Will Salesforce’s stance on agent pricing frame AI pricing generally?

Steven Forth is a Managing Partner at Ibbaka. Connect on LinkedIn

Mark Benioff, founder and CEO of Salesforce redefined the software industry with his ‘no software’ positioning back in the naughts (the first decade of this century). In doing this he helped create what we now think of as the SaaS industry.

Now he is shaking up the software industry again with his position on software agents and how they should be priced.

Most SaaS, even today, relies on per-user pricing. What will happen as AI agents begin to replace users in key functions? The more successful the software the fewer the users for that specific function. If people continue with pricing as usual, user-based pricing, revenue could collapse. Salesforce is faced with this situation as are many other SaaS companies that are introducing generative AI applications.

Investors are going to worry about the number of seats being used coming down [because of] those productivity gains," the analyst said. On the earnings call, Marc Benioff said that Salesforce would respond by introducing a new charging model.

He said his personal goal – rather than official guidance – was that by the end of fiscal 2026, the company would introduce 1 billion AI agents into user environments through what Salesforce calls its Agentforce AI platform.

"Let's just think: it's like a manifestation of all these agents talking to all these consumers. When we look at pricing, it will be on a consumption basis … we think about saying to our customers, it's about $2 per conversation," the CEO said.

"It's a very high margin opportunity, as you can imagine … this is how your customers are going to be connecting with you in this new way, and we're going to be helping our customers to manage these conversations.

Quoted from “Salesforce mulls charging per AI chat as investors sweat over fewer seats” The Register August 29, 2024

Salesforce is betting its future on Agentforce, an agent ecosystem, with a goal of deploying one billion agents in 2025! See “Salesforce Introduces Agentforce Partner Network — The World’s First Agent Ecosystem.”

What will it mean for the rest of us as Salesforce moves to consumption pricing?

What will Salesforce’s strategic decision to move to consumption-based pricing mean for the rest of us? We asked the community with a LinkedIn poll that ran on Sept. 30 and Oct. 1. The poll was shared across personal networks, the Professional Pricing Society, Design Thinking, Product Development Management Association, and the impressively large Artificial Intelligence, Deep Learning, Machine Learning group (2,235,000 members). Most of the responses, 90/152 came from this last group.

We asked “How will Salesforce agent pricing impact other pricing?” with the possible answers being:

  • Shape pricing for AI agents - 42%

  • Shape pricing for AI apps - 19%

  • Only impact direct competitors - 20%

  • Only relevant to Salesforce - 19%

Ibbaka Value & Pricing Blog - Salesforce consumption pricing LinkedIn Poll

N = 152 Data collected through LinkedIn poll Sept. 30 and Oct. 1, 2024.

Note that more than 60% of respondents feel that this will shape pricing one way or another. It will do this in two ways:

  • Giving a big boost to consumption (usage based) pricing - Ibbaka predicts that virtually all companies selling agents will include some form of consumption/usage pricing by the end of 2025

  • Setting an anchor point for agent pricing of US$2 per action - this does NOT mean that companies should use this as a price point, most should not, but it does anchor the price perception of buyers and they will have this at the back of their mind as they come to a willingness to pay (WTP)

Note that Salesforce is behind companies like Zendesk and Intercom that are using outcome-based pricing for Agent interactions (only pay if a customer issue is resolved - see AI applications edge towards outcomes-based pricing).

We also probed with the following prompt on Perplexity.

“What has Salesforce announced at Agentforce regarding pricing and monetization?”

Link to the Perplexity thread.

Salesforce's decision to adopt a consumption-based pricing model for AI agents is likely to have significant impacts on the AI pricing landscape, especially for B2B applications. Here are some key ways this move could influence the industry:

Setting a Benchmark

Salesforce's $2 per conversation pricing for AI agents sets a reference point that other companies may use when determining their own pricing strategies

As a major player in the CRM and enterprise software space, Salesforce's approach will likely be closely watched and potentially emulated by competitors.

Driving Adoption

The consumption-based model could accelerate AI adoption by reducing upfront costs and allowing customers to scale usage as needed

This may pressure other vendors to offer similar flexible pricing to remain competitive.

Shifting Away from Seat-Based Pricing

Salesforce's move signals a potential shift away from traditional per-seat licensing models for AI capabilities

. Other B2B software companies may follow suit, especially if they see customer demand for more usage-based options.

Emphasizing Value and Outcomes

By tying pricing directly to AI agent usage, Salesforce is putting focus on the tangible value and outcomes delivered

This could push the industry towards more outcome-based or value-based pricing models for AI applications.

Impacting Vertical AI Applications

For B2B vertical AI applications, Salesforce's approach may serve as a template. Companies offering industry-specific AI solutions may adopt similar consumption-based models, potentially priced based on industry-relevant metrics or outcomes.

Encouraging Transparency

The consumption model requires clear tracking of AI usage, which could lead to greater transparency in AI pricing across the industry

. This may benefit customers by providing more visibility into the costs and benefits of AI implementations.

Potential for Hybrid Models

While Salesforce is emphasizing consumption-based pricing, we may see other companies adopt hybrid models that combine elements of subscription and usage-based pricing to balance predictability and flexibility

Challenges to Address

Salesforce and other companies adopting similar models will need to address potential issues like bill shock and difficulties with purchase orders for variable costs

How they navigate these challenges could influence broader industry practices. It's worth noting that not all companies will necessarily follow Salesforce's exact approach. For example, Oracle recently announced AI enhancements to NetSuite at no additional cost, demonstrating that there are multiple strategies being explored in the market. Overall, Salesforce's move is likely to accelerate the trend towards more flexible, consumption-based pricing for AI capabilities in the B2B space. This could help frame generative AI application pricing around actual usage and value delivered, rather than static licensing fees.

Steps to take today to prepare for the future of pricing

  1. Find out where your customers and ICP are on generative AI and agent adoption.
    Ask ”What is driving adoption?” and ”What is blocking adoption?”

  2. Update your value model for the new value created by AI (no new value, hmm, you may have a problem)
    Are there new value drivers? Are existing value drivers strengthened? Can value be extended to new customer segments or markets?

  3. Connect usage (consumption) and value by finding a consumption metric that correlates with value delivered to customers

  4. Communicate your value and how the new pricing better connects price and value improving outcomes for everyone

  5. Document the value that your AI agents and applications are delivering

Contact Ibbaka today to discuss how Ibbaka
can help you to transform your pricing

Previous
Previous

B2B SaaS discounting: Should you discount fixed or variable pricing metrics?

Next
Next

The Crucial Role of Value Models in Value-Based Selling