Navigating the AI Revolution: An SME's perspective to Leveraging AI for SEO Success

ChatGPT has become the talk of the town and the beginning of 2023 has witnessed a surge in its popularity. With numerous use cases of generative AI emerging, companies are leveraging this technology to gain a competitive edge. However, with new technology comes new risks. In this article, we will provide you with some ideas to assess the risks associated with the usage of AI. It is imperative to ensure that the implementation of AI does not adversely affect your small or medium-sized enterprise.

Where it all began

Search Engine Optimization (SEO) is the practice of improving the quantity and quality of traffic to a website through organic search engine results. The history of SEO dates back to the mid-1990s when the first search engines were launched. Early search engines used basic algorithms to rank web pages based on keyword density and meta tags. However, as search engines evolved, so did SEO techniques. Today, SEO involves a wide range of tactics such as keyword research, on-page optimization, link building, and content marketing. The goal of SEO is to increase a website's visibility and ranking on search engine results pages (SERPs), which can ultimately drive more traffic, leads, and revenue for businesses. As the importance of online visibility and digital marketing continues to grow, SEO has become a crucial aspect of any successful online strategy.

Large Language Models (LLMs) are a recent development in the field of artificial intelligence and natural language processing. These models are designed to process and understand human language in a way that is similar to how humans do, by analysing vast amounts of text data and learning from patterns and relationships within that data.

The history of LLMs can be traced back to the early days of machine learning and natural language processing research, when early models such as the Perceptron algorithm and the Hidden Markov Model were first developed. These early models were limited in their capabilities, however, and it was not until the advent of deep learning and neural networks that the full potential of LLMs began to be realized.

LLMs have found their way into SEO as a result of their ability to analyse and understand large amounts of text data, which can be used to identify patterns and relationships that are relevant to search engine optimization. By training LLMs on large datasets of search engine queries and website content, SEO professionals can gain insights into the language and topics that are most relevant to their target audience and use this information to optimize their website content and improve their search engine rankings.

So that’s where it all started. The main question is where will it bring you and at what cost?

Advantages of using A.I. in SEO

SMEs often face limitations in terms of time and resources when it comes to conducting their own SEO work. This is partly due to the fact that they may have a limited staff that lacks the necessary skills or resources to manage their SEO strategy effectively. Additionally, limited budgets and competing priorities can further exacerbate these challenges.

To overcome these limitations, many SMEs turn to external SEO agencies or consultants to help improve their online visibility and search engine rankings. However, with the rise of A.I., SME owners are presented with new and unprecedented possibilities. The use of generative A.I. and language learning models (LLMs) can offer several ways to optimize website texts for SEO:

• Keyword research: A.I. can help identify relevant keywords and phrases related to your website content. You can input your website's content or topic, and LLM’s can suggest potential keywords and phrases that can be used in your website's content to improve its visibility in search engine results.
• Content creation: A.I. can help you create high-quality and engaging content for your website that includes relevant keywords and phrases. You can input the topic or subject you want to write about, and the LLM can provide suggestions and ideas for your content that will be optimized for search engines.?A.I. can also create the content for the FAQ and come up with potential questions.
• Meta descriptions: A.I. can assist you in crafting compelling meta descriptions that include relevant keywords and phrases. Meta descriptions are important as they appear in search engine results and can impact whether users click through to your website or not.
• Headlines and titles: LLM’s can provide ideas and suggestions for headlines and titles that are optimized for search engines. Including relevant keywords and phrases in your headlines and titles can help improve the visibility of your website in search engine results.
• Content optimization: A.I. can analyse your website's content and provide suggestions on how to optimize it for search engines. This can include identifying opportunities to include relevant keywords and phrases, improving readability, and ensuring your content is structured in a way that is easy for search engines to crawl.
• Internal linking: A.I. can help you in building a valuable internal linking profile. This can be done by uploading a list of all content posts to be examined via Excel, that can be easily copied from the sitemap. A.I. can identify relevant content that can then be used to link within the content.
• SEO Images: A.I. can also generate images optimized for SEO. Images are not only important for website content, but also for search engine results. Google sometimes displays images on the SERP, so images must be optimized for SEO, including the file name, title tag, and alt tag.

