In a recent analysis video from the "AI, SaaS & Agentic Pricing with Monetizely" YouTube channel, pricing strategist Ajit Ghuman breaks down the common pricing model issues plaguing generative AI companies today. Using Synthesia's pricing structure as a case study, Ghuman identifies a critical problem that could be costing GenAI companies revenue and creating poor user experiences.
The Three Key Components of Any Pricing Model
According to Ghuman, a solid pricing model consists of three distinct components:
- Packages (the different plan tiers)
- Pricing metrics (how you'll charge - per seat, gigabyte, minute, etc.)
- Price points (the actual dollar amounts)
"A pricing model has three components: your packages, the metric by which you're going to charge, are you going to charge on a per seat basis, per gigabyte basis, per voice minute, per video minute generated, what have you, and the price point itself," explains Ghuman.
The core issue he identifies is that many generative AI companies are mixing these components together, specifically conflating packaging with pricing metrics, leading to confusion and potential revenue leakage.
The Synthesia Case Study: Where Pricing Goes Wrong
To illustrate this issue, Ghuman analyzes the pricing page of Synthesia, a popular AI video generation tool. Synthesia offers several plans:
- Basic (free): 1 seat, 36 minutes of video per year
- Starter ($18/month): 1 editor, 3 guests, up to 2 hours of video per year
- Creator ($64/month): 2-3 editors, 6 hours of video per year
- Enterprise: Custom pricing with more extensive features
While there are some feature differences between plans (more AI avatars, personal avatars, AI dubbing, etc.), Ghuman points out the problem:
"Between the enterprise and the creator and starter, there really isn't enough feature differentiation when it comes to a generative AI product. The real key difference is this: How many videos do I get to produce?"
The Volume-Based Pricing Problem
The most glaring issue Ghuman identifies is that Synthesia charges $18/month for 2 hours of video (Starter plan) but $64/month for 6 hours (Creator plan). Simple math suggests the Creator plan should cost around $54/month (3 × $18) if the pricing were consistent.
"They are selling two hours for $18 and they are selling six hours for $64. All right. But what if I just multiplied this? It should ideally be 18 into three, right? So really they're charging $64 per month where based on the price of the starter plan, it should only be $54 per month."
This creates a perception of unfairness for customers who primarily care about usage volume. Synthesia might argue they're offering more features in higher tiers, but Ghuman questions whether customers truly value those additional features separately from the core generation capability.
No Overage Fees: A Critical Flaw
Another significant issue is the lack of overage options. Synthesia's policy states that if users exceed their plan limit, usage is capped until renewal. The only option to get more minutes immediately is to upgrade to the next plan.
"If I want just a little bit more usage, I have to get the next plan, which is $64 per month. That is a clear problem, right? If I could spend $25 and you're now making me spend $64 a month, that is no bueno, I'm just going to stop using the service."
This rigid approach forces customers into an all-or-nothing decision, which can lead to lost revenue and user dissatisfaction.
How GenAI Companies Should Structure Their Pricing
Ghuman recommends several fixes for generative AI companies:
- Separate packaging from pricing metrics: Create distinct plans based on customer needs, with consistent volume-based pricing within each plan
- Implement a three-part tariff model: Allow for smooth scaling with fair pricing as usage increases
- Add overage options: Don't force customers to upgrade to an entirely new plan for slightly more usage
- Find real feature differentiation: Identify what truly separates different customer segments beyond just usage limits
"The challenge generative companies face is how do we add in more functionality that is beyond just the products that have been generated through AI? It could be a video editing stack. It could be partnering with other vendors."
Learning from Better Examples
Ghuman points to 11 Labs as a company that handles this pricing challenge better:
"Many companies like 11 Labs solve this much more better. There is less of a jump from plan to plan, and they do offer overage fees and they do let you scale."
Conclusion
The key takeaway for GenAI companies is to carefully separate packaging decisions from pricing metric decisions. When usage is the primary differentiator between plans, pricing should scale appropriately without significant jumps that feel punitive to customers.
As Ghuman concludes, "It is not in Synthesia's interest for you to cap out your $18 per month plan, wait and either wait or then go to the next plan. That doesn't make any sense. It's a poor user experience and it's a lose for both the consumer and Synthesia."
For GenAI companies looking to optimize their revenue and customer experience, rethinking the relationship between plan tiers and usage-based pricing could unlock significant value.