Google Integrating Generative AI Into Search Ad Creation
Google Integrating Generative AI Into Search Ad Creation
Digital Marketing Trends: AI, Audio, Consumer Products, Politics, SEO, Social Media, Visual Communications & Video
AI-Powered Advertising Creative
I ran across this a couple weeks ago while setting up a Google search ad campaign:
Google Ads told me: "Google AI can help you create your ads" and offered to "suggest ad headlines and descriptions based on information about your business."
Write a short description and Google will create your ad. I declined, of course, given this was a campaign for a client, the feature is in beta, and I had not run across any coverage of this feature in trade pubs.
This test popped up for me when setting up a Responsive Search Ads campaign but Search Engine Land now reports that Google is integrating generative AI into Performance Max campaigns. The article explains:
Advertisers can submit creative content such as images, videos, and text related to a campaign, and the AI will “remix” these materials to generate ads that target specific audiences and meet objectives like sales targets, according to the presentation.
However, there are concerns that the tool could disseminate misinformation, as AI-generated text can confidently assert falsehoods. One individual familiar with the presentation commented that the AI is optimized for converting new customers and does not have an understanding of truth.
While it has its updsides, over the years I’ve learned a hard lesson about the downsides of being a first-adopter (some day I’ll share the photo of me wearing my Google Glass), so you’ll forgive me if I don’t jump on the generative AI ad creation bandwagon just yet.
That last paragraph quoted above illustrates why I’m content to let others go first with this new feature.
My main concern about AI writ large is that it is increasingly becoming a black box, making it ever more difficult to understand why it does what it does. There’s little algorithmic transparency. So we should expect unexpected results.
When generative AI “can confidently assert falsehoods,” that’s a serious problem for any organization or brand deploying it in ad campaigns.
There’s the obvious problem with generative AI lying about your offering in order to optimize your ad campaign objective.
But there’s a deeper issue that taps into the larger concern about how AI effects society at large.
Embedded within the vast trove of data these generative AI tools have been trained on are the psychological biases and blindspots we human possess. It doesn’t take a vast leap of the imagination to foresee that generative AI may tap into those lies we tell ourselves in order to manipulate us to into a purchase we wouldn’t otherwise have made or to support a political candidate who is clearly contrary to our own interests.
Granted, it’s not like advertising doesn’t already use such psychological triggers (fear of missing out is a classic advertising tactic) but the psychological associations AI is able to make is likely much more subtle than a single human mind is capable of.
The generative AI suggestions offered by the Responsive Search Ads example I cited at the beginning are just that, suggestions. Google has been offering AI-generated suggestions throughout its products for years. They are often helpful but you can take them or leave them.
But handing over the keys to the Corvette to a generative AI Performance Max campaign does not seem wise at this point.
Trends I’m Following
I’m trying something new with this “what I’m reading” section. Rather than include random articles that I find interesting, I’m going to try narrowing them down to digital marketing trends and a digital marketing “changelog.” Reply to let me know what you think.
Artificial Intelligence
The Atlantic - Our relationship to the written word is fundamentally changing. So-called generative artificial intelligence has gone mainstream through programs like ChatGPT, which use large language models, or LLMs, to statistically predict the next letter or word in a sequence, yielding sentences and paragraphs that mimic the content of whatever documents they are trained on. They have brought something like autocomplete to the entirety of the internet. For now, people are still typing the actual prompts for these programs and, likewise, the models are still (mostly) trained on human prose instead of their own machine-made opuses.
But circumstances could change—as evidenced by the release last week of an API for ChatGPT, which will allow the technology to be integrated directly into web applications such as social media and online shopping. It is easy now to imagine a setup wherein machines could prompt other machines to put out text ad infinitum, flooding the internet with synthetic text devoid of human agency or intent: gray goo, but for the written word. Read the full article: Prepare for the Textpocalypse
TechCrunch - As the European Union gears up to enforce a major reboot of its digital rulebook in a matter of months, a new dedicated research unit is being spun up to support oversight of large platforms under the bloc’s flagship Digital Services Act. Read the full article: Europe spins up AI research hub to apply accountability rules on Big Tech
Bloomberg - Milo Van Slyck missed an appointment with his therapist in early April, so he decided to try something new: telling ChatGPT about his problems.
