Lately, there has been a lot of news about artificial intelligence (AI) and how it is going to drastically change many facets of everyday life. A variety of advanced AI tools have made the leap from concept to reality, including natural language generation (which allows for AI-generated content), speech recognition, chatbots, and machine learning (algorithms). Artificial intelligence has been a growing part of our digital lives since the dawn of the machine learning based technology, the search engine. Now, AI has come to revolutionize our search experience.
Over the last several years, the search ad platforms have increasingly encouraged its advertisers to embrace automation tools (i.e.: smart bidding strategies, broad match keywords, responsive search and display ads, performance max campaigns) that are designed to save time managing campaigns, improve campaign performance, and help brands reach their objectives more efficiently and effectively. They have also helped ease us into the future we now find ourselves in where more ad components, from creative asset generation and ad copy writing, to targeting and optimizations, are controlled, or at least greatly influenced, by AI technology.
Microsoft Beat Google to the AI Search Race
Microsoft announced in early February that they had rolled out an updated version of their Bing search engine, incorporating an artificial intelligence chatbot powered by OpenAI, the company behind ChatGPT, in which Microsoft invested $10 billion. Since that announcement, according to Similarweb.com, estimated weekly traffic to Bing has seen a noticeable increase, beating out traffic to Google some weeks. While Google is still the dominant player (approximately 70% of the U.S. search ad market), Microsoft’s growth is expected to outpace Google’s this year. On the heels of Bing’s OpenAI integration, Google has announced this spring that they too have an AI-powered, conversational chatbot as part of their new Search Generative Experience (SGE).
Google is slowly releasing access to their SGE, currently available by joining the Google Labs waitlist. While Bing’s AI chatbot is available to everyone now with a Microsoft account.
Both Search Engines Are Testing Generative AI Experiences with Ads
Google and Microsoft both claim advancements in large language models (LLMs) enable them to better understand intent behind complex search queries, which help power their AI-generated results. The updated search engines are programmed to learn what users want to know based on what they’re searching and provide them options for buying products as well as answers to their search queries right on the search engine results pages (SERPs). All this while being more conversational, like talking to a person, often suggesting follow-up searches. Paid search and shopping ads may appear within the generative AI experience. The platform’s AI will then draw on the campaign’s landing pages to automatically generate the ad content and match the ad to the user’s journey.
The way ads will display in AI search results is different than we are used to because overall search results pages will look and function differently. At this time, not all advertisers have visibility into delivery and performance, nor any greater control over how and when search ads serve in AI-generated results. Greater transparency and control should come as these offerings evolve and are released for all advertisers.
Part of the changes to SERPs, which Google and Microsoft have been prepping us on for years, includes creating shoppable and interactive experiences that can drive both transactions and leads from within the search engine experience. This does not necessarily mean less revenue for advertisers or fewer prospects in the funnel; rather, more customers may not need to click a link out to an advertiser’s website and instead complete their conversion on the SERP.
eMarketer’s Insider Intelligence estimates in 2022 that Google generated over $132 billion dollars in search ad revenue globally. While AI generated search results might reduce the need for searchers to click through to a website from the results page, none of the search engines are going to make changes to their SERPs that will negatively affect their ad business.
Additionally, when search ads first rolled out, organic click rates declined; that trend continued as competition for search traffic grew and the search engines continue to prioritize paid placements ahead of organic. This evolution has not eliminated the need for strong SEO; if anything, increased competition for search traffic has fostered the growth of search engine marketing businesses. Like with the introduction of paid ads on the SERPs, experimental generative AI search experiences will be disruptive. Expect a learning curve as we navigate these new waters together.
New AI-Powered Ads Coming From Google
At this year’s Google Marketing Live event on May 23rd, we learned of what to expect from Google’s artificial intelligence-powered ads in the coming year.
Create campaigns using conversational AI
Google will make available a new search campaign setup that will allow for AI-driven chat search ads. Advertisers will be able to edit the assets generated by Google’s AI before launching.
Generative AI coming to Performance Max campaigns
One of Google’s newest campaign types will soon allow for custom asset creation in the Google Ads platform.