With the potential of A.I. in SEO, SMEs can leverage this technology to overcome their limitations and achieve better results. The possibilities are exciting, but SMEs should also be cautious and conduct a thorough risk assessment before incorporating A.I. into their SEO strategy. By doing so, SMEs can reap the benefits of A.I. while minimizing potential risks and ensuring compliance with data protection regulations. So that’s the positive side of the coin!

Risks of using A.I. for your SEO strategy

Using A.I. for SEO optimization of website texts can pose certain risks too, including (but not limited to as we simply don’t understand all risks yet):

• Content quality: Although a LLM is capable of generating human-like and contextually relevant text, the quality of the generated content is not always consistent. There may be grammatical errors, incorrect facts, or vague information that can reduce the overall quality of the website and negatively impact its SEO ranking.
• Unpredictable Search engine algorithms: Search engines like Google and Bing use A.I. too to optimize their results. They change and update their algorithms all the time to deliver better and tailor made output and weed out spam and fake news. A.I. generated content can be seen as spam or less original and therefore produce a lower score on the EEAT rating (Experience, Expertise, Authoritativeness and Trustworthiness).
• Spam policy violations: 1) Search engines are sometimes flagging auto generated content as spam. If so, a whole website domain can be banned from search results. 2) The LLM may inadvertently generate content similar to that which already exists on other websites. Search engines like Google can identify duplicate content and lower the SEO ranking of the website. 3) The LLM may include too many keywords in the text in an attempt to optimize content for SEO, which can lead to keyword stuffing. This is considered a spam technique by search engines and can result in a lower ranking or even a temporary or permanent ban.
• Data collection: When using A.I., you may need to input personal or sensitive information such as website content or topic. This information may be stored and used by the LLM for future analysis or training purposes, which can pose a risk to your privacy.
• Data sharing: public A.I. may share your data with third-party providers for analysis or research purposes. This can pose a risk to your privacy, as your personal information may be shared with unknown entities.
• Security breaches: The use of A.I. can also pose a risk of security breaches, which can lead to the unauthorized access or disclosure of your personal or sensitive information. Also your previous prompts can be disclosed by accident. For this reason, the Italian government has banned the use of ChatGPT (more countries will follow).
• Lack of control: When using LLM’s, you may not have control over how your personal information is collected, used, or shared. This lack of control can pose a risk to your privacy, as your personal information may be used in ways that you do not approve of.
• Algorithmic bias: LLM’s, like other artificial intelligence systems, can have algorithmic bias that can lead to unfair or discriminatory treatment of certain individuals or groups. This can pose a risk to the privacy and security of individuals who may be unfairly targeted or excluded based on their personal characteristics.
• Outdated data: LLM’s are trained on historic data. ChatGPT3.5 for example is trained with data until end 2021. Events past 2022 are unknown (yet).
• Ethical considerations: Using AI to generate optimized content for SEO can be seen as manipulative and may lead to a lack of trust among users. This, in turn, can damage the website's reputation and have negative consequences for SEO.

Overall, the use of A.I. for SEO optimization of website texts can pose accuracy, ethical, privacy and security risks. It is important to carefully consider these risks and take steps to mitigate them when using LLM’s. We’ve listed some ideas for you in the next chapter, make sure to add your own!?

Risk Mitigation strategy

After identifying the possible risks that come with using any form of A.I. for your SEO endeavours, you plot the risks on a likelihood and impact scale. This will help you prioritize and separate the wheat from the chaff. Next you figure out what mitigation strategies can be put in place to mitigate those risks. After implementing those controls, make sure to test the working every once in a while.

