The search engine optimization (SEO) community loves to discuss how new technologies, algorithm adjustments and the like will change everything as we know it. As an industry we’re often looking to the horizon making overly-adamant statements about what tomorrow will look like – most of us are futurists to the fullest.
And while looking ahead is critically important, especially in regards to technology, it can also cause us to lose sight of what’s right in front of us. So, in this article we’re going to ignore the doomsday predictions and hopes for a utopian future and focus on how generative AI is changing SEO today.
SEO Still Isn’t Dead
Every few years an article about how SEO is dead will garner attention. It has been happening since the industry took its first breath. Sure, much of it is just clickbait, but occasionally even respected marketing experts will claim SEO is on its last leg. It should be no surprise then that many are claiming the generative AI revolution will be the death of SEO.
While some SEOs may have an incentive in claiming that their industry isn’t dead, there are numerous reasons to believe that AI isn’t replacing SEO. As Britney Muller of Data Science 101 pointed out in this Search Engine Land article – we’re likely to see the shine diminish as limitations of large language models (LLMs) are observed. To phrase it another way, the more we use generative AI as a society, the more we will see its limitations and the less impressive it will be.
Interest and excitement surrounding generative AI seems to be wearing off less than a year since Chat-GPT’s blockbuster release. Similarweb reported that traffic to the Chap-GPT website (chat.openai.com) dropped by almost 10% globally between May and June 2023. And according to Google Trends data, search interest in “AI” peaked in April 2023 after a meteoric rise and has been trending slightly downward since (screenshot below).
It’s important to remember that generative AI isn’t some magic technology that came to life overnight. Generative AI tools and experiences are built using LLMs models that can perform a variety of natural language processing (NLP) tasks. Once you learn the basics of how these technologies work it becomes apparent that they generative AI is just another tool – not an industry killer.
AI is just another tool
Generative AI is just another tool in our collective technology arsenal. What’s particularly interesting about AI’s influence on SEO though is how it has impacted things on both the front and back end. Generative AI is changing how both search engines and SEOs work.
While tools like Chat-GPT, Google Bard and Bing Chat can be genuinely impressive, they aren’t a magic bullet for society’s (or digital marketers’) problems. Thankfully even our new AI overlords understand their own limitations. This is what Google Bard replied with when asked how generative AI is changing SEO:
“Generative AI is changing SEO in a number of ways, including:
- Creating high-quality content: Generative AI can be used to create high-quality content that is both informative and engaging. This can help to improve a website’s SEO by attracting more visitors and increasing the time they spend on the site.
- Optimizing for long-tail keywords: Generative AI can be used to identify and optimize for long-tail keywords. These are keywords that are less competitive than short-tail keywords, but they can still drive a lot of traffic to a website.
- Personalizing search results: Generative AI can be used to personalize search results for each individual user. This means that the results that a user sees will be tailored to their interests and needs.
- Generating new content ideas: Generative AI can be used to generate new content ideas for websites. This can help to keep websites fresh and up-to-date, which can improve their SEO.
- Creating interactive content: Generative AI can be used to create interactive content, such as quizzes, calculators, and games. This type of content can be engaging and informative, and it can also help to improve a website’s SEO.
Overall, generative AI is a powerful tool that can be used to improve SEO in a number of ways. However, it is important to use generative AI responsibly and to ensure that the content it generates is high-quality and relevant to the user’s search intent. According to Google’s SGE, here are some of the challenges of using generative AI in SEO:
- It can be difficult to ensure that the content generated by generative AI is original and unique. Search engines like Google value original content, so it is important to be careful not to use generative AI to create content that has already been published elsewhere.
- Generative AI can be expensive to use. The cost of using generative AI depends on the specific tool or service that is being used. However, in general, generative AI is more expensive than traditional SEO methods.
- It can be difficult to measure the effectiveness of generative AI in SEO. The effectiveness of generative AI in SEO can be difficult to measure because it is a new technology and there is not yet a lot of data available. However, there are a number of metrics that can be used to track the performance of generative AI, such as traffic, engagement, and conversion rates.
