Q & A

What is the research you conducted?

We know that young people around the world are interacting with extremist content online, and we wanted to conduct an experiment to see if we could curate pre-existing content online that could serve as a counter-weight against some of ISIS’s propaganda and recruiting material. This pilot project was an attempt to curate existing content and serve it to people who were searching for extremist content online.

How is this different from previous attempts?

Many previous efforts to push back against extremist propaganda involve creating new content -- writing, videos, etc -- to dispel extremist narratives. Our method uses pre-existing content, including content that wasn’t created expressly for the purpose of counter-messaging, to push back against extremist propaganda using authentic content that effectively reaches its target audience.

How does the method work?

We begin by working with partners to curate pre-existing content online that in some way undermines extremist group’s recruiting efforts and propaganda. In this case, we were focused on ISIS and the ways that they spread their messages online and recruit new members. As you might expect, there is already a huge amount of video and written content online that directly confronts many of ISIS’s claims (for example, the real experience of living in the caliphate). We work with partners to create curated playlists and collections of the best, authentic content that challenges ISIS’s narrative.


Once we have this content in place, we create targeted advertising campaigns that serve this content directly to people who are searching for information about ISIS and the caliphate, as well as people who are trying to view extremist content online. We serve these ads the same way that businesses have been doing for years—we serve advertising against certain keywords that people tend to use when they’re looking for jihadist content online.

What did you learn from the research?

Most importantly, we learned that the most effective way to reach people who are interested in extremist content online is to use content that already exists online, especially content that presents alternative messages or perspectives on the issues that individuals who seek extremist content care about—the reality of living in the caliphate, for example.

Was the method effective?

Yes. We were pleased by how many people we reached, and how deeply engaged people were with the content. An estimated 320,906 unique users clicked on the ReDirect Method's pilot ads. To evaluate how engaging our ads were for these users, we compared the "click through rate" (the ratio of people who click on the ad vs. people who see the ad) with the CTR of all ads that ran against similar search terms in the 12 months prior to our Pilot launch. From this comparison, we learn that on average our Arabic ads did 79% better in CTR than others similar ads in this space, and that our English ads did 76% better.

How will you apply your insights?

We’re starting by sharing our insights with the community of people and organizations who are working on stemming the tide of violent extremism. We plan to continue our research exploring new ways to undermine extremist recruiting and propaganda online.

So you’re tracking people’s behavior online? Isn’t that what law enforcement is supposed to do?

No. We are using the same tactics that businesses use to advertise. The term “targeted advertising” just refers to serving content against certain keywords—the same way a shoe company might serve ads against the keywords “running” or “jogging”—that indicate an interest in extremist content. While this method fuels a trillion-dollar ecommerce business, it’s hardly been used as a tool to provide alternative messages to people who are looking for extremist content online.

Aren’t you discriminating against people with a certain opinion? Shouldn’t people be able to access content without Google interfering?

This project was incubated by Jigsaw, a think tank and incubator within Alphabet, using Google products (YouTube and AdWords). This research doesn’t discriminate in any way. We aren’t blocking anyone’s access to anything or preventing anyone from expressing an opinion. All we’re doing is trying to identify the keywords that people use to search for extremist content online and serving our curated content against those keywords. This is the same tactic that businesses use to advertise to consumers.

How will you imagine other organizations using this method?

We hope that people will be inspired by this research and realize that all sorts of non-governmental organizations can contribute to the struggle against violent extremism.

Engage & Collaborate

Recruitment narratives are tailored to specific audiences and evolve quickly. We’re interested in other approaches and findings into the main drivers of recruitment videos, and the counter-points that can be found in videos uploaded by others online. We’re also eager to hear about relevant keywords and insights that would help others replicate the Method. If you conduct relevant research, manage YouTube channels and AdWords campaigns, want to fund other iterations of the method, or can contribute in other ways, please get in touch.

Connect with us Contribute