Price the Job, Not the Tool
Part 2 - A New Philosophy of Value: Pricing the "Job-to-be-Done"
Chapter 3: From Features to Function: The JTBD Revolution
The failures of the subscription and usage-based models are not failures of accounting; they are failures of philosophy. Both are built on a fundamental misunderstanding of why a customer buys anything. They are rooted in a product-centric worldview that sees the thing being sold—the software, the tool, the feature, the API call—as the source of value. This is a vendor-centric perspective, and as we’ve seen, it leads to a chronic misalignment of interests.
To build a new model, we must first adopt a new philosophy. We must find a new unit of analysis. That philosophy is the Jobs-to-be-Done (JTBD) theory.
The term, often associated with Harvard Business School professor Clayton Christensen, is built on a single, profoundly simple insight: Customers don’t “buy” products or services; they “hire” them to get a “job” done. A “job” represents the customer’s ultimate goal, their objective, the future state they are trying to realize. They don’t want a quarter-inch drill bit; they want a quarter-inch hole. The drill bit is the solution they hire for the job. And if a better solution comes along—a laser borer or a chemical compound—they will fire the drill bit and hire the new solution, without a moment’s hesitation.
This perspective is revolutionary because it is solution-agnostic. It forces you to stop defining your market by your product—a “software company”—and start defining it by the customer’s problem—a “company that helps people achieve a goal.” A horse breeder, a railroad tycoon, and an automobile manufacturer were not in the horse, train, or car businesses; they were all in the business of “traveling from one place to another”. This stability is the key: while solutions evolve and technologies become obsolete, the core human job often remains the same for decades, even centuries.
The Jobs-to-be-Done framework provides the common language, the “Rosetta Stone,” that allows a vendor and a customer to finally speak about the same thing. It moves the conversation away from the vendor’s inputs (our features, our platform, our API) and onto the customer’s outputs (the job, the outcome, the goal).
This is the precise vocabulary we need to price AI.
Let’s apply this lens. A company doesn’t buy an “AI-powered marketing automation platform.” That’s the solution. The job they are hiring for is “generating qualified leads for the sales team.” A hospital doesn’t buy an “AI-powered radiology diagnostic tool.” The job is “accurately identifying tumors to improve patient outcomes.”
This simple reframing has radical implications for pricing.
If a vendor is selling an “AI marketing tool” on a subscription, their incentive is to pack it with features to justify the subscription. If they sell it on a usage-basis, their incentive is to have the AI “analyze” as much data as possible to maximize usage.
But if the vendor is “hired” to generate qualified leads, their incentives are, for the first time, perfectly and mathematically aligned with the customer’s. The vendor now wins only when the customer wins. They are no longer selling a tool; they are selling a result.
The JTBD framework is underpinned by a set of core principles that make it uniquely suited for this new economy:
The “Job” is the Unit of Analysis: The framework shifts the focus from the product, the customer profile, or demographics to the job itself. This is critical. You don’t build a product for “a 35-year-old marketing manager” (a persona). You build it for the job of “creating a go-to-market plan.” This focus on the job provides a stable, comprehensive view of customer needs that isn’t dependent on a fluctuating technology stack.
Jobs Have Functional, Emotional, and Social Dimensions: This is not just about a cold, rational calculation. When a job is performed, the customer also wants to feel a certain way (emotional) and be perceived a certain way (social). When a parent hires a solution to “impart life wisdom to their children” (the functional job), they also want to “feel they are contributing positively” (emotional) and “be regarded as a responsible parent” (social). An AI pricing model that only solves the functional job but ignores the emotional job (e.g., an AI that makes the human user feel incompetent or obsolete) will ultimately fail.
A Job-to-be-Done is Stable Over Time: This is perhaps the most important principle for building a long-term business. Technology is volatile; jobs are durable. The job of “managing personal finances” existed long before spreadsheets, and it will exist long after the next AI model. By anchoring your value proposition and your pricing model to this stable job, you are building a business that is resilient to technological disruption. You are no longer a “spreadsheet company” (vulnerable) but a “financial management company” (durable).