As Van Slyck, a paralegal in Charleston, South Carolina, typed, he found that he felt comfortable discussing a range of deeply personal issues. He told OpenAI’s chatbot about his fears and frustrations as a transgender man at a time when transgender rights are under attack in much of the country. He mentioned conflict with his parents, who aren’t supportive of his gender identity, and his preparations for an upcoming visit. Read the full article: AI Therapy Becomes New Use Case for ChatGPT
Cornell University - Public opinion reflects and shapes societal behavior, but the traditional survey-based tools to measure it are limited. We introduce a novel approach to probe media diet models -- language models adapted to online news, TV broadcast, or radio show content -- that can emulate the opinions of subpopulations that have consumed a set of media. To validate this method, we use as ground truth the opinions expressed in U.S. nationally representative surveys on COVID-19 and consumer confidence. Our studies indicate that this approach is (1) predictive of human judgements found in survey response distributions and robust to phrasing and channels of media exposure, (2) more accurate at modeling people who follow media more closely, and (3) aligned with literature on which types of opinions are affected by media consumption. Probing language models provides a powerful new method for investigating media effects, has practical applications in supplementing polls and forecasting public opinion, and suggests a need for further study of the surprising fidelity with which neural language models can predict human responses. Get the full study: Language Models Trained on Media Diets Can Predict Public Opinion
- We need a new mental model for the AI enterprise, one that reflects AI’s emerging role as an infinitely flexible brain sitting at the center of the organization, the people inside of the organization as planets that move in its orbit. Read the full article: A New AI Organization BeckonsAudio
NPR - Texting can muddle meaning, and calls can trigger anxiety. But for many, short voice recordings offer an easier, low-pressure alternative in a world that's grown more accustomed to audio mediums such as Clubhouse and podcasts.
The ability to communicate tone is a big part of the appeal.
According to a recent YouGov survey conducted by Vox, 62% of Americans say they've sent a voice message, and about 30% communicate by voice message weekly, daily or multiple times a day. And 43% of 18- to 29-year-olds who responded to the survey said they use the feature at least weekly.
WhatsApp revealed last year that an average 7 billion voice messages were sent daily on the app. Read the full article: For Gen Z, are voice notes are the new text?
Pew Research - The survey asked adult podcast listeners (those who said they listened to a podcast in the past 12 months) in the United States to name the podcast they listen to most. Overall, 61% of listeners volunteered the name of a podcast, but there was relatively little common ground in these answers.
Only six podcasts were mentioned by at least 1% of those who volunteered an answer: The Joe Rogan Experience (named by 5% of those who gave a response), The New York Times’ The Daily (2%), Crime Junkie (1%), The Dan Bongino Show (1%), NBC’s Dateline (1%) and The Ben Shapiro Show (1%). Looking at all U.S. podcast listeners, as opposed to just those who gave a podcast, no single podcast was named by more than 3% of listeners. Read the full article: Americans differ widely in the podcasts, topics they listen to most often
Consumer Products
Insider - Some younger viewers said they'd gladly traded pandemic TV binge-watching habits for social and outdoor activities. Others said they were open to renewing — but probably just to watch a few new shows or movies before canceling again.
