AI and the Big Five: Product-Led Growth's Failure

AI and the Big Five: Product-Led Growth's Failure

In the world of technology, there are two types of innovations: sustaining technologies and disruptive technologies. Sustaining technologies improve the performance of established products in a way that aligns with mainstream customer values. On the other hand, disruptive technologies introduce completely new market innovations that often come from new entrants in a market.

The Internet and cloud computing are examples of disruptive technologies. The Internet created new companies that disrupted industries, particularly those involving information. Cloud computing, although part of the Internet, deserves its own category because it was extremely disruptive. However, it is essential to note that the core infrastructure for cloud computing was primarily built by established companies like Amazon, Microsoft, and Google.

The mobile industry was also disruptive, dominated by Apple and Google. Apple introduced a new UI paradigm, while Google focused on using mobile as a moat for their advertising business. This shows that disruptive innovations can come from both startups and incumbent companies.

Smart companies understand the importance of commoditizing their product's complements to increase demand and charge higher prices. By lowering the price of complements, companies can attract more customers and make more profit. This strategy is particularly evident in large companies investing in open source software, as they recognize the value in supporting complementary products.

Apple's efforts in AI have been mainly proprietary, using traditional machine learning models for recommendations and photo identification. However, they received a gift from the open-source world with Stable Diffusion, a small but powerful model. Stable Diffusion can be deployed on-device, protecting user privacy, eliminating the need for an internet connection, and reducing server-related costs for developers.

This advancement in AI technology could lead to built-in image generation capabilities in Apple devices, benefiting both Apple and independent app makers. Centralized image generation services and cloud providers may face challenges in this new landscape.

Amazon, like Apple, uses machine learning across its applications. However, its direct consumer use cases for image and text generation are not as obvious. Amazon's prospects in the AI space will depend on the usefulness of its products and its ability to compete on price. While AWS is a major partner for Nvidia's offerings, Amazon could also develop dedicated hardware for AI models.

One challenge with AI is the cost of inference, making it challenging to achieve product-market fit. OpenAI's ChatGPT, a breakout product, was successful because it was free for end-users and had a sweetheart deal with Microsoft for compute capacity. The long-term solution lies in building probabilistic models that understand target audiences and ad conversion rates.

Meta, formerly Facebook, has faced challenges with its advertising business due to changes like Apple's App Tracking Transparency. However, Meta's investment in AI can lead to better targeting, recommendations, and revenue growth. This massive investment sets Meta apart from its competitors and deepens its moat.

Meta's AI also requires ongoing customization, which can be costly. However, the use of AI in content recommendations and experimentation aligns well with Meta's advertising goals.

Google, as a leader in AI and machine learning, faces challenges with its business model. The company's search and assistant products have difficulty incorporating ads effectively. While Google has been successful in cramming more ads into search results, the peak of this strategy seems clear.

Microsoft, on the other hand, is well-positioned in the AI space. Its cloud service sells GPU and is the exclusive cloud provider for OpenAI. By incorporating ChatGPT-like results into its productivity apps, Microsoft can leverage its subscription business model and gain market share.

The future of AI may involve open source models proliferating in text and image generation. This could lead to AI becoming a commodity, impacting individual companies economically. Nvidia and TSMC are likely to be the biggest winners in this scenario.

In the story of Qualtrics and SurveyMonkey, the former's success can be attributed to its full-stack approach and focus on experience management. While SurveyMonkey stayed focused on survey software, Qualtrics built a complete solution that addressed customer outcomes. Qualtrics' investment in sales and marketing, particularly in Utah, played a significant role in winning the market.

Product marketing was crucial for Qualtrics' success, as their positioning empowered their sales teams to win over customers. Even if SurveyMonkey replicated Qualtrics in every way, they would still struggle because they are perceived as "just surveys."

In conclusion, AI and product-led growth have their successes and failures. Companies must understand the market dynamics, commoditize complements, invest in sales and marketing, and focus on delivering outcomes to succeed in the AI landscape.