Technology which converts text cues to headlines, conversational passages, and photographs is just starting to be incorporated into digital ads. OpenAI's ChatGPT is amongst the various AI-driven tools that the primary online ad companies are testing. "It's going to have a massive effect on electronic advertising," declared Cristina Lawrence, the executive vice president of Razorfish's consumer and content experience.
Last year, ChatGPT made headlines for its capacity to interact with users in a convincing, lifelike manner. Entrepreneur Shane Rasnak was quick to test the potential of generative artificial intelligence (AI) for digital marketing. With the help of AI, jobs that would usually take him up to an hour now only require 15 minutes. And that's just the beginning.Rasnak has also been using software like Midjourney, which can turn text into graphics, to come up with exciting visuals for his Facebook ad campaigns. Such programs are especially helpful to those without design qualifications, while popular tools like Canva and Adobe Photoshop are also used.Rasnak feels that generative AI is like the introduction of social media in its influence on digital ads, allowing marketers to target customers based on their interests. It has enabled them to create customised messaging and visual content. “The quality, output and level of personalisation that can be achieved completely changes the game”, says Rasnak.His enthusiasm for AI is shared by the bigwigs in online advertising. Alphabet, Meta and Amazon have each announced plans to develop applications to help businesses produce messages, pictures and videos for their platforms.
Though their products are still largely in trial phases and some have been criticized for being too hastily introduced, ad experts have told CNBC that, when taken together, generative AI represents the following step in online advertising.Cristina Lawrence, executive vice president of consumer and content experience at Razorfish, a major digital marketing agency part of the renowned Publicis Groupe, expressed that this kind of technology “is going to have a massive influence on digital advertising”.In May, Meta launch the AI Sandbox testing suite, which makes it easier for companies to take advantage of generative AI software to create background images and to test different advertising copies. Furthermore, Meta also made alterations to the Meta Advantage service, which applies machine learning in its endeavour to enhance the efficacy of promotions and ads that appear on several social programmes.Given that Apple's 2021 privacy update prohibits its capacity to monitor and track users' activities on the web, Meta firmly argued that the Advantage suite can aid companies to deliver better outcomes from their campaigns.
A bicycle company, for example, could theoretically use Facebook to target users in Utah with AI-generated graphics of people cycling through desert canyons while users in San Francisco could be viewing advertisements which feature cyclists cruising over the Golden Gate Bridge, ad experts predict. Additionally, the text of the advertisement can be personalized based on the person's age and interests. Lawrence commented "You can be using it for that sort of personalization at scale."According to data from Varos, an online marketing firm, Meta's Advantage service has been gaining traction as retailers use it for automated shopping ads. In May 2023, roughly 2,100 companies spent an estimated $47 million, or around 27.5% of their combined total Meta advertising spend, on Advantage+. This is a noticeable increase from the previous month where these same companies allocated 26.6% of their budget, or $44.9 million, to Advantage+. When Meta formally launched Advantage+ last August, companies used less than 1% of their Meta ad budget for the service.’s Fleets
Yarden Shaked, the CEO of Varos, noted that Facebook has had some success in coaxing advertisers to utilize its automated advertisement systems, yet he added that he is "not sold on the creative piece yet," in terms of the fledgling Meta's attempt to offer AI instruments to advertisers. Rasnak mentioned that Midjourney's tool isn't "quite there yet" as far as generating realistic visuals that can be part of an online ad, but is proficient at developing "cartoony designs" that are appealing to some smaller clients. Forrester analyst Jay Pattisall stated that a few noteworthy obstacles are hindering generative AI from having a major effect on the online advertising industry straight away. One of these is brand safety, as companies are uncertain about relying on generative AI to generate visuals and phrases which can often be biased, offensive, or imprecise. Bloomberg News discovered that AI-generated imagery from the popular Stable Diffusion tool produced visuals that contained a number of stereotypes, with darker skin tones assigned to jobs such as "fast-food worker" or "social worker" while lighter skin tones were attributed to higher-paying jobs. Additionally, there may be potential legal issues pertaining to models trained on data from sources such as Reddit, Twitter, and Stack Overflow, who can charge AI companies for the use of data on their sites. Long-time marketing writer and consultant Scott McKelvey referenced other limitations associated with the quality of the output. He stated that, based on his experience with ChatGPT, the OpenAI chatbot, the technology doesn't make the kind of long-form content that can be used in promotional pieces. An OpenAI spokesperson declined to comment, while a Meta spokesperson said they have taken steps to reduce any bias in their AI systems, alongside brand-safety tools to provide advertisers more control over where their ads appear online. Furthermore, they added that they would eliminate any AI-generated material that goes against their rules.
Online advertising and Facebook ads consultant Stacy Reed has just started to incorporate generative AI into her work. She is implementing the software to come up with ideas for Facebook advertising headlines and short copy, and has found it useful in an era when it is increasingly difficult to trace users online. She described generative AI as a good "starting point," but believes companies and marketers still require their own brand messaging strategy and not depend on generic content. According to her, companies must not depend on the technology to manage the overall thinking of understanding which themes resonate with different audiences or how to execute significant campaigns across multiple platforms. For now, major ad agencies and big companies are using generative AI mostly for trial runs while waiting for the technology to evolve, according to industry experts. Just this year, actor and co-owner Ryan Reynolds used AI-writing for an ad for Mint Mobile. Reynolds went on to describe the AI-created script as "mildly terrifying, but compelling" – an apt description for the industry's current attitude to generative AI.
Comentários