New research released Monday by Deloitte illustrates what this phenomenon means for streamers. Nearly a third of millennial respondents (32%) and 30% of members of Gen Z said they'd canceled at least one paid entertainment subscription within the past six months to save money, the firm's 2023 Digital Media Trends survey found. Read the full article: Millennials, Gen Z Cancel Streaming Subscriptions Amid Inflation
Politics
- It's no accident that the transformation of the GOP into an autocratic entity during Donald Trump's presidency accelerated the Republican war on education. All of the fronts of this ongoing war draw on authoritarian history: the book bans, the censorship of curriculums, the surveillance of syllabi and lesson plans, the demonization of professors, teachers, and librarians as perverts and radical leftist indoctrination machines, the assertion of "parents' rights" as a cover for right-wing activism, and the establishment of pathways for a culture of informing to take hold. Read the full article: Seeking to Turn Students into Informers, Republicans Continue Authoritarian HistoryThe Daily Beast - Last week, investigative journalist Saurav Das shared the fact that—in response to legal demands—Twitter blocked access to two tweets he had posted about India’s Minister of Home Affairs, Amit Shah. Censorship demands originating from the Indian government are nothing new, and Twitter boss Elon Musk has thus far agreed to grant them, blocking the material from view within India.
This is in line with Musk’s faulty understanding of “free speech” as a simple reflection of an individual country’s laws, no matter how oppressive.
But this time, Twitter did something different: It blocked the tweets not just within India, where Indian law applies, but everywhere. As with many of Twitter’s moves of late, it’s unclear if this decision is a result of Musk’s personal directives; understaffed teams; slipshod, off-the-cuff policy making; or all three. Twitter has yet to offer an explanation, despite requests from the journalist whose tweets were blocked. Read the full article: Tech Bosses Are Letting Dictators Censor What Americans See
Foreign Affairs - In the seven years since Russian operatives interfered in the 2016 U.S. presidential election, in part by posing as Americans in thousands of fake social media accounts, another technology with the potential to accelerate the spread of propaganda has taken center stage: artificial intelligence, or AI. Much of the concern has focused on the risks of audio and visual “deepfakes,” which use AI to invent images or events that did not actually occur. But another AI capability is just as worrisome. Researchers have warned for years that generative AI systems trained to produce original language—“language models,” for short—could be used by U.S. adversaries to mount influence operations. And now, these models appear to be on the cusp of enabling users to generate a near limitless supply of original text with limited human effort. This could improve the ability of propagandists to persuade unwitting voters, overwhelm online information environments, and personalize phishing emails. The danger is twofold: not only could language models sway beliefs; they could also corrode public trust in the information people rely on to form judgments and make decisions. Read the full article: The Coming Age of AI-Powered Propaganda
Search Engine Optimization
Whitespark surveyed 44 different local SEO experts, asking them to rank 149 potential local search ranking factors in order of importance across both the local pack and organic rankings.
The local SEO experts considered the following the most important ranking factors for Google’s local pack:
Primary Google Business Profile category.
Keywords in Google Business Profile title,
Proximity of address to the point of search,
Physical address in the city of search,
High number of Google ratings,
Additional Google Business Profile categories, and
Verification in Google Business Profiles.
Get the full report: Whitespark
Digiday - Publishers like Ingenio, Team Whistle, BuzzFeed and Gannett are experimenting with how generative AI technology can support their SEO strategies, including using it to optimize headlines and keywords for search and to find new topics to create content around to drive traffic.
It’s unclear how successful these tactics will be. Traffic hasn’t grown exponentially since Ingenio started experimenting at the end of 2021 with generative AI for SEO, though president of media Josh Jaffe said anecdotally at the Digiday Publishing Summit in Vail, Colorado last week that traffic has “gone up.” Read the full article: Publishers test generative AI
Social Media
The Drum - Since the early days of social media, follower counts have held a kind of cultural currency.
That currency became more literal when popular users became influencers and celebrities in their own right – paid to be brand marketing tastemakers. Follower counts, and our perception of them, have given rise to a $21bn industry (per Influencer Marketing Hub data). But time will tell if they will play the same role in influencer marketing’s next evolution.
More recently, though, content algorithms on platforms such as TikTok and Instagram have started prioritizing discovery, allowing smaller creators to get better reach. This has made ’followers’ an even less relevant metric. These days, an influencer with millions of followers can get a small fraction of that number in views, while a post from a micro-influencer has a chance to become the next TikTok sensation. Read the full article: Are Influencers Ready For A World Without Follower Counts?