A Market is a Group of People Trying to Get the Same Job Done: This redefines your competition. Your true competitor is not just the other company with a similar AI tool. It’s any solution a customer might hire to get the job done. This could be a rival AI, a legacy software platform, a team of human analysts, or even a cobbled-together Excel spreadsheet. An outcome-based price must deliver a better, more predictable, and more cost-effective result than all of these other solutions.
The JTBD framework, in effect, provides the briefing document for hiring an AI agent. It allows a customer to clearly state: “This is the job I am hiring you for. This is what success looks like. This is how I will measure your performance.”
This is the antidote to the SaaS mirage. We are no longer buying a promise of potential wrapped in a subscription. We are no longer paying for the activity of a tool, measured in tokens. We are entering into a new commercial pact: we are paying for the function.
This shift is profound. It changes the identity of the AI provider from a “vendor” to a “partner.” It changes the sales conversation from a “demo” to a “diagnosis.” And it changes the contract from a “license agreement” to a “performance contract.”
Of course, this immediately raises a critical question. Why is this shift happening now? The Jobs-to-be-Done framework has been around for decades. Why has it not already taken over the software industry?
The answer is that, until now, we were missing the other half of the equation. We had the language to describe the job, but the tools we were selling were still passive hammers. The customer still had to do the work. The software could help you get the job done, but it couldn’t do the job for you.
That has all changed. We are now building, for the first time, a tool that can swing itself. We are moving from passive tools to autonomous agents. And this single, technological leap is the catalyst that makes the “Jobs-to-be-Done” economy not just a theory, but an impending reality.
Chapter 4: From Tool to Agent: Why This Time Is Different
The Jobs-to-be-Done framework is not new. The core concepts were being developed in the 1990s. This raises the critical, obvious question: If this philosophy of value is so powerful, why isn’t the entire software industry already priced this way? Why are we still trapped in the SaaS mirage?
The answer is simple: we had the philosophy, but we lacked the mechanism.
Until now, the technology we sold was fundamentally passive. It was a tool, and a tool, no matter how sophisticated, is an inert object. It has no agency. It cannot, by itself, do a job. It can only assist a human who is doing the job.
This distinction is the single most important technological and economic shift of our lifetime. The revolution of modern AI is not that it’s a “smarter” tool; it’s that it’s not a tool at all. It is the first piece of software that can be accurately described as an agent.
This shift from passive tool to autonomous agent is the catalyst. It is the technological breakthrough that finally makes the “Jobs-to-be-Done” economy a practical, structural, and inevitable reality.
To grasp the chasm between these two concepts, let’s draw a clear line.
A tool is a passive lever for human intent. Think of a spreadsheet, a CRM, or a word processor. A spreadsheet is one of the most powerful tools ever invented, but it does absolutely nothing on its own. It is a structured environment, a digital workbench, that waits patiently for a human to imbue it with purpose. A human analyst must act upon it—provide the data, write the formulas, build the pivot tables, and, most importantly, interpret the results. The analyst is the agent. The spreadsheet is the object. The value is created by the analyst.
Because the tool is passive, the vendor’s responsibility is limited. The vendor of the spreadsheet is responsible for making it available, reliable, and functional. They are responsible for the workbench. The analyst—the agent—is responsible for the outcome.
This clean separation is precisely why the subscription model worked so well. The customer was, in effect, engaging in two separate transactions:
They “rented” the workbench from the software vendor for a flat, predictable subscription fee.
They “hired” an agent (an analyst, a designer, a marketer) to use that workbench to produce an outcome, paying them a salary or a fee for the result.
The vendor sold access. The human sold the outcome. It was a stable, rational, and well-understood economic model.
An agent, by contrast, is an active, autonomous system that executes an objective. An agent is not an object; it is a proxy for the agent. It is a non-human carpenter, analyst, or marketer.
You do not use an agent in the same way you use a tool. You delegate to it. You hire it. You do not give it a step-by-step command (”sum column C”); you give it a high-level objective (”find the three biggest risks in our quarterly sales data”).
This is what modern AI can do. An autonomous AI agent can be given a goal, a budget, and a set of constraints, and it can then perform the job on its own. It can plan, research, write code, execute tasks, analyze the results, and iterate, all without direct human intervention for each step.
