinavia. Some of these instead mutter about the “need for” or “fairness in” having an International Women’s Day at all.

As you can imagine, we at Ceretai believe there is a need for this day. We also believe that a happy “congratulations!” may not be the ideal way of dealing with the results of our recent study that can be found further down in this post. 

In fact, in some feminist communities, saying “congratulations” to a woman on International Women’s Day is comparable to saying the same to a random Christian person on Good Friday. I would like to make an attempt at explaining why.

International Women’s Day was established in 1910 as a strategy to promote equal rights for women. It was founded by female revolutionary activists who fought, among other things, for the right to vote. During the entire 20th century it has been a day of protests, strikes, marches and the like, but in the past decades, it has turned more into a commercialised day of “congratulations” in our part of the world.

We at Ceretai think it should be possible to celebrate women’s achievements while also recognising that we still have a long way to go. We need to keep calling out inequalities and creating awareness around them – whether they are inequalities towards women, men, non-binaries, or any other of the seven grounds of discrimination defined in Swedish law* on which we base all of our data and analyses. 

We do this particularly in the film and TV industries. We recently completed a study of Swedish and international movies from 1970 onward, where we complemented our automated machine learning analysis with looking at Bechdel Test scores**, comparing plot summaries on IMDb, and studying movies’ financial success. 

And like so many times before, the results are unfortunately still disadvantageous towards women.

Results from the study

First, we looked at available statistics from the Swedish Film Institute between 2012 and 2018 (which is when they started collecting gender data on all movies shown in Sweden). Our first conclusions:

• Even though movies are still heavily male dominated both behind and in front of the camera***, no financial benefit of having a male lead actor can be seen when studying the 100 movies (Swedish and international) with most cinema visits in Sweden 2012-2018. 
• If anything, movies including a female lead outperform those with a male lead in this category of highly successful movies.

While this is of course a small dataset, it does tell us that the gut feeling of many within the film industry, that if you want to make a really successful movie you should cast Mr. so-and-so, should not be taken as a truth. 

When looking more closely at the actual content of these movies and their portrayal of men and women, we stumbled upon a very interesting find that we call the smile factor:

• Women smile on average 2.2 times more than men in movies. 
• There are movies that have a smile factor of up to 7, but only 8% of movies have a smile factor below 1, meaning that men smile more than women.
• A movie has the highest probability of financial success if its smile factor is slightly below the average of 2.2.

What this last point actually tells us is very interesting; why do we as audiences prefer women to smile twice as much as men? It’s a hen-and-egg question – is it because we are taught to like this from the movies we see, or do we make movies that way because audiences like it? Personally I think it’s both, and the smile factor says something about our unconscious biases as well as our movie-making. We all need to ask ourselves if we are okay with these different expectations on men and women.

We all need to ask ourselves if we are okay with these different expectations on men and women.

We then looked at 8000 movie plots at IMDb to see what these movies were about and how they were described. 695 of these movies had four men in the four main roles, and 213 had a top-four cast that was all female. Based on keyword analysis of these plots we could conclude that both movies with male dominance and movies with female dominance reinforce stereotypes.

Of course, since these are user-generated descriptions, we can ask ourselves whether this is rather a measure of how stereotypically the clientele of IMDb describes movies rather than how stereotypical the movies actually are. Either way, with IMDb being the world’s largest and most popular movie site, this is a problem.

We then looked at the Bechdel Test scores of these 8000 movies, to see if this could help us distinguish movies that defy stereotypes.

Our conclusion is that also in movies that pass the Bechdel test stereotypical portrayal is very common, and so the Bechdel test is not nearly enough for us to find movies that show true equality.

We have also put a lot of work into defining criteria for movies that break the norm. When comparing these criteria to the financial success of the movies, we actually do see some optimistic results! In the past five years, the financial success of movies that break norms has increased dramatically among the top performing movies in Sweden. From only 15% between 1999-2013, norm-breaking movies now make up more than half of the top performing movies in Sweden every year.

From only 15% between 1999-2013, norm-breaking movies now make up more than half of the top performing movies in Sweden every year.

Something worth mentioning about this part of the analysis is that almost a third of the movies that were labelled as norm-breaking in this dataset (1970-2018) are adaptations of Astrid Lindgren books. We have Pippi Longstocking, Ronia the Robber’s Daughter and the perhaps less internationally known but very loved Lotta on Troublemaker Street. Imagine what the world would have looked like without her!

So to conclude, the most important finding of this study is that the audiences are ready for a change. They are hungry for more movies questioning our normative world, the traditional narratives and stereotypical portrayal. Even though this is of course not the only factor determining the success of a movie, mapping out and understanding what we are watching is immensely important. And until we see real equality and diversity on screen, let’s keep International Women’s Day, but without congratulating women for being born with a vagina.

Because, don’t forget – “better” is great, but “better” doesn’t mean “good enough”.

Footnotes and sources

* The seven grounds of discrimination are: gender, gender identity, sexuality, age, ethnicity, religion, and disability.

** Bechdel Test

*** Gender distribution in international movies shown in Sweden 2012-2018 (statistics from the Swedish Film Institute)

The post Why a study of the film industry tells us we need to keep International Women’s Day appeared first on Ceretai.

*Vinnova is Sweden’s innovation authority.

The post Can the film industry become more equal and inclusive by analyzing the impact of norms on the industry’s business models?<span class="badge-status" style="background:#0cbdb4">Press release by: Vinnova</span>  appeared first on Ceretai.