Summary of Key Points
This news article discusses the downfall of Medallia, a leading customer experience SaaS company. In 2021, it was privatized by private equity giant Thoma Bravo for $6.4 billion. However, just two years later, it lost control of its finances due to a debt crisis, resulting in the nearly complete loss of its $5.1 billion worth of equity. The core reason for this failure was that the AI analysis capabilities developed over 20 years—used to process customer feedback and identify customer emotions—became commonplace features offered by larger models such as ChatGPT, instantly eroding Medallia’s competitive advantage. This incident serves as a warning to all software companies: “Don’t rely on having AI as your core competency; instead, offer something that customers cannot do without, even with AI.”
Detailed Analysis
How Did Medallia’s “AI Advantage” Get Overthrown by Larger Models?
Medallia’s strength lay in its advanced AI analysis engine, which it had spent 20 years building. This engine could process vast amounts of customer data, identify customer emotions, and predict churn, serving many Fortune 500 companies. It was acquired at a high price precisely because this AI technology was considered unique and difficult for competitors to replicate.
However, with the emergence of larger models like ChatGPT and Claude, tasks such as text classification and emotion analysis became easier to perform, more efficiently, and at lower costs. Medallia’s years of investment suddenly made its features seem basic and unnecessary. As a result, it was unable to manage its growing debt and ultimately went bankrupt.
Relying on “Having AI” as a Core Selling Point Is Like Entrusting Your Future to Others
Many software companies now highlight AI in their marketing materials, claiming they offer “AI digital employees” or “automated tasks.” But this can be a dangerous strategy:
- Loss of Pricing Power: As models continue to evolve, the features you showcase today may become standard offerings from larger model providers, reducing the value of your products.
- Customers Care About Results, Not Just AI: Sales managers buy systems to improve conversion rates, not just because they have AI capabilities; customer service teams want to increase satisfaction, not just automated responses. If your product only provides an AI tool, customers will switch to more advanced alternatives when the tool is upgraded.
AI Shortens Your Lead Time, and Customers Can Do It Themselves
In the past, software companies gained a competitive edge through longer development times; for example, if it took a year to develop a new feature, they could monopolize the market during that period. But with AI, competitors can quickly catch up using more powerful models and specialized tools. Additionally, customers can use larger models to develop their own solutions, potentially rendering your investment useless.
To Survive, Software Companies Need to Offer Three Valuable Assets
To continue attracting customers even in the presence of advanced models, companies must offer something that they cannot obtain or manage on their own:
- Infrastructure That Customers Are Unwilling to Rebuild: AI requires a supporting infrastructure (e.g., enterprise systems) to function effectively. Customers are reluctant to replace such essential systems; for example, SAP’s ERP platforms, which restrict third-party AI access, do not fear customer churn because they are core business assets.
- Industry-Specific Knowledge That AI Cannot Learn: Companies possess expertise in areas like identifying valuable leads or resolving sales issues, which larger models cannot replicate. What you sell is not just the ability of AI to generate messages but the knowledge and strategies needed to win deals, which customers are willing to pay for.
- A Service System That Delivers Actual Results: Simply installing a system is not enough; companies need to provide solutions that directly impact business outcomes (e.g., increasing sales or reducing customer support costs). This requires a dedicated team with industry expertise, which model-based companies often lack, and customers are unlikely to develop on their own.
Conclusion
The future of software companies lies in offering something more valuable than just having AI: “Being needed by customers even when they have access to AI.” This is the true competitive advantage in the era of large models.