.svg)
Get a Demo
Interested in trying Goodie? fill out this form and we'll be in touch with you.
Oops! Something went wrong while submitting the form.
.svg)
.svg)
Google searches are changing: users are spending less time on search engines, and the majority of searches are zero-click – a query that ends without the user clicking on any of the websites provided as answers.
In many cases, users are turning to other platforms for their searches, like AI-powered large language models (LLMs). Given that LLMs are designed to provide a succinct answer to a search query, users may find that the functionality of LLMs aligns better with their needs compared to a traditional search engine.
The functionality of LLMs aligns with how users are comfortable searching – with natural questions that are simple and straightforward. For example, LLMs can directly answer a question by providing a synthesized answer of the top results it finds online, which is what most people look for from Google. Then, since it can remember the context of the conversation, it can continue to expand on this answer.
To maintain a competitive edge as the theaters in which consumers find products, experiences, and services change, marketers will need to boost brand visibility on LLMs by optimizing their brand’s material to AI search.
While Google lists all potential websites that could help answer a user’s question, LLMs provide a more curated answer and show only the websites that are most relevant to the original query. Brand visibility on LLMs refers to how often your brand shows up for a query that’s relevant to your product or service in the answer that’s curated by the LLM.
Marketers who want to ensure that their brands show up on LLMs in response to certain types of questions must understand why LLMs highlight some brands and not others.
LLM optimization consists of creating content that’s more likely to be mentioned and picked up by LLMs. Content that’s mentioned by LLMs can look like text-based mentions, links, quoted content, statistics, videos, or visuals.
.png)
When we think about traditional marketing techniques such as SEO, LLM optimization is an extension of this – it not only provides a list of websites that can answer a query, like traditional search engines do, but also a recommendation for a certain product, service, or website.
More LLM recommendations can lead to more traffic generated for a website or landing page. By identifying clear goals for traffic directed to a landing page or social media site, marketers can convert this initial interest into more brand awareness. The more people who are aware of a brand and the problems a product or service can solve, the higher the likelihood that this interest will build to more sales or brand engagement.
By investing in LLM optimization now, marketers can begin to drive awareness among people using LLMs and have the advantage of being a first-mover in the space.
Marketers can optimize content for LLMs by evaluating which brands appear in LLM outputs for what search queries, how favorably brands are represented, the negatives of the brand according to LLMs, and the visibility of the brand compared to competitors.
A platform like Goodie can streamline this analysis by tracking brand mentions and competitor performance across multiple LLMs. It consolidates real-time data on sentiment and visibility, helping you identify exactly which content changes or optimizations will yield the most impact for your brand.
Marketers can use this information to optimize their content by analyzing the similarities and differences between their website and competitors who are showing up in a way that they want to replicate on LLMs. Then, marketers can incorporate more elements that their competitors use to get ahead. Blog content with original images, statistics, and cited sources, for example, all help a website show up more favorably on LLMs.
.png)
LLM algorithms pull content from different types of media including text, image, and video as well as content types such as reviews and user-generated content to develop an answer.
Another way to increase the chances that a brand is recommended by LLMs is by inserting text into an information page about the product. According to a working paper called “Manipulating Large Language Models to Increase Product Visibility,” published by Harvard University in September 2024, the researchers explained that vendors can influence which products are recommended by LLMs by inserting a text sequence on the product information page to steer LLMs toward their product page.
Inserting this block of text, which they described as a “strategic text sequence,” can turn a product that was never recommended by LLMs into the most frequently recommended product. Tactically, this involved inserting token sequences into the product information page. This is a form of manipulating the LLM recommendations originally used in adversarial attack algorithms and has potential ethical implications.
PR ensures that a brand is associated with topics that align with their marketing objectives. LLMs recommend content that’s closest to the search topic. They do this by converting the data into tokens that represent words, fragments, spaces, and punctuation.
These tokens are then translated into numeric representations and mapped to a “space” where the angle of cosine similarity between the different tokens in the space is used to judge which tokens are most relevant given a search. This means that it’s essential that websites highlight the words in press releases that match the queries of people who marketers would like to use their product.
Organic reviews can also drive topical awareness as can paid affiliate programs and sponsorships. Although there is an overlap between PR and SEO strategies, PR focuses on a brand’s public-facing persona and SEO elevates a website’s ranking on search engines. Ensuring PR is focused, includes keywords that are easy to optimize around for SEO, and drives to a particular action improves performance in LLMs.
.png)
Certain methods increase visibility in LLMs more than others. For example, a study on 10,000 search engine queries found that quotes, statistics, fluency, citing sources, and technical terms were the top five methods that boosted brand visibility in retrieval augmented generation (RAG) chatbots. Marketers can therefore consider how LLMs perceive brands by reviewing where LLMs view the brand as an authority and what topics the brand wants to show up for.
