Building a viable pricing model for generative AI features could be challenging

In October, Box Inc. announced a new pricing strategy for its generative AI capabilities. Instead of a standard flat fee, the company has developed a unique usage-based model. Every user receives 20 credits each month, which can be used towards various AI tasks equating to 20 events, with a single credit charged for each task. Beyond that, users can access an additional 2,000 credits from a shared company pool. If this is exceeded, discussions around purchasing extra credits with a sales agent would occur. As per Box's CEO, Aaron Levie, this innovative model aims to charge based on usage, acknowledging the variation in AI feature usage among users while also factoring in the cost of utilizing the OpenAI API, their choice for the large language model that underpins the system.

On the other hand, Microsoft opted for a more conventional pricing structure, announcing in November that they would impose a $30 monthly charge per user for the usage of its Copilot features, in addition to the regular Office 365 subscription cost that differs across customers.

The rise of generative AI features within enterprise software companies became evident last year. During a panel discussion in November at the Web Summit on the impact of generative AI on SaaS companies, Christine Spang, co-founder and CTO of communications API startup, Nylas, and Manny Medina, Outreach's CEO, expressed concern over challenges SaaS enterprises could face while adopting this technology. Spang stated, despite the buzz surrounding generative AI, it indeed represents a significant advancement. She added that software organizations should strategically integrate this technology into their offerings, emphasizing that its true potential will be revealed through its interconnection with other systems and applications, eventually driving value in diverse use case scenarios.