AI Pricing Strategies for SaaS Companies Offering Copilots
Pricing an AI product will be a defining question in software for the next few years. AI products offer productivity gains. But greater productivity may reduce the demand for seats over time, ultimately decreasing the size of software markets.
We can observe the market trends today across some of the larger SaaS companies who offer AI pricing.
The table above lists the company ; the product ; the base price per-seat for the enterprise plan if available, otherwise the team plan ; then the price for the AI or co-pilot add-on ; and finally the ratio between the AI price and the base price.
Sometimes the price is hard to compare, but I’ve tried to do my best to create a fair comparison.
Plotting the ratio illustrates the variance in the market today. Google charges more for their AI features than the base seat. While Loom charges about a 33% premium.
领英推荐
There’s no relationship between a more expensive seat & a greater ratio of the AI add-on. The R-Square is 0.08 : no correlation at all.
Overall, I’d characterize the ecosystem as iterating. OpenAI and GitHub launched their features at roughly $20-30 per month. This initial pricing has anchored the market at least for now in that range.
Microsoft & ServiceNow have stated AI features increased productivity by approximately 50 percent. If buyers act rationally & reduce headcount by 50%1 which we know is probably not true, then to maintain the same revenue per customer, price would need to double. We can observe that in three of the companies’ pricing strategy above.
If pricing really does provide information (see the work of Mauboussin), then these companies are pricing in a 40% productivity gain.
This is for copilots. Agents, which fully automate work or at least claim to fully automate work, may have more disruptive pricing. Instead of hiring a sales development rep, hire a robot. I’ll write about that in tomorrow’s post.
1 It’s highly debatable whether this will happen. Most companies will likely leverage efficiency gains into more growth, but let’s consider the downside scenario.
Iterate 50% faster. Turn customer data into structured output. CEO @ BuildBetter.ai
4 个月Strong believer in not just usage-based pricing but usage-based trials over time-based trials: https://www.dhirubhai.net/pulse/stop-using-time-based-trials-spencer-shulem-ai-prod--ezsge/ Companies should WANT usage-based pricing because our incentives are aligned with the value a company gets. Per-seat pricing's incentives are to maximize seats, often by adding value which is often tracked by usage. But when usage means lower margins, companies prefer lower usage but higher seats. Bad incentives. With usage based pricing, you only use the product if it provides value, which means the way only we get paid is when and if we can provide more and more value to you by using our product. It's a win win.
CTO @ LIMU | AI chatbots for Digital Health
4 个月Arnout Ulenberg
Post Grad in Chemistry
4 个月Pricing an AI product is indeed a defining question in software for the coming years. While AI products promise significant productivity gains, these gains may paradoxically reduce the demand for seats over time, potentially shrinking software markets. Observing market trends in larger SaaS companies that offer AI pricing can provide valuable insights into this dynamic. As we navigate these changes, it's crucial to stay informed and prepared. For personalized interview preparation, check out interviewjarvis.com. It can be an invaluable tool in your career journey.
Chief Revenue Officer (CRO) at Sentry (sentry.io)
4 个月Maybe they can sustain ACV, but margin? Based on my understanding of how these systems work, with the exception of Google and Microsoft/Github, for today's tasks, everyone just acts as a pass-through to one of the core LLMs. In the mid-term that means zero margin for them as competition heats up. Differentiation comes from a rapidly commoditizing RAG loaded up with some level of "proprietary" data (your support ticket history, your customer interaction history,...). Existing vendors will try to wall off that data as much as possible, but their customers will see through it and get frustrated. So my take is that pricing power will be eroded drastically for incumbents who will have a hard time adjusting their per seat pricing models and are dependent on 3rd party LLMs/RAGs without differentiation. Those that work around it and find ways to accomplish tasks in innovative ways, without the constraints of legacy, will have a better shot at capturing the gold.
Principal Director of Subscription Strategy @Zuora & Start-ups investor/mentor >> Enabling companies to shift to the Subscription Economy in every industry
4 个月Tomasz Tunguz - excellent - you might be interested in this research I recently published : https://www.zuora.com/resource/the-4-genai-monetization-avenues/ There are also benchmarks on price ratios when GenAI is sold as a "super tier" (and of course as an add-on) Fyi this is the 2nd installement of a series of articles, benchmarking 70+ offers -- starting here : https://www.zuora.com/resource/monetizing-gen-ai-why-saas-companies-are-missing-out/