Suddenly, the two separate transactions of the old model—renting the tool and hiring the agent—have collapsed into one.
The AI vendor is no longer just renting you the workbench. They are, in fact, providing the agent itself. The “carpenter” is now in the code. The “analyst” is in the algorithm. The vendor is no longer selling a passive tool that assists a human; they are selling an autonomous agent that delivers a result.
And this, right here, is why the old pricing models are now obsolete.
You would never pay a human carpenter a flat monthly “subscription” fee simply for the privilege of having them on retainer, regardless of whether they build anything. You would never pay a human analyst “per-key-stroke” (a usage-based model) as they inefficiently bang away at a problem. You pay them for the job they get done. You pay for the finished deck, the completed audit, the successful campaign.
The AI agent must be paid in the same way.
This technological leap—from tool to agent—solves the single biggest blocker that prevented the JTBD economy from ever taking root: the problem of attribution.
In the old tool-based world, a vendor could never plausibly sell an outcome. If a marketing manager used a SaaS email platform and had a 10% conversion rate, who was responsible for that outcome? Was it the human’s brilliant copy? Or the tool’s “A/B testing” feature? It was impossible to disentangle. The human and the tool were a hybrid team, and the human was the senior partner. The vendor, therefore, had to retreat to the only thing they could definitively prove: access. They charged a subscription, a safe, defensible, and non-committal price for their piece of the puzzle.
The autonomous agent solves this attribution problem. The agent is the service. The human’s role shifts from operator to director. You hire the agent to do the job. The agent either does the job or it doesn’t. The outcome is a direct, measurable, and contractual consequence of the agent’s performance.
Let’s revisit our “generate qualified leads” job.
Old Model (SaaS Tool): A Marketing Manager uses a CRM tool. She builds the lists, writes the copy, and designs the nurture sequence. She is the agent. The tool is the object. She is paid a salary for the outcome (leads). The CRM vendor is paid a subscription for access.
New Model (Autonomous Agent): A Marketing Director hires an AI agent. She gives it an objective: “Generate 100 qualified leads this month, with this budget, for this persona.” The AI agent is the agent. It performs the entire job: it analyzes the market, writes the copy, runs the ads, and delivers the 100 leads.
In this new model, what is the logical price? It is not a subscription. It is not a per-token usage fee. The only rational, aligned, and sustainable price is a fee for the outcome. “We will pay you $100 for every qualified lead you generate.”
This is the future. The vendor and the customer are now perfectly, mathematically aligned. The vendor is incentivized to make their agent as brilliant and efficient as possible, because they only get paid for success. The customer is ecstatic, because they have eliminated all performance risk. They are not paying for activity; they are paying for achievement. They are not buying software; they are buying results.
This is why this time is different. It’s not a philosophical preference; it’s a technological forcing function. The emergence of autonomous AI as a viable commercial product makes the “Jobs-to-be-Done” framework the only one that makes sense. The agent is the service. The job is the contract. The outcome is the price.
Now that we have established this new foundation, we must move from theory to practice. If we are to build a business on this model, we must become experts at one thing: finding the true outcome. Before we can price a “job,” we must be able to deconstruct a customer’s vague, surface-level “need” and drill down to the fundamental, measurable, and priceable outcome at its core. This is not a sales skill; it is an intellectual one. It requires a set of precise, powerful mental tools.
I make content like this for a reason. It’s not just to predict the future; it’s to show you how to think about it from first principles. The concepts in this blueprint are hypotheses—powerful starting points. But in the real world, I work with my clients to de-risk this process, turning big ideas into capital-efficient investment decisions, every single time.
Follow me on 𝕏: https://x.com/mikeboysen
If you’re interested in inventing the future as opposed to fiddling around the edges, feel free to contact me. My availability is limited.
Mike Boysen - www.pjtbd.com
De-Risk Your Next Big Idea
Masterclass: Heavily Discounted $67
My Blog: https://jtbd.one
Book an appointment: https://pjtbd.com/book-mike
Join our community: https://pjtbd.com/join