Collectively, the underlying theme of the methods is that they reinforce brand authority and credibility. The more a website is linked, the more it’s perceived as credible to the LLM, and the more likely the LLM is to cite it.
Beyond the keywords that are common in PR, quotes, and statistics, LLMs also consider entities. Entities are pieces of information that are important to a search. They’re important to complete an action requested by a user.
Working through an example can show in practice what entity extraction looks like. In a search, “buy 3 tickets to New York,” the entities extracted are “3” and “New York.”
Then, marketers can create brand content to build that association. A few brand entity research tools that can be helpful are Google’s Natural Language API, Inlink’s Entity Analyzer, and Ahref’s AI Content Helper.
The LLM that’s the best for building brand visibility is the one that drives the highest amount of traffic to the desired page. It makes the most logical sense that the platform with the highest number of users would drive the most traffic to a page, but this is not necessarily the case. LLMs and search engines favor sites with different attributes in the results pages.
.png)
Among the top three LLMs, they have different benefits and drawbacks for building brand visibility. LLMs developed by OpenAI, for example, allows brands to appear more prominently on search results within their LLM ChatGPT, for example. Google-developed LLMs such as Codey and Chirp have the benefit of carrying the brand reputation and recognizability of Google.
The LLMs also often have more specific purposes than those created by other developers. Codey, for example, is specifically a coding assistant and Chirp is a speech to text model. Finally, Llama can boost brand visibility by generating content that incorporates into an existing brand to increase brand awareness and solidify brand awareness.
By taking market share away from traditional search engines, LLMs are shaping what content looks like and how people engage with traditional search engines. Google introduced AI Overviews, not a traditional search engine nor an LLM, but rather a summary of the top results. Changes, or lack thereof, of search behaviors on Google following May 2024 reveal how people responded to Google’s AI integration.
.png)
The data suggests that the number of mobile searches per searcher ranged between approximately 200 and 250 from January through April then dropped to around 175 in May. In comparison, the number of searches per searcher on desktop remained more consistent between 100 and 150 between January and May.
This may be because users found their answers faster with AI overviews and didn’t need to search as much. Another possibility is that users did not like the addition of AI overviews. More long term data will reveal which of these two possibilities – or if another possibility – is more likely.
To get a better understanding of the number of searches on Google versus ChatGPT, we can look at the number of ChatGPT visits in comparison to Google and project growth rates.
.png)
As of January 6th, 2025, the total number of visits was still dominated by Google at 82.19 billion. ChatGPT in comparison had 3.8 billion page visits. The compound annual growth rate (CAGR) for ChatGPT between its launch date of November 30, 2022, and January 6th, 2024 is 88.84% per year. In comparison, Google’s CAGR is 18.21% per year since its launch on September 4th, 1998.
In reality, Google and ChatGPT’s growth is not linear and an exponential or logistic growth model would make more sense. It’s helpful to start with the CAGR because it uses information directly pulled from Google and ChatGPT and the exponential and logistic growth models include more assumptions.
Likely, ChatGPT is still growing at an exponential rate, whereas Google has neared its maximum capacity and has fully saturated the market. We can therefore estimate that ChatGPT has a lower exponential growth rate of 64% per year than Google at the same point in its timeline, which likely had a higher growth rate, but fewer visits.
It’s impossible for marketers to perfectly predict what the future of ChatGPT and Google will look like. Instead, it’s essential to continue traditional SEO marketing and prepare for changes brought about by LLM marketing.
AI search optimization is complicated by the fact that there are different types of LLMs: self-contained LLMs and retrieval augmented generation (RAG) LLMs. For pre-trained responses, with self-contained LLMs, visibility will not change and links will not consistently show up since the data is already set.
For retrieval augmented generation, marketers can optimize content for each search engine and check that the results from core queries are updated regularly. The shift from self-contained LLMs to retrieval augmented generation LLMs reveals the need for marketers to begin optimizing for their brand to rank favorably on these platforms.
When ChatGPT first came out, for example, it was a self-contained LLM trained on a historical and fixed dataset. At present, most of the models are still self contained, and the ChatGPT model GPT-4 search feature is the only model that can access the internet.
Google’s Gemini and OpenAI’s SearchGPT are two examples of RAG LLMs, which are also referred to as fine-tuned LLMs.
Brands can enhance their brand by using LLMs to streamline processes and analyze large quantities of data. For example, brands can innovate in product development by using LLMs to analyze their data sets. LLMs can also track irregularities in network behavior to stay ahead of security threats. Another benefit is improved content because LLMs can enhance communication and improve business strategy.
