Brand Knowledge Will Level Up Your AI-Driven Search Engine Marketing

As ChatGPT, Bard and AI-enhanced Bing have rolled out to the public, the question on every marketer’s mind has been, "When can we advertise in these results?" OpenAI, Google and Microsoft have been tight-lipped about when these capabilities will be available, but they’ve certainly alluded to their future. Google has already provided examples of shopping results within its conversational AI-based chat feature, and OpenAI just launched ChatGPT Enterprise, so we know they’re coming. All of us who perform any level of search engine marketing (SEM)—bloggers, solopreneurs, ecommerce sellers and, yes, SEO professionals—will need to adapt our strategy or risk becoming obsolete. So how can we prepare for something we know relatively little about? How SEO Agencies Are Adapting to a World Without Traditional Search Quality over quantity With Google’s nascent search generative experience in Search Labs, the goal is to make your business or brand eligible to show up in the results. A person asking one question doesn’t need 14 pages of results; they need one or two excellent answers from the most qualified sources. Looking back at the classic Quality Score measure, we can assume the higher the quality of content, the more likely it is to be chosen for the results. Google recommends investing in high-quality creative assets for your website, which will likely also help your chances of being used as a source by other LLMs. The new technology is meant to support a helpful and reliable search experience, not detract from it. Context clues That handy checklist released at Google Marketing Live in May contains clues about how to adapt SEM strategy—not just to Bard but potentially to Bing and ChatGPT as well—such as: Whatever LLM you’re using will need sufficient contextual information about your product or expertise to know when it might be relevant to a searcher. For example, it’s not enough to know that Premier Protein makes protein shakes—the LLM needs to know that protein shakes can be a healthy breakfast or snack, that protein intake is imperative for building and maintaining muscle mass, that it’s effective after a workout, that it appeals to fitness-conscious consumers on a budget, and so on. It’s not so different from regular SEM in this way. Use Google Ads to experiment with broad match, long-tail keyword combinations and descriptions. Learn how your target audience naturally speaks, building your keywords around how a user would phrase their query in conversation. Align your #goals Like any search engine, an AI chatbot needs to know your campaign goals to find searchers whose needs align with them. Set up proper goals (by managing your tags) and attribution (how you’ll report and measure ad success in relation to your business goals). Set up enhanced conversions and value-based conversion bidding based on your business goals (revenue, profit margins or lifetime value) well in advance to provide the LLM with rich data points that guide its automated bidding and machine learning. Google’s conversational copywriting feature for search ads isn’t widely available yet, but select agencies can apply for early access through their Google reps. By leveraging an AI tool that learns from 8.5 billion search queries a day, we’ll learn how to better speak to targeted users. In terms of getting to know your brand, ChatGPT’s brand plug-ins will be unmatched in efficiency and accuracy—when they become available to everyone, that is. Kayak, Instacart, Expedia and a handful of others have been selected by OpenAI to develop the first brand plug-in architecture, and there’s a waitlist for others who want to follow. By feeding ChatGPT detailed information directly from your knowledge base, a user can obtain correct, consistent answers without having to seek out your brand (or even know about it) first. For example, a user could ask for a flight itinerary from Los Angeles to Bali on a given set of dates, and ChatGPT can provide real-time flight information straight from Kayak. While OpenAI works out UX kinks and data security concerns, marketers should play with the existing plug-ins for inspiration. Prep your team Advertisers will need to align internally as an agency and externally with clients to agree on a brand-safe approach to generative AI SEM. Create a “culture of experimentation,” as Google puts it, in which everyone feels comfortable testing and learning from mistakes. What’s AI Actually Good for Right Now? Your shared goal is to adequately leverage the LLM’s data to see efficient results without giving away too much proprietary information. All of the AI-powered chat tools warn not to put any personal or private information into the system because it may end up relayed to someone else. For a business, that means not giving away the secret sauce. Start streamlining your process for ad copy approval such that you can leverage the quick copy outputs from Google while maintaining brand guidelines. Search engine marketers will likely adapt to the new technology before brands (or even other marketers) can, so prepare to patiently educate clients and colleagues on your reasoning for strategy adjustments. Show them you’ve done your homework and come ready with examples of other brands that have participated in these early testing phases. Any paid media will need to flow with the conversational experience as naturally as possible, so continue to focus on quality, relevance and keeping up with the evolution of AI search tools.