New Google tool makes AI image creation a breeze
Merchants without many creative assets will be able to use Google’s new Product Studio to create unique product imagery.
Merchant Center Next eliminates the need for a product feed
Automatically update product info from your website with no need for a feed.
2 new video-focused campaign types coming soon
First, for YouTube, a new Video View campaign will become available, combining skippable in-stream ads, in-feed ads, and Shorts ads in one campaign. In addition, a new Demand Gen campaign will show video ads across YouTube Shorts, in-stream, and in-feed placements as well as in Google Discover and Gmail, with the objective of driving conversions. A new lookalike segment builder will also be incorporated.
Hype and Uncertainty Are High Right Now
AI-generated content is still in its early stages. Results from the technology range from very impressive to completely wrong. And, there are plenty of searches where the engines either cannot produce generative AI results or are choosing not to display them to avoid sensitive topics (see the screenshot below). The technology is likely to improve in effectiveness and use cases over time.
Concerns with AI generated content are abundant and will need addressing before we start to see a greater adoption.
Trust issues persist due to misinformation, over personalization, and many unknowns
AI generated content is particularly worrisome right now for a variety of reasons including plagiarism, accuracy, recency, and relevancy. While the future of generative AI search results is exciting, the imperfections of the technology today means AI creators have a long way to go to build trust and comfortability with consumers and advertisers.
Measurement solutions will have to catch up
It is possible click rates may drop as more searchers will get their answers or make a purchase from the search engine. This will mean a reduced reliance on measuring performance from clicks. This could have profound impacts on advertisers, publishers, and all those who publish web content expecting search traffic. Although, more than half of Google’s revenue comes from ads; so, do not expect disruptions that would reduce interactions that Google will charge advertisers.
Expect regulation and refinement
Right now, AI is in its "wild west" phase. Government agencies, the makers of the AI tools, and industry bodies will work together to answer the important ethical and legal questions raised by the technology.
Top 3 Ways Advertisers Can Incorporate AI Into Their SEM Strategy
Embrace automation & AI tools
Shift your mindset and lean into the promise of automation and AI by creating a culture of experimentation. Test, learn, and stay agile. Advertisers who avoid the AI revolution may be left behind and struggle to catch-up with competitors.
The AI chatbots are suggesting very specific searches. This highlights the need to adopt either keyword broad matching or a have a long-tail keyword strategy to capture more of the traffic coming from lengthy, specific, AI chatbot-generated searches.
Initial testing of the new search experiences seem to be resulting in fewer, in theory better, ads being served in AI-generated results. If this trend continues, this would further highlight the need for search advertisers to focus on in-platform performance indicators like ad strength, keyword quality scores, and campaign recommendation scores – all factors in if & when the search engines show an ad and how much the advertiser pays for the exposure.
Build a strong measurement foundation by combining first-party data with AI
With third-party cookie deprecation comes signal loss. Counter this by enabling as much data sharing with the search ad platforms as possible.
Ensure robust sitewide tagging enabling advertisers to capture important data for measuring performance and optimizing campaigns.
Assign values to conversion actions. This can include variable revenue or estimated lead values to help calculate and optimize towards actions most valuable to your business.
Utilize advanced data matching functionality like enhanced conversions to help Google fill in the data gaps on your converters.
Upload hashed customer data in a privacy-centric way to educate the ad platforms on who your customers are beyond what the ad platforms are able to capture via the tags and enhanced conversions. Doing so is likely to improve campaign performance.
Maximize results with AI-powered campaigns & features
There are several best practices performance advertisers can employ to take advantage of existing AI tools. First, activate both Search & Performance Max campaigns to drive more conversions.
Use value-based automated bidding strategies to generate the greatest return on ad spend or lowest cost per conversion.
Incorporate a variety of strong creative assets. The best performing ads are typically ones where the platform’s machine learning can optimize against a sampling of text, image, and video assets.
The search landscape is changing rapidly as artificial intelligence tools continue to be released, tested, and refined. Remain nimble and embrace the possibilities that AI may bring – complete campaign tasks quicker than ever before, allow a bot to generate assets for you, and expect improved campaign performance.