Read the blog post that goes with this video at https://jonloomer.com/broad
Broad targeting can be effective, especially for Purchase campaigns. But it is bound to fail for any type of top-of-the-funnel optimization. You'll probably get what appear to be good results, but they'll be low quality.
The reason for that is that the algorithm is literal. The primary focus of Meta's optimization is satisfying your performance goal. If you say you want link clicks, the algorithm will find you plenty of link clicks by going after people who click everything, accidental clicks due to placement, and even bots or click fraud before they are detected.
This is already an issue normally, but broad targeting makes it worse. Many advertisers will tell you it's a complete waste of money to run traffic campaigns, and I'm one of them. But at least if you can add targeting guardrails, there's a possibility that you could reach the right people (in theory). Allow Meta to go after clicks that don't satisfy your targeting, and it opens up a big potential issue.
You could just not ever run Traffic campaigns or other top-of-the-funnel optimization, especially when using broad targeting. But I'm stubborn and I want there to be a solution.
I've experimented with custom events for driving traffic before. In this video, I'm doing it again, but this time using broad targeting.
Here's the setup...
1. Engagement objective
2. Website conversion location
3. Conversion event is VideoWatched custom event that fires when someone plays an embedded YouTube video
4. Attribution Setting: 1-day click
5. Advantage+ Audience with suggested audience
6. Excluding people who viewed the posts promoted in these ads
7. Advantage+ Placements
These ads drive people to custom post types for video posts where there's a short-form YouTube video embedded at the top and a short blog post below. The hope is that people watch the video (the optimization event), but they could certainly read the blog post, too.
Early returns have been good. While the number of watches of that video isn't huge, the percentage of quality traffic continues to be solid. I'll continue to monitor it over time as the audience is bound to expand more.
