I interviewed for a head of product job at a startup once where the CEO had a laundry list of items he wanted the new product leader to sort out. One of the items on the list was “pricing.”

When I asked him what needed to be addressed concerning the pricing, his answer was profound:

“The only thing I’ve learned about B2B product pricing is that you’ll get it wrong on the first try, and you’ll need to iterate to get it right.”

That quote has stayed with me over the years. Pricing (and packaging) is this squishy domain between product building and customer outcomes that is shared between PM, Finance, and GTM and tends to suffer from being pulled in all directions.

It can be frustrating to feel you have a product-market fit and a scalable GTM motion, only to be hindered by the pricing and packaging being dissonant with your customer’s perception of value.

Now you’re probably wondering…why can’t you just change it? Because changing it inevitably solves the problem for one cohort of customers while creating problems for another. You end up with legacy plans, grandfathered customers, operational overhead, and an all-around cluster that account and success teams end up having to deal with.

One person that I had many discussions with around pricing/packaging was Aaron Levie at Box (when I worked there). One of the interesting learnings we had when I led the developer products business was the perceptions/expectations of buyers changing as you got closer to “the path of revenue.”

Said more clearly, if you’re helping an enterprise be more productive, they look to you on how to think about productivity and translate that into an atomic unit that you can price around. But if you’re in the critical path of their workflow, they are well versed in how their business makes money. Hence, your product’s value is measured solely in terms of lift to that overall metric.

So when I saw this tweet from Aaron, I immediately recognized how he was applying that old thought to a new idea: AI agents.

Here’s the TL;DR:

• Per-seat pricing is the dominant model in B2B SaaS.
• This pushes vendors towards horizontal SaaS to maximize users using your tool.
• AI agents unconstrain the number of people doing the job.

For a second, I thought he was going to propose charging a seat for each AI agent. 😬

But what’s potentially happening in the not-too-distant future is that AI could help reimagine the processes that are happening within an organization in order to accomplish the core business workflow.

Let’s look at an example from my recent session with Anurag Wadehra about Layering AI into Your B2B Product and GTM:

In this example, we walk through how a customer success leader might have a business objective to maintain CSAT as a metric while reducing costs by 50%.

People are using AI today to eliminate redundant tasks and automate repetitive work, which definitely helps with productivity. But with the evolving capabilities of AI agents, you’ll start to see improvements in terms of real-time, relevant responses and eventually, the ability to reflect on and reframe the business goal itself.

In other words, you’ll eventually prompt your AI agent for customer service that you want to “maintain CSAT while reducing costs by 50%”, and it’ll tell you all the ways you might accomplish that – and through prompt iteration, you can arrive at the most viable strategy for your business.

One of the larger themes Anurag and I discussed in the full video is that the human involvement in the feedback loop is what slows down most workflows:

If you’re only thinking about AI as a tool to accelerate your existing processes, that’s a mistake. You’re falling into the trap of using a new tool with your old mindset. When your worldview is your current job to be done, of course, you’re going to think about any new tool as a productivity play.

But AI opens the door to starting with the end business outcome and working backward. And that requires unlearning what previously served your business.

AI is going to unlock a second act for a lot of companies and formalize the last act for many more. A quote from Aaron’s Tweet that drives this home:

“All of a sudden small businesses, under-resourced teams in large enterprises, and all new geographies begin to open up as markets. AI will enable otherwise niche categories of software to become much larger, and already large categories of software to become even bigger.”

So how does this connect back to pricing?

Well, today, as a PM/GTM leader pitching value to a customer, your talk track looks something like this:

• Your company does primary job A to make money.
• To do A, you need B number of employees doing several secondary jobs B1, B2, B3, etc.
• The effort/efficiency of these sub-jobs is measured as C1, C2, C3, etc.
• Our software makes these various sub-jobs X% more efficient.
• This leads to an overall efficiency gain of Y% across your workforce.
• This (insert fuzzy math) results in Z% more money.
• Oh, and let’s not talk about the costs to:
  • Set up, configure, and deploy the software.
  • Train admins on ongoing maintenance.
  • Educate users on a new tool/process.
  • Integrate with other tools in the stack.

Now imagine the ROI in the future with an AI agent as a major part of your product:

• Your company does primary job A to make money.
• Tell us how much more money you want to make.
• On what time horizon.
• With what constraints.
• Here’s a blueprint – how much is that worth to you?

AI is a shortcut into the path of revenue, and a much more direct play. If you can generate tangible value, then pricing is based on a cut of value generated (most pricing today is abstracted away from the business value and more oriented on the internal employee job to be done).

And why do we care about value-based pricing?

Check out this snippet from The Product Compass – some of the most valuable companies in the world (Apple, Tesla) utilize this model. And it has historically eluded B2B software companies…until now.

I expect this trend to first show up downmarket because AI helps unconstrain “small businesses, under-resourced teams”, so startups and lean teams will be more likely to adopt vs companies with large workforces who are not incentivized to disrupt a functional workflow (again, change management is hard). As we know, tools and talent move upmarket over time in B2B. This is what Scott Belsky is calling the era of scaling without growing:

“As smaller companies modernize using a new AI-native stack of technology, they no longer need to grow their team to scale their ambition. What a wild concept, but it will be increasingly true. Small teams will gain superpowers once limited to the world’s largest organizations. As a result, small teams will increasingly be able to run — and compete with — big businesses.”

As Anurag Wadehra said in our session, “90% of the action is in pure cost reduction” but a lot “remains to be seen” because AI has the potential to “eliminate jobs, transform the workflow, and change the buyer”.

Based on my experiences and learnings, and stitching together points of view from B2B experts like Aaron Levie and Scott Belsky, I now have the conviction that:

If AI can get your B2B product closer to the critical-path, revenue-generating workflow of your customer, then you can charge a percentage of the business growth you enable.

You can read the original piece by Ibrahim Bashir in his Substack.

Don’t Miss Out on Expert Knowledge That Keeps You Ahead.

Read More From Ibrahim