The Future of Search: Generative AI and Evoked Sets

The Future of Search: Generative AI and Evoked Sets

In today's digital age, where information is at our fingertips, search engines have become an indispensable tool for finding what we need. For decades, we have relied on search engines like Google to provide us with relevant results based on our queries. However, the way search engines work has remained largely the same since the late 1990s, and it's time for a change.

"The next Google search engine will be Generative AI." This statement challenges the current paradigm of search engine design and suggests that there might be a better way to search. The existing search engine model is based on using a large database and searching for specific keywords or phrases. But with the advancements in technology, particularly in the field of generative artificial intelligence (AI), there is an opportunity to revolutionize the way we search.

The content we consume has evolved significantly over the years. We now interact with social networks, data streams, video platforms, e-commerce websites, and authoritative knowledge sources like Wikipedia. These diverse forms of content require a different approach to search. Instead of relying on a traditional search algorithm, we can leverage the vast amount of data available to us as training data for a neural network.

Trained models, powered by generative AI, are capable of generating results based on the provided data. These models are incredibly efficient, requiring minimal storage space compared to the massive training data they rely on. For example, Stable Diffusion, a generative AI model, is only around 2 gigabytes in size, while the training data it uses is a staggering 100 terabytes. This highlights the potential of generative AI in transforming the search experience.

Imagine a world where instead of searching for something and sifting through multiple search results, users can simply generate the answer they are looking for. This revolutionary approach would bypass the distribution monopoly and advertising business of current search engine incumbents. While training a generative AI model may be expensive, the marginal cost of running it, i.e., conducting a search, is negligible. This shift could disrupt the existing advertising landscape and open up new possibilities for businesses and users alike.

Now, let's shift our focus to the concept of "Evoked Sets." In an article titled "これからは「一番最初に思い出してもらえるブランド」しか生き残れない｜池田紀行@トライバル｜note," the author discusses the importance of being the first brand that comes to mind when consumers think about a specific product or service.

In today's highly competitive market, where product differentiation is increasingly challenging, being able to secure a position in consumers' evoked sets is crucial. The evoked set refers to the subset of brands that consumers consider when making a purchase decision. According to Miller's research, the human brain can accurately rank and remember around 5 to 9 stimuli, indicating the limited capacity for recall and decision-making.

The key to success lies in being the brand that consumers recall first within their evoked set. This first position holds significant advantages, as it becomes a shortcut in the decision-making process for many individuals. The "law of double jeopardy" further reinforces this idea. It states that brands with a higher market share tend to have more customers, and these customers exhibit higher levels of both behavioral and attitudinal loyalty.

For market leaders, strategies that focus on minimizing the number of brands in the evoked set or maintaining the first-choice position through extensive advertising campaigns can be effective. By reducing the number of brands in the evoked set, companies like Kiwi achieve a competitive advantage. Others invest heavily in advertising to ensure they maintain their position as the first brand that comes to mind.

Interestingly, the concept of evoked sets aligns with the Zero Moment of Truth (ZMOT) theory, which states that consumers make a majority of their purchase decisions before even stepping foot in a physical store. This means that the battle for a place in consumers' evoked sets and the first-choice position is often determined before they visit a store. The power of branding and strategic positioning becomes evident in this context.

To summarize, the future of search lies in leveraging generative AI to provide users with the ability to generate answers rather than sifting through search results. This shift could disrupt the advertising business of search engine incumbents. Simultaneously, the concept of evoked sets highlights the importance of being the first brand that comes to mind when consumers think about a particular product or service.

Here are three actionable pieces of advice for businesses and individuals:

1. Embrace Generative AI: Stay ahead of the curve by exploring the possibilities of generative AI in your industry. Consider how this technology can enhance the search experience for your users or customers.

2. Focus on Brand Recall: Invest in branding strategies that aim to secure a position in consumers' evoked sets. Be the brand that comes to mind first, and differentiate yourself from competitors in the minds of your target audience.

3. Optimize the Zero Moment of Truth: Recognize the power of pre-purchase decision-making. Prioritize your marketing efforts to reach consumers before they even step foot in a physical store. Ensure your brand is part of their evoked set and positioned as the first choice.

In conclusion, the combination of generative AI and the concept of evoked sets presents an exciting future for search and branding. By embracing these concepts and incorporating them into our strategies, we can adapt to the evolving digital landscape and stay ahead of the competition.