Overall, generative AI is a powerful tool that has the potential to revolutionize SEO. However, it is important to be aware of the challenges involved in using this technology and to use it responsibly.”
As the response above (and countless other resources have pointed out) generative AI can be used for SEO in a variety of ways. Tools like Chat-GPT can help SEOs generate ideas, write content, analyze keywords, optimize meta data and even facilitate link building. As a response from Bing Chat points out, generative AI tools can help SEOs by allowing them to delegate and automate technical SEO tasks leaving them more time and energy to invest in strategic and creative initiatives.
Generative AI can be great for analyzing keywords, writing content, generating ideas and similar tasks. But it’s also important to remember the value of originality and uniqueness. Furthermore, Google values the experience and expertise (the first two E’s in E-E-A-T) of authors as it is conveyed through their content. AI writers can’t really have experience and expertise in the same way human writers can and the performance of their content is likely to be reflective of this.
Generative AI is far from perfect
From this author’s experience most generative AI responses are unnecessarily verbose. Statements and sentences are often repeated or varied only slightly. In some cases, it reads like a high schooler’s essay trying to accomplish a specific word count.
The response above from Bard is also a great example of how generative AI responses have a tendency to contradict themselves. Bard listed “creating high quality content” as the first way generative AI is changing SEO, but also mentions challenges of making AI content original and unique. To most digital marketers, original and unique content are synonymous with high quality, but Bard doesn’t seem to make the connection in this response.
It’s evident that generative AI is causing moral and ethical questions posed by Quick Answers in the past to be resurfaced with more intensity. While harmless hallucinations like saying God drove a Plymouth Fury and Jesus drove a Honda (screenshot below) are bound to pop-up; others are unlikely to be tolerated long.
Sometimes very problematic
It’s common for new technologies to make mistakes and be misused, but some of the responses from generative AI experiences have been incredible concerning.
This article from Tom’s Hardware shows how Google’s Search Generative Experience (SGE) and Bard have touted the benefits of slavery and genocide in addition to listing historical figures like Hitler and Stalin as “great leaders.” One of the SEO industry’s most recognized experts, Lily Ray, has done extensive testing and research to show that Google’s SGE responses aren’t just imperfect – they are dangerous.
It’s very unlikely that Google’s shareholders, advertisers, the general public and regulators will allow “mistakes” like this to continue for long. Generative AI experiences are likely to become a liability to tech companies if they aren’t able to establish better guardrails. And guardrails themselves are likely to diminish the impressive nature of the experiences as well simply by limiting what queries AI results are provided for – which Google SGE has been oddly inconsistent about already.
And oddly inconsistent
In addition to outright hallucinations, it’s likely that opinion-based responses will be viewed as more and more troublesome for advertisers. In many instances Google appears to be inconsistent in their approach to YMYL (your money, your life) search queries – displaying results some and not others. For example, Google SGE provides tons of results for “best car” and related search queries, but nothing for “best car insurance” (screenshots below).
Conclusion
Generative AI won’t be the end of SEO, but navigating these new experiences will be a challenge for businesses and marketers alike. For right now most SEOs are trying to navigate this paradigm shift just like they have before – by doing everything they can to improve their clients’ organic search visibility while protecting their brands.
But how can SEOs really optimize results for their clients in an environment where wild hallucinations are presented as facts? How can our clients’ brands be protected in an experience where Hitler is considered great? How can SEOs hope to compete with bots that know best practices?
Artificial intelligence has the power to transform how humans work and live, but generative AI experiences in their current form seem like an unstable novelty. While AI may help SEOs crank out content, do keyword research, write code etc. it’s unlikely to rival the quality of work done by a human SEO. Generative AI is also unlikely to be loved by users if it continues to struggle to provide more utility than traditional experiences (e.g., SGE over traditional Google Search) while making inexcusable mistakes. Maybe things will be different tomorrow, but this is the state of generative AI & SEO today.