Why Search – and SEO – Is Important

Why Search – and SEO – Is Important

In 2009, Bill Gates made a statement regarding the future of search, stating that it is all about verbs. And he couldn't have been more accurate. Today, search is not just about keywords; it has transformed into a means for users to solve problems, accomplish tasks, and "do" something. It's about guiding users from the start of their journey to the completion of their task. Search has become a crucial element in providing the right content to the right user at the right time.

The Evolution of Generative AI

Generative AI has experienced remarkable growth, driven by its novelty and a wide range of use cases. Already, we can identify three product categories that have surpassed $100 million in annual revenue: image generation, copywriting, and code writing. However, despite this growth, many companies in this field struggle with retention, product differentiation, and gross margins. Infrastructure vendors have emerged as the primary beneficiaries, capturing the majority of the market's revenue. While application companies grow rapidly, they face challenges in maintaining customers and achieving product differentiation. Model providers, although responsible for the existence of this market, have not yet achieved significant commercial scale.

The Role of Hosting Services and APIs

In the realm of generative AI, it is becoming increasingly evident that commercialization is closely tied to hosting. The demand for proprietary APIs is growing rapidly, with companies like OpenAI witnessing significant growth. Additionally, hosting services for open-source models, such as Hugging Face and Replicate, are emerging as valuable hubs for model sharing and integration. These services even foster indirect network effects between model producers and consumers. Considering the potential benefits and risks associated with generative AI, many model providers have adopted a public benefit approach, incorporating the public good explicitly into their mission. Surprisingly, this has not hindered their fundraising efforts.

The Flow of Money in Generative AI

Although generative AI applications generate substantial revenue, a significant portion ultimately flows through to infrastructure companies. On average, app companies allocate around 20-40% of their revenue to inference and per-customer fine-tuning. This expenditure is typically directed towards cloud providers or third-party model providers, who, in turn, dedicate approximately half of their revenue to cloud infrastructure. As a result, it is reasonable to estimate that 10-20% of total revenue in the generative AI market goes to cloud providers. Behind the scenes, Nvidia has emerged as a major winner, reporting billions in data center GPU revenue, a significant portion of which stems from generative AI use cases. Infrastructure has proven to be a lucrative and durable layer in the generative AI stack.

The Elusive Defensibility in Generative AI

When considering the generative AI landscape, it is challenging to identify structural defensibility throughout the stack. Traditional moats, such as scale, supply-chain, ecosystem, algorithmic, distribution, and data pipeline moats, exist but are not necessarily durable over the long term. It remains uncertain whether strong, direct network effects will take hold in any layer of the stack, and a long-term, winner-take-all dynamic is still uncertain. However, both horizontal and vertical companies are expected to thrive, with the approach dictated by the specific end-markets and end-users. If the AI itself is the primary differentiating factor in the end-product, verticalization is likely to prevail. On the other hand, if the AI is just one component of a broader feature set, horizontalization is more probable.

Actionable Advice for Success in Generative AI

1. Focus on customer retention and product differentiation: While rapid growth is essential, it is equally important to prioritize customer retention and develop unique features that set your product apart from the competition. This will help ensure long-term success in the generative AI market.

2. Embrace hosting services and APIs: Recognize the growing demand for hosting services and proprietary APIs. By leveraging these platforms, you can easily share and integrate models, ultimately driving more value for your customers and your business.

3. Explore verticalization and horizontalization strategies: Assess your end-market and end-users to determine whether verticalization or horizontalization is the best approach for your generative AI business. Consider the primary differentiating factor of your AI and align your strategy accordingly.

Conclusion

The importance of search and SEO cannot be overstated. It has evolved from a mere keyword-based approach to a means of guiding users through their journey and helping them accomplish their tasks. Similarly, generative AI has witnessed impressive growth, with infrastructure vendors reaping the most significant benefits. However, challenges persist for application companies and model providers. By focusing on customer retention, product differentiation, and leveraging hosting services and APIs, businesses can navigate the generative AI landscape and position themselves for long-term success.