From SEO To GEO: What GPT Marketers Need to Know
If you are 25 or younger, chances are high that you never encountered the paper version of Yellow Pages but throughout the 20th century, print directories were among the primary ways for consumers and businesses to connect. Established in 1886, Yellow Pages posted its final print issue in January 2019 closing the chapter on 130 plus history of print directory marketing. In the late 1990s, the new exotic profession of online directory marketing emerged with the rise of Yahoo! and other online directories, and quickly disappeared as the search engines took over. Search engine to be precise, since Google quickly took the lion’s share of the market in the early 2000s. Since then, every business is being bombarded by armies of search engine optimization (SEO) marketers offering to analyze and optimization of your websites, social networks, and all kind of tricks designed to get the business to the top of search results. According to Artios, there were over 200 thousand SEO professionals on LinkedIn alone and almost half of the people used a search engine when looking for products or services. With the advent of social networks, another notable profession emerged – Social Media Marketer (SMM). And many SEO “professionals” added SMM to their offerings.
At the end of 2022, OpenAI’s ChatGPT took the market by the storm beating Google at its own game in transformer neural networks, and triggering the massive transition from search to AI-generated answers and content. At present, Microsoft, Google, Amazon, and every other major technology giant is racing to advance its generative AI capabilities and introduce Large Language Models (LLM) and conversational AI into their products. This transition also elevated the many generative AI startups, including mine, which published in this area since 2016 and got the first drugs imagined by generative AI into the human clinical trials in 2021. Even for experts in this field, the release and rapid adoption of ChatGPT was largely unexpected and served as a call to accelerate and change. In order to stay competitive in generative AI, companies need to develop much better generative tools, produce specialized training data sets, train larger, more capable multi-modal systems, combine them with highly-accurate expert systems, incorporate expert human feedback, provide sticky user experience, and rapidly grow the user base. In this race for dominance, the developers often overlook the ability to integrate advertising and monetize the link between the customer and the product.
This rapid transition from search to generation coupled with the developers lack of clear guidelines for marketers poses is very confusing to the SOE and SMM marketers.
From SEO to GEO – What Can Marketers Do Today?
In this article, I speculate on some of the 5 steps SOE and SMM marketers can take to transition to Generative Engine Optimization (GEO).
1. Learn the Large Language Models
It is very difficult to optimize for something you do not understand. And understanding how generative systems work is essential. The systems that are most relevant to GEO are LLMs and multi-modal transformers.
I recommend read the first fundamental paper on transformer neural networks by Google titled “Attention is All You Need”. This title is has relevance to marketing as well, even though, in my opinion, is us usually the great product which is all you need.
Here are a few articles for the general audience that you may want to start with:
ChatGPT fundamentals by OpenAI
ChatGPT Explained by Towards Data Science
And for a slightly more advanced users, Andrej Karpathy, one of the heroes of the deep learning revolution, produced a very short youtube video showing how to quickly make your own GPT.
2. Monitor the Key Sources for Training Data
The current-generation of generative engines such as ChatGPT were trained on mostly publicly-available data coming from several corpuses of books, open repositories of articles, and Wikipedia.
Google’s crawling real-time crawling capabilities will provide it with a competitive advantage in its own generative systems development and it is surprising that it did not use its copy of the Internet to train a ChatGPT-like system. Most probably, it is due to the internal firewalls, copyrights and restrictions and the size and formatting of that data. However, it is reasonable to expect it to use this data in the future. But since the new generative tools work very differently from search and present the link functionality, most of the SOE techniques will no longer work. But to be good at GEO, marketers will need to focus on the most data-rich open copyright-free resources. These resources include Wikipedia, the many open-access publishers.
While it is arguably the largest volunteer-created source of human knowledge, Wikipedia has many biases, inaccuracies, and, previously, was not considered to be a reliable source by most academic institutions. Also, the articles that do not attract massive public attention can be easily manipulated by the tight group of senior editors. Ensuring accuracy of content in Wikipedia without breaching the rules, engaging in self-promotion or hiring paid help (even though it is allowed, the senior editors may discredit the edits by a person with the disclosed COI), is a challenge. Hopefully, the rise of generative AI will increase the activity in Wikipedia, improve governance, and improve the quality of content. Expert data curators should get more actively involved in Wikipedia editing and policing. New industry groups ensuring quality content are likely to be created and marketers may benefit by helping build and improve these groups.
3. Establish Working Relationship With the Key Generative AI Vendors
In the near future, Microsoft and other vendors that did not play in the search advertising business due to Google’s near-monopoly on search are likely to expand into generative engine marketing and explore the different advertising models. Following their thought leadership in the digital marketing space and participating in the early pilot projects will increase competitiveness in the future. These early experiments are likely to result in successful generative marketing startups that will inevitably be created in the next few years. By monitoring and collaborating with the generative AI vendors in the early days may help you become one of these successful generative marketing entrepreneurs.
4. Follow the Key Opinion Leaders (KOLs) in Generative AI and in Generative Marketing
While the field of generative engine optimization (GEO) and generative marketing is very new, the top people in the field. I recommend following:
- Sam Altman, the founder of OpenAI
- Andrej Karpathi, deep learning pioneer
- Ian Goodfellow, “father” of the generative adversarial networks
- Andrew Ng, one of the most prominent online AI educators
- Thang Luong, a scientist at Google Brain specializing in NLP
- Yann LeCun, one of the “fathers” of deep learning
I also recommend following KDnuggets and Data Science Central for borderline-technical content.
5. Get Better at Creating Unique Human-Generated Content
The text generation capabilities of ChatGPT are very impressive and it is possible to very rapidly generate human-like text for SEO, SMM, and other marketing purposes. However, most generative engine developers have already developed or are developing tools to detect machine-generated content. This content is likely to be discarded or deprioritized by the search engines, generative engines, publishers, and data providers. For example, Forbes.com already introduced new policies for the contributors prohibiting the use of generative tools for text generation. To prepare for the generative engine revolution it is important to get better at producing unique, valid, human-generated content that is more likely to capture the attention of the deep neural nets. Of course, there is also an opportunity in training the adversarial systems that would produce unique valid machine-generated content that would be indistinguishable from the original human content but I would leave this research direction to the experts. My advice to marketers is to avoid the generative digital pollution as it will decrease the quality of the AI systems in the short term and in bans and restrictions in the future.Consider Pursuing a New Career
With generative systems eating search and providing a more personalized and intimate experience for the users to connect with the products they need, the role of marketers in general may be deprioritized. Instead of pushing the products that consumers do not need it may be a good time to re-think your role in the world and consider creating or contributing to the creation of products and services that will positively impact human lives and the environment.