Today in Digital Marketing - The Myth of "99% Brand Safe"

Starting point is 00:00:00 It is Wednesday, June 12th. Today, brand safety tools claim platforms they measure are 99% safe, but can marketers trust those numbers? Probably not. Also, YouTube adds the single biggest item from marketers' wishlists. TikTok's new search feature could sell more of your products. And garbage in, garbage out. The sad and dangerous state of AI search summaries.

Starting point is 00:00:28 I'm Todd Maffin. Here's what you missed today in digital marketing. One of the changes that marketers have most wanted of YouTube was the ability to test thumbnails in a proper multivariate test. Well, they've been trying it out for a year now with a handful of creators, and now they say they are ready to roll it out to everyone in the next one or two weeks. Here's how it'll work. You will give YouTube three images to use as thumbnail tests. YouTube will show those thumbnails evenly across your video's audience, and will select a winning thumbnail based on which one generates the most watch time. That's an interesting design choice, since at first you'd think they'd want to test based on which one generates the most watch time. That's an interesting design

Starting point is 00:01:05 choice, since at first you'd think they'd want to test based on clicks, but YouTube has always used watch time as its North Star metric, so I suppose it's not surprising. There's also no way to have it measure any other numbers. You can check the status of your test by going to the video details page and clicking view test report. From there, you'll see how much time is left for the test and how they're performing. You can also stop the test and manually set one of the thumbnails as the winners. If you let the test run though, the results can take a few hours or maybe even a couple of weeks to be finalized. The more thumbnail impressions that your video receives, the faster your test will finish. If there is a statistically significant result,

Starting point is 00:01:46 you will see a little label that says winner and YouTube will automatically update your video thumbnail to this winner when the test is done. You might also see a label that says preferred, which means the results were not statistically significant, but there was a thumbnail that performed slightly better. If there's no clear winner, the first thumbnail will be selected

Starting point is 00:02:06 and shown to your audience. You can also manually select the video thumbnail you want to use. It should be noted this new test will not work on shorts. It won't work on live videos, and it also won't be available for videos intended for children or mature adults. YouTube also had a few other updates,

Starting point is 00:02:22 including a way to use AI to skip to the good parts of a video that is running now on the Android YouTube app for premium subscribers. We have a link to videos showing off both of these in today's email newsletter, which is free to get. Just tap the link at the top of the show notes or go to todayindigital.com slash newsletter. Reddit this week announced a new partnership with DoubleVerify to provide brand safety and suitability measurement for advertisers. Quoting from their announcement, quote, Once DoubleVerify starts reporting on an advertiser's activity, they can hone their unified service and analytics platform, allowing brands to leverage keywords from post text and subreddits related to flagged posts within negative keyword exclusions or Reddit community exclusions, unquote. And while Reddit said that more than 99% of impressions appeared next to content it determined was brand safe, some analysts question that 99%. Quoting social media today, quote,

Starting point is 00:03:33 That is obviously a super high number and super impressive for Reddit in this context. But Double Verify also recently reported that more than 99% of measured impressions on X also appeared adjacent to content that was deemed safe. That's despite many third-party reports suggesting that X's ads are being shown alongside harmful content. So where's the discrepancy? Well, the most significant issue would appear to be DoubleVerify's methodology for reporting at platform level, which is seemingly not much more than a reflection of each app's own reporting on brand safety. As recently reported by Business Insider, quote, unlike on the open web, verifiers don't have a direct line of access to crawl the so-called walled gardens such as Meta, YouTube

Starting point is 00:04:22 and X. That leaves open the possibility that tech platforms can cherry pick the data for the likes of DoubleVerify and Integral AdScience to rubber stamp, unquote. So while this specific safety number might be a little less trustworthy than we'd like, even just having this partnership will likely help Reddit continue to grow through its adolescence into a more mature platform. TikTok is testing letting users take or upload a photo to find similar products in TikTok shop. It shows up as a camera icon in the search bar once you're in the shop UI. Quoting TechCrunch,

Starting point is 00:05:00 quote, say you're at a restaurant and you really like the plate you're served on. You could open up the TikTok app and snap a picture of it to find similar items available for purchase on TikTok Shop. Or say you're online shopping and you find a top you like that's too expensive. You could save the item and then upload it to find something similar with a cheaper price tag, unquote. This kind of tech is already in Google's Lens tool and Amazon's app. TikTok shop now has more than 15 million sellers worldwide, more than a half million of those in the US. We all had a bit of fun with Google's AI advising people to add glue to their pizza to keep the cheese in place. Many journalists, myself included, made fun of it all. Katie Notopoulos from Business Insider actually made and ate it. And many people watched her video, including apparently Google's

Starting point is 00:05:53 AI training bot. Because now when The Verge asked Gemini how much glue to add to pizza, rather than having been corrected by the AI engineers with a hard-coded don't do that answer. Gemini now uses Katie's post as a source for its answer. Quoting from one of Google's AI overview summaries, quote, according to Katie Notopoulos of Business Insider, you should add one eighth of a cup or two tablespoons of white non-toxic glue to pizza sauce to keep cheese from sliding off. Notopoulos says that mixing glue into the sauce didn't significantly change its consistency and the resulting pizza sauce had an orange color

Starting point is 00:06:32 similar to vodka sauce. Quoting The Verge, just phenomenal stuff here, folks. Every time someone like me reports on Google's AI getting something wrong, we are training the AI to be wronger. It's a big day in video gaming world, at least for fans of the video game that I play a lot of, which is Fallout 76. A new season came out today, so I'm super excited to get to it and grind it.

Starting point is 00:07:00 You know, it's one of these always online games, so there's a lot of grinding for end game gear. You know, right now I'm trying to build what they call a bloodied build, which is the less health you have, the stronger you are. And it's probably going to be another 50 hours for me to grind enough, do enough events to get all the gear that I want. And then, you know, they'll and then they'll nerf the weapon. That's a video game term for basically making it weaker, which means you'll have to grind for another weapon. I should hate it. I don't know why. I kind of like the grind.

Starting point is 00:07:32 There's some deep psychology happening. Anyway, that's it for today. Thanks for listening. I'm Todd Maffin. See you tomorrow.

Today in Digital Marketing - The Myth of "99% Brand Safe"

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.