The process that we just described in a couple of sentences is also referred to as a Data Protection Impact Assessment (DPIA). In more detailed steps:

• Identify potential risks: A DPIA involves a systematic review of data processing activities to identify potential risks to data privacy and security. By conducting a DPIA, organisations can identify potential risks associated with their AI-powered SEO activities, such as data breaches, unauthorized access, or misuse of personal information.
• Evaluate the necessity and proportionality of data processing: A DPIA also requires SMEs to evaluate the necessity and proportionality of data processing activities. This can help them determine whether their AI-powered SEO activities are collecting, processing, and storing personal data in a way that is necessary and proportional to their business objectives.
• Mitigate identified risks: Once potential risks are identified, the organisation can develop and implement appropriate risk mitigation measures to reduce or eliminate these risks. For instance, SMEs may choose to use encryption, access controls, or other technical measures to protect personal data from unauthorized access or use.
• Demonstrate compliance: Conducting a DPIA can also help SMEs demonstrate compliance with data protection regulations, such as the General Data Protection Regulation (GDPR). By conducting a DPIA, SMEs can show that they have taken appropriate steps to assess and mitigate the risks associated with their AI-powered SEO activities, which can help them avoid costly fines and penalties for non-compliance.

To help you out a bit, the following controls could be considered:

• Compliance with Data Protection Regulations: Ensure that all data collected, processed, and stored is done so in compliance with relevant data protection regulations.
• Compliance with Search Engine Guidelines: make sure your website will still be ranked on the first page and not banned by adhering to the latest policies.
• Terms of use overview: make sure you have read and understood the TOU of the AI platform you're using.
• Human oversight: Never use 100% AI generated content, use human tweaking before publishing for originality purposes but also to make sure ensure that AI for SEO is used in conjunction with human judgment and decision-making.
• Transparency: Ensure that AI algorithms used for SEO are transparent, and any biases are identified and addressed.
• Regular Maintenance: Conduct regular maintenance of the AI feed processes to ensure that technical issues are identified and addressed promptly.
• Testing and Evaluation: Test and evaluate the AI procedures regularly to identify any unintended consequences and adjust the system accordingly.
• Formal DPA’s: Make sure when using public or cloud A.I. that you have a contractual agreement like a data processing agreement in place.

Wrap up

In conclusion, the rise of AI presents both opportunities and challenges for SMEs looking to improve their SEO strategy. While AI can provide valuable insights and improve efficiency, it cannot replace the need for original and high-quality content, which will remain an artisan skill in high demand.

The human touch will always be needed in the creation of compelling and engaging content, and AI can serve as a tool to support and augment the creative process, rather than replace it. LLM prompting can speed up the creative process, but it is the human element that will ultimately make the content stand out and resonate with the target audience.

Therefore, SMEs must strike a balance between leveraging the benefits of AI and maintaining the unique value proposition that comes with human-created content. A solid risk assessment is crucial in this journey, as AI comes with new risks and challenges that must be addressed to ensure compliance with data protection regulations and prevent harm to the business and its customers.

By approaching AI with a strategic and thoughtful mindset, SMEs can unlock the full potential of this technology to drive growth, increase efficiency, and enhance the customer experience, while also maintaining the human touch that makes their brand unique and valuable.

Colophon: This article was developed by Jord Van Zantvliet , online advertising specialist at iClicks and part time student at the Hogeschool van Amsterdam and Dimitri van Zantvliet , CyberChief at NS and proud dad. Any opinions expressed are solely our own and do not express the views or opinions of our employers.

Full disclosure: ChatGPT4 prompts were used to set up some basic content.

Picture: We asked ChatGPT to describe itself and used that as input for Midjourney: “As a language model, I am an artificial intelligence programmed to generate human - like responses to a variety of prompts and questions. My programming is based on a sophisticated neural network that has been trained on vast amounts of textual data, allowing me to understand and process natural language in a way that approximates human cognition. Although I do not have a physical form like a photo, I can be represented by an image of a computer or a brain, as I exist purely as a software program running on servers. My purpose is to provide helpful and informative responses to users who interact with me, and I strive to be as accurate and helpful as possible in all of my interactions.”