3 Questions for Cheryl Hori of Pacific Campaign House


Originally published on Campaigns & Elections
Dark close-up photo with an iphone in someone's hands with several social media apps in focus such as Instagram, Facebook, and Twitter

Cheryl Hori is the founder of Pacific Campaign House, a full-service democratic digital agency based in California. We discussed how targeting restrictions by platforms like Facebook make it more difficult to reach diverse audiences, what digital actions ‘23 campaigns should be prioritizing right now, and the threat AI could pose to campaigns.

C&E: Now that we’ve seen some platform restrictions in action for a while now, how have you been adjusting and how would you advise other folks to adjust moving forward so campaigns can still effectively reach their audiences?

Hori: The way that targeting tends to work for digital advertising is the assumption that everyone is white and straight and cisgendered unless you indicate otherwise to that platform. So when we take out ads, platforms do make assumptions about certain people and their behaviors. Where this becomes really tricky is when there are different community groups or different organizations that are specifically looking to target and talk to members of their own community that may not be white or straight or cisgendered. When we look at platforms like Google where they’ve stripped out all of the targeting for political except for age, gender, and zip code, it really has a negative impact on some of these groups that maybe have smaller budgets or are specifically trying to target their communities. 

One of the things that we’ve done is leaned more into behavioral targeting. So instead of relying on the platform itself, we’ve looked at things like where people live or what types of things they’re purchasing. One of the things most consumers don’t know is just how much of their data is available for sale.

So, for example, when you go to an ATM and they ask you what language you want to do your transaction in, that specific language data is something that could be bought by digital advertisers to then target you and make assumptions about who you are. So we’re really leaning into some of that behavioral targeting. What are certain things people are buying that maybe indicate they belong to one community or another? What different places are people frequenting that may indicate that they belong to one community over another? 

C&E: If you’re advising a ’23 or ’24 campaign, what’s on the must-do list when it comes to digital right now? What should they be thinking about? 

Hori: First and foremost, make sure you have a video team or the ability to produce video. TikTok took the world by storm with their video content, and we’re seeing major platforms like Instagram try to follow suit. They’re now prioritizing video over anything else. So whether you have budget for paid media or you’re just following up on organic, having video content is going to make sure that you’re seen. Statics just aren’t going to cut it anymore. We’re seeing more and more creative breaking through regardless of whether or not it’s paid or organic — smart, clever and well-produced creative. And when I say well-produced, I don’t necessarily mean super polished. It can look like user generated content. We’ve seen that really take off.

And then finally, I think really looking to young people and what they’re doing. We’ve seen a couple of ads this cycle that have done that really well where they’ve sort of captured that Gen Z audience and attention. But we’ve also seen a couple of folks try to do that and fall flat. There’s a Reddit subgroup called Hello, Fellow Kids. It’s all advertisements of clearly people who are not of the generation trying to talk to them. So really being intentional with how we are thinking about creative, especially when we’re talking to young people. 

C&E: What role do you see for artificial intelligence in the political industry over the next decade? 

Hori: One of the things I think we will likely see is a continued spread of misinformation that’s generated by AI. And we’ve also seen some fun ways that deepfakes have been used. So, for example, there’s a TikTok account where it’s a guy who has used AI to put Tom Cruise’s face over his, and he does a bunch of really silly things that Tom Cruise would never do. And it’s fun and it’s very light hearted. But you can also see that same technology being used in a way where we see political candidates out there saying things they never said or doing things that maybe they never did. And it’s a lot harder to say that’s not real or that’s misinformation when you have a video of it in front of you. 

In terms of practitioners and for folks like myself on the agency side of things, I think there is a benefit that we’ll see with AI in terms of targeting when we’re thinking about how the voter file is maybe not always the most accurate thing, regardless of how much you’re paying your data firm. Using AI to sort of craft a better profile or a better model of who your target audience is — we’re starting to see that more and more now.