Pricing as the #1 competitive advantage for AI startups
The emergence of outcome-based pricing
It's been a long time since pricing was as important as it is today.
With the current shift in technology driven by AI, pricing will be a critical factor for both startups and established companies as they adapt their business models to the new landscape.
There are two ways pricing will change the way startups can challenge incumbents (the first is the focus of this article).
Charge for the output: Companies can charge for the output instead of the user (because of AI doing +90% of the work)
Compete on price: Startups can compete on price vs incumbents because of different cost structures
The startup opportunity: Outcome-based pricing or selling the work product
LLMs have enabled startups to create products that can complete entire pieces of work or workflows. This is a massive shift from the past two decades where we charged a subscription fee for a user to access a piece of software.
Whilst incumbents have been slow at adopting this new pricing model we are seeing more and more companies pricing based on the work or task instead of focusing on the end user.
Below are several companies implementing some form of outcome-based pricing:: (to highlight what that might have looked like a pre-AI era I have added that version in grey)
Tying value and pricing closer together is not a new concept. Google, Meta, LinkedIn and the other large ad networks were some of the first tech companies to move towards outcome-based pricing when they changed from charging per impression to per click/conversion.
Why pricing will become the #1 competitive advantage
For certain transactional workflows and tasks, AI enabled solutions are able to complete 90-100% of the work. In instances where the output is primarily driven by the software, and not the human, the traditional pricing model of charging per user does no longer make sense. Instead the price should be based on the value of solving the work and replacing human hours.
Outcome-based pricing aligns itself much closer with value creation for the customer and is advantageous for startups to use for several reasons:
Lower Risk for Customers: Customers are only required to pay if the startup delivers measurable results. This reduces the financial risk for clients, making it easier for startups to acquire new business.
Establishing Trust: Startups that tie their revenue to performance demonstrate confidence in their product. This builds trust faster than incumbents who often rely on subscription or license models that don’t guarantee outcomes.
Flexibility and Scalability: Many incumbents have fixed pricing models that are not adaptive to the rapidly changing needs of customers. Startups can offer flexible, scalable solutions that allow businesses to pay only for what they use or what they achieve.
The revenue potential is 15-30X higher with outcome-based pricing than with per user pricing
When you complete a full task or offer a solution that a human would have previously handled, you can price it relative to the cost of the human labor. Instead of charging based on a 10-30% productivity improvement, you can charge 25-50% of what it would cost for a human to perform the same task.
Lets use EvenUp as an example. They are company which helps injury lawyers by creating demand packages. If they had created a “SaaS product” they would have created a solution that would have enabled the lawyer through some interface to complete their work product faster. The problem with this is that the productivity increase would have been limited to 10-30% and they would not have been able to price the product relative to a human doing it.
I’m not sure about EvenUp’s exact pricing model, but here’s an example comparing the revenue potential when pricing based on the user versus pricing based on the work. The first line represents pricing for the user, and the second line represents pricing for the work. You can imagine an injury lawyer handling 50-100 demand packages a year (or potentially even more).
If outcome-based pricing is so good - why are incumbents not doing it yet?
I foresee incumbents will struggle to adopt new business models like outcome-based pricing for several reasons:
Revenue Cannibalization: Shifting from subscription or license-based pricing to outcome-based models might result in lower immediate revenue, as customers will only pay for proven results. Incumbents are often reliant on predictable cash flows from their existing pricing models, and changing this can disrupt their financial stability.
Complex Organizational Structures: Large companies typically have rigid internal structures and long-standing contracts with customers. Making a shift in pricing requires buy-in from multiple departments, including sales, finance, and customer service. This complexity makes incumbents slower to adapt compared to agile startups.
Resistance to Change: Incumbents often resist changing their pricing models because of the risks involved, such as alienating existing customers or failing to accurately measure outcomes. As seen in Zendesk's experience, changing pricing can lead to customer pushback if not managed properly.
Incentive Structures: In large organizations, sales teams and executives are often incentivized based on metrics like upfront sales revenue, making it difficult to transition to a model that emphasizes long-term performance. This creates friction in implementing outcome-based models, which are more focused on achieving results over time.
In the coming years and months, as AI permeates through their products, I expect to see increasing pressure on incumbents to change their business and pricing models. I expect them to slowly start this transition by offering new features and products priced based on usage and/or outcomes rather than users.
It will be interesting to see if they can adapt quickly enough and are willing to cannibalize their existing revenue in order to compete with newer, much leaner competitors than they've encountered in the past.
🙌 George Boretos has been saying this as well…but is struggling to get startups to listen.
Keep up the good work Palle 👏
Nice 👌