Discoverability: Working with Filter Bubbles

Often perceived as a hindrance to discoverability, filter bubbles can also be helpful allies for both creators and content producers. Stitch Media’s Evan Jones explains how Facebook can be used to maximize the success of projects and marketing campaigns.

In our recent article about audience-building strategies, experts agreed that, right from the start, the audience must be the core focus of your project. In order to do this, the web can become a very useful tool to test your concepts among highly targeted groups – all at little cost.

« This is how the rest of the Internet has been refined, Evan Jones, Creative Director and Producer at Stitch Media, explains. Facebook, Amazon, Netflix and Google are constantly creating variations of their interfaces and the content that they serve to see which is performing better. »

Narrowcasting vs. Broadcasting

If it’s easy to test content on the web today, be it a movie trailer or a website’s home page, it’s mainly because of the advent of narrowcasting, which has replaced traditional broadcasting methods on the web.

Every time someone views content online, a single connection is established between the source (i.e. the server that hosts said content) and the viewer. This unique connection makes it possible to determine what the viewer will see, and to modify the content in real time. It’s basically the same principle as the A/B tests commonly used in marketing to compare original and modified versions of various content.

There are many tools for A/B tests currently available. However, similar tests can also be conducted directly within a social network environment such as Facebook, where the ad auction system can be used to assess the popularity of specific content.

Using Social and Search Algorithms to Test an Idea

Whenever an ad appears on its website, Facebook determines the level of interest for the product, service or content by the targeted audience. The more interested these people are in what you are promoting, the more the social network will make it easy for them to see your ad.

As Facebook explains: “An ad that's high quality and very relevant can beat an ad that has a higher advertiser bid, but is lower quality and has less relevance.”

In other words, Facebook adjusts the cost of advertising based on the level of “risk” the ad represents. To assess this risk, it relies on each user’s past behaviour and current preferences, such as movies or artists they “like”.

This auction system allows the social network to find a proper balance between creating value for advertisers and providing relevant content for users. Evan Jones puts it this way: « Great content that fits with the interest of their users is deemed positive for Facebook. They’d be happy to receive it anyway. »

Creators and content producers can use these same algorithms for their own projects and marketing strategies in order to reach their audience.

Together Tales: Tweaking Content and Targeting the Right Audiences

How does all of this work? Evan Jones uses as an example Together Tales, a series of interactive books aimed at children that combine games and reading activities.

From the start, the team tested many elements of the project, such as its title and the target audience. For the title, Jones took a photo of his kids and created three different versions of a potential ad, each with its own title. All three were then published on Facebook.

The verdict was indisputable. The title Together Tales yielded a significantly greater engagement than the other two alternatives, Imaginary Friends and Live Action Books.

The team also made use of Facebook’s auction system in order to identify which audience was the most receptive to the project. They proceeded to select highly targeted users – Facebook makes it possible to identify such criteria as age, gender and areas of interest – toward whom it would then promote the project.

Having gathered all this data, the creators of Together Tales were able to determine that their project mostly attracted mothers who enjoy geocaching (a treasure hunt-type activity involving GPS technology), as well as fathers who like live-action role-playing games, or LARPs.

As mentioned earlier, the interests of each separate target audience can be measured based on the cost per thousand impressions (CPM) that Facebook charges to advertise on its platform. The lower the CPM, the greater the audience’s interest for your ad.

By identifying your target audience this way, you end up reducing your advertising costs, while at the same time getting a more accurate profile of your project’s first users. « This is your perfect audience, and we can define it by age, affinities and many other criteria to get a deeper picture, Evan Jones says. So when you create projects, you have this perfect audience member in your mind. »

To learn more about the development of Together Tales, you can watch this short presentation made by Evan Jones at the 2017 OMDC Digital Dialogue conference:

Filter Bubbles: Friend or Foe?

While Facebook’s ad auction system is indeed a powerful project development tool, it also tends to lock users in filter bubbles, that is, personalized environments where the content they receive is filtered based on their web-surfing habits. The current model makes it difficult to post an ad that targeted users might not enjoy or be attracted to.

As an example, Evan Jones mentions the dilemma faced by the creators of Road to Mercy, a documentary on medically assisted dying. One of their goals was to raise awareness amongst young Canadians.


However, Facebook’s algorithm determined that millennials had no interest in this topic. As a result, the cost to advertise the film skyrocketed, reaching somewhere between $4 and $5 per thousand impressions. Indeed, popping out of the filter bubble can be quite expensive…

Filter Bubbles: A Springboard to Success

While Jones believes this sort of situation is quite unfortunate, he also points out that current mechanisms also make it possible to go beyond filter bubbles.

Going back to Road to Mercy: the marketing costs can be significantly reduced by targeting adults whose parents suffer from a degenerative disease. Once you reach this audience that is already engaged with the subject of your project, the film gains in popularity. In turn, it becomes easier for you to appeal to an even larger audience.

Jones’s own Threads audience development service works in much the same way. « It sees where you’re going to get the first wins for your product, he explains. You can get these people now because they are already interested in your work. And then, as things start to snowball, we can start challenging people outside of these filter bubbles. »

Internet giants, such as Facebook, have already undertaken to find solutions to the problems brought about by algorithmic personalization. Until then, creators and content producers can use the current mechanisms to their advantage. « If you can develop that filtered audience, the audience that comes afterward sees the number behind the product and the social proof carries it forward, » Jones concludes.

Gaëlle Engelberts
A former television journalist at Radio-Canada, Gaëlle Engelberts is the editorial coordinator for CMF Trends. In parallel to her research and editing work with the Canada Media Fund’s Industry and Market Trends team, she recently completed a master’s degree in communication at Université du Québec à Montréal (UQAM) focused on new narrative forms in journalism and documentary, namely interactive webdocumentaries and serious games.
Read Bio
This site is registered on as a development site. Switch to a production site key to remove this banner.