How users are engaging on Google is becoming increasingly similar to how LLMs operate. For one, users are looking for answers without clicking through multiple websites, and secondly, users are spending more time on Google to ensure they get the information they’re looking for. However, LLMs are a new space and there are limitations in the way they work.
As technology usage continues to increase, digital literacy becomes even more important. People are spending more and more time on search engines without finding answers because sources are constantly pushing new pieces of information their way. As a result, LLMs are becoming a potential location to find answers more quickly, parse through large quantities of information, and synthesize findings.
More people are able to use digital products and search engines than ever before. With the rise of technology usage, users are spending more time on Google to ensure that they get answers to the questions they’re looking for from reputable sources.
In a survey conducted and processed by the market research partners Scorpion and Dynata published in June 2024, 54% of the participants stated that they browsed search engines for longer periods and looked through more results than they did five years ago.
However, just because people are spending more time on search engines does not mean that they’re able to distinguish between sponsored sources and reputable articles. A study released by the Stanford History Education Group in 2016 found that students struggled to distinguish between reputable and non-reputable sources with more than 80% of those surveyed believing a native ad with the tag “sponsored content” was a news story.
Bringing this back to LLMs, it’s essential that marketers boost their content on LLMs as users are likely to turn to LLMs for fast and reputable responses and trust LLM outputs. 90+% of surveyed students in 2022, for example, use LLMs to increase their understanding. People are putting more trust in LLMs because they consolidate complex information and make it accessible and easy to understand. In addition, people are being selective with what information they use LLMs for and often have a general idea of what kind of response they’re expecting in the output.
There’s an opportunity for marketers to get ahead of the curve and drive brand awareness while also being cognizant of its limitations. Since users are spending more time parsing through information on Google, there’s an opportunity for LLMs to distinguish themselves as search platforms that provide a response quickly while also maintaining quality.
To avoid common errors in AI search optimization, marketers should ensure they’re creating content with both AI and traditional search engines in mind. Content that is thorough, authoritative, and in-depth is more important, and ranked higher in LLMs, than many short and keyword-focused pieces.
Some common errors and how to avoid them include:
LLMs can still make a lot of mistakes, and one issue in this uncharted space is that people are turning to LLMs for searches and not verifying information. By trusting blindly in LLM outputs, users are acquiring incorrect information and harming their trust in LLMs.
One step in the right direction for building brand reputability is with the advent of SearchGPT. In November 2024, OpenAI, the creators of ChatGPT, announced the pilot of SearchGPT. As a search engine built with reputable news partners including The Associated Press and News Corp, SearchGPT was positioned as direct competition for Google.
In comparison, OpenAI trained ChatGPT on historical data. More recently, OpenAI implemented the features of SearchGPT that got the most favorable feedback directly into ChatGPT.
Although it’s still important for marketers to devote time and resources to SEO optimization since Google still dominates the search space, it would be a mistake to only focus on traditional SEO metrics.
Quality content and relevance must be prioritized alongside traditional keyword optimization. Given the conversational nature of LLMs and their ability to understand the intent behind user questions, marketers should make sure to consider how users might phrase their questions and the underlying issues the product addresses.
In the rapidly changing space of LLMs, reading about recent developments in articles like this one is a great way to stay up-to-date on trends. Choose a few suggestions from this article to experiment with, whether it’s doing an audit to avoid common errors or including a strategic text sequence on your website. It’s an exciting time to be in the marketing world, so get ahead of the curve, strengthen your brand, and add Goodie to your AI search optimization toolkit.
As AI continues to reshape how consumers discover and engage with brands, Goodie AI equips marketers with the tools they need to excel in LLM-driven search. Our focus right now is on AI visibility, AI content writing, and AI optimization, but our offerings are growing!
Goodie enables you to monitor your brand’s real-time performance across top LLMs. By tracking both brand mentions and competitor activity in AI-generated answers, you can stay ahead of shifting trends and user behaviors. These insights help you make data-driven decisions to elevate your ranking.
It’s also possible to produce high-impact content tailored for AI-focused platforms with Goodie. By incorporating our “Author Stamp,” Goodie ensures your brand is recognized as the authoritative source, boosting credibility in LLM results. Our platform aligns your brand voice with user-centric prompts, delivering thoughtful, relevant material that meets evolving AI standards.
There are also opportunities to identify key opportunities to refine your content strategy and enhance visibility with Goodie. We evaluate both owned and earned sources across tracked topics, and offer semantic text integrations to help your brand outshine the competition in AI-generated results, strengthening your reputation and authority.
From monitoring brand visibility on new AI search engines to generating content tailored for conversational queries, Goodie streamlines the entire process. By integrating audience personas and prompt variations into our solutions, we empower you to stand out in AI-driven search results, so you can focus on connecting with customers when it matters most.
Get started with Goodie today.
.svg)
.svg)
.svg)
.svg)
