Google

    Friday Marketing Links | 031524

    1. TikTok’s potential U.S. ban stirs marketers, spurs contingency planning

    Meta could capture between 22.5% and 27.5% of TikTok’s U.S. ad revenues in the event of a ban.

    YouTube stands to gain an additional $1.24 billion to $1.53 billion, with $410 million to $500 million of TikTok’s ad revenues redirected to Google’s display and search businesses

    One of the early thought exercises I was given at Blue Ion was: what happens if Meta was shut down tomorrow?

    It’s a good question to occasionally ask about any important distribution channels.

    1. Apple Buys Canadian AI Startup as It Races to Add Features

    DarwinAI has developed AI technology for visually inspecting components during the manufacturing process and serves customers in a range of industries. But one of its core technologies is making artificial intelligence systems smaller and faster. That work that could be helpful to Apple, which is focused on running AI on devices rather than entirely in the cloud.

    Apple was launching Vision Pro while the rest of the Valley Giants were pivoting to AI. But the benefit of a massive bank account is the ability to buy whatever you want.

    Apple has also been really secretive about their AI plans, claiming “disclosure of strategic plans and initiatives harmful to our competitive position and would be premature in this developing area.”

    Apple is at the forefront of ambient computing, and on-device AI will be a key component. Plus, Apple is the only one of the giants that isn’t really a cloud company and is most definitely a hardware company.

    1. Report Finds No Correlation Between Social Media Engagement and Content Readership

    Social media apps are gradually becoming more valued as entertainment sources, while actual interaction shifts to smaller, enclosed chats and communities.

    Notice how all the platforms focus on “discovery.”

    Across all the articles and topics we analyzed, we found no clear connection between social engagement and actual readers of the news.

    Understand vanity metrics vs. brand metrics vs. performance metrics.

    1. Podcast Frenzy Report

    podcasting is taking over traditional media consumption time, with respondents reporting 28% of them watch less TV and 24% browse social media less often. Gen Z podcast discovery is a mix of methods. 46% of Gen Z respondents rely on social media recommendations, and 33% of younger Gen Z browse top charts and “best of” podcast lists.

    Audio! Audio! Audio!

    1. What We Learned About Creative From Analyzing $3M in Podcast Media by Caroline Culbertson

    Findings include midrolls outperforming both pre-rolls and post-rolls for placement. A quiet value-add for host-read contracts is hosts tend to go over their contracted ad length. Right Side Up found the sweet spot for “60 second” host-read ad performance was host-read creatives that landed between one to three minutes.

    Podcasts foster parasocial relationships which gives host-read ads some extra oomph in the persuasion department.

    1. The bad ad ecosystem: Here’s what the research says

    five types of bad ads, each varying in harm for the marketer: malicious ads, spoofed ads, scam ads, heavy ads and miscategorized ads.

    The easiest thing is [ad buys] are cheap. [Bad ad creators] don’t wanna spend a ton of money on it. So they proliferate in places with really low CPMs

    marketers should work on making good ads. Ensuring the proper ads for the right environments is key, along with keeping on top of creative

    1. Layoffs could be coming as debt-laden firms navigate the pain of higher rates, economists say

    Higher rates spell trouble for US companies with near-term debt maturities.

    Rate changes and inflation measures are the important indicators this year.


    Are the algorithms starting to turn on generative AI?

    Google has updated Search to derank spam including:

    • Scaled content abuse (aka made by AI)
    • Site reputation abuse (junk third-party content on quality sites)
    • Expired domain abuse (grabbing an expired domain and starting a spam farm)

    The goal is to:

    better understand if webpages are unhelpful, have a poor user experience or feel like they were created for search engines instead of people. This could include sites created primarily to match very specific search queries.

    This change actually has teeth too, with plenty of manual actions.


    Another Podcast’s episode Google Gemini and AI bias unpacks the weights and training data stuff I briefly mentioned yesterday.

    Hidden patterns in data and models as bias amplifiers.

    Also the excellent quote “the fake is in the caption,” because what is a prompt but a caption for a future image?


    Quick Hit Google Bits

    Google Analytics 4 launches new trend detection insight

    trend change detection focuses on slower changes happening over a longer time. This gives users a detailed view of data changes, making it a useful tool to spot both quick and long-term trends.

    GA4 is introducing two dedicated spaces – one for marketers to track and analyze campaigns, and another for behavioral insights.

    • The Reports section provides insights into how users engage with your websites and apps so you can improve your product and user experience.
    • The Advertising section will become the hub to monitor and analyze your campaigns whether you’re a publisher or an advertiser.

    Reddit signs AI content licensing deal with Google

    reportedly paying Reddit $60 million per year to train its artificial intelligence models (e.g., Gemini).


    OG journo-blogger Matt Welch on legacy media, like the Village Voice & LA Times:

    Never could adapt to the new reality. It’s a success curse.

    When you’re making that much money, you’re that powerful, the people who work for you—including yourselves—are not going to solve the industry once the industry turns and changes.

    This is not just a media problem.

    Same could be said of Google right now.

    How do you change when the change threatens your business model?


← Newer Posts Older Posts →