The Verge - Unlike most other businesses on Earth that live and die by their customers’ demands, social media services are caught trying to satisfy both their users and the people actually paying for it all: investors and advertisers.
The needs of these groups are dramatically different. Users want what the platform was originally for — be it ephemeral messaging, sharing photos, or otherwise. Surprising, energized spaces to connect with friends in a new way. But these use cases inevitably have a limit. You can only post so many photos. You only have so many friends to message. And for investors and advertisers, that’s a problem. So each social network has to find ways to make you send another photo, or it has to deploy a brand-new feature and encourage you to use that, too. More usage, more space for ads, more money for investors. Read the full article: Social media is doomed to die
Visual Communications
Motherboard - The Midjourney subreddit is being flooded with images that depict images of historical events like “The infamous Blue Plague Incident" that occurred in the 1970s in the Soviet Union, the “July 2012 solar superstorm and blackout” in the U.S., and “The 2001 Great Cascadia 9.1 Earthquake & Tsunami” that devastated the West Coast of North America.
Each post showcases a slideshow of images that share various perspectives of the event, as well as important events following the catastrophe, including press conferences and cleanups. The catch is: These are all AI-generated images and none of these events have ever occurred. Read the full article: People Are Creating Records of Fake Historical Events Using AI
Ars Technica - Not long after the wonder of generating solo images emerged, some artists began creating multiple AI-generated images with the same theme—and they did it using a wide, film-like aspect ratio. They strung them together to tell a story and posted them on Twitter with the hashtag #aicinema. Due to technological limitations, the images didn't move (yet), but the group of pictures gave the aesthetic impression that they all came from the same film. Read the full article: Artists astound with AI-generated film stills from a parallel universe
Video
A report by Vidmob analyzed more than 800 million ad impressions from brands executing paid video advertising campaigns on LinkedIn and found videos that include the following elements performed better than those that did not:
Videos that include messaging in the first quarter of the content enoyed more than 149% lift in view-through 25%,
More than 54% lift in engagement rates with 7-15 sec video duration,
More than 102% lift in view-through 25% with high text contrast,
More than 175% lift in view-through 25% when showing a person in the opening quarter,
More than 33% lift in click-through rate when offer includes ‘Get a Quote’, and
More than 17% increase in click-through rate when featuring brand logo in the opening 2 sec
Get the full report: Vidmob
New York Times - Ian Sansavera, a software architect at a New York start-up called Runway AI, typed a short description of what he wanted to see in a video. “A tranquil river in the forest,” he wrote.
Less than two minutes later, an experimental internet service generated a short video of a tranquil river in a forest. The river’s running water glistened in the sun as it cut between trees and ferns, turned a corner and splashed gently over rocks…They represent the next stage in an industry race — one that includes giants like Microsoft and Google as well as much smaller start-ups — to create new kinds of artificial intelligence systems that some believe could be the next big thing in technology, as important as web browsers or the iPhone. Read the full article: Instant Videos Could Represent the Next Leap in A.I. Technology
Wired - You may have noticed some impressive video memes made with AI in recent weeks. Harry Potter reimagined as a Balenciaga commercial and nightmarish footage of Will Smith eating spaghetti both recently went viral. They highlight how quickly AI’s ability to create video is advancing, as well as how problematic some uses of the technology may be.
These clips suggestt we are at an inflection point for AI video making. As with AI image generation, a growing rush of memes could be followed by significant improvements in the quality and controllability of AI videos that lodge the technology in all sorts of places. Read the full article: Free AI Video Generators Are Nearing a Crucial Tipping Point
Commercial Earworm
The weather in Minnesota is not playing ball so far, so we’ll need to get a hint that it’s Spring from the heavy rotation of Miracle-Gro commercials that have been running on TV lately.
I gotta say, the ad the company is currently running is as catchy as a Taylor Swift song.
No, it’s not the music in the ad that’s stuck in my brain. It is the British woman voicing the tagline ending on the high note: “Miracle-Gro: All you need to know to gro.”
Don’t tell anyone, but it’s so cute, I find myself repeating the line each time the commercial comes on.