Sign up to save your podcasts
Or
Share You Don't Know What Your Customers Are Actually Trying to Do
Share to email
Share to Facebook
Share to X
Share to email
Share to Facebook
Share to X
Or
You Don't Know What Your Customers Are Actually Trying to Do
Uphoff on Media | Field-Tested Frameworks
When I became President and CEO of ThomasNet.com, I inherited a research project already underway with a firm called Strategyn. They were conducting over 300 individual interviews with Engineers, Procurement professionals, and MROs, all focused on the jobs to be done around custom manufacturing sourcing.
The core value proposition of ThomasNet was supplier discovery and custom manufacturing sourcing. The goal of the research was simple: give us user insights to inform our product roadmap, and surface intelligence we could use when marketing to our customers.
We thought the number of distinct jobs would be relatively small. Maybe a dozen. Perhaps two dozen if we stretched it.
We were stunned.
Strategyn’s research surfaced over 200 primary jobs to be done, and another 150 secondary ones. They ranged from the mission-critical: an Engineer searching for a certified manufacturer to produce a key component for a medical device, where the wrong supplier choice could cost lives. To the granular and human: a Procurement professional simply trying to “get my global team all on the same page” about a potential new supplier.
Two hundred primary jobs. One platform. One team that had been building features based on what we thought mattered.
That research didn’t just inform our roadmap. It demolished it. We went back to first principles and redesigned and reengineered the entire ThomasNet platform around the actual jobs of our audience. It was clarifying in the way that only being genuinely wrong can be.
And here’s what happened next: active registrations grew. Platform usage grew. Revenue followed. The JTBD research became the north star of every product decision we made, not a one-time exercise but a living operating framework. It ultimately led to an unsolicited offer to buy the company at an eye-popping multiple.
I’m not telling you this to brag about an exit. I’m telling you because the causal chain matters. Understanding what your customers are actually trying to accomplish — with rigor, specificity, and humility about how wrong you probably are — is not a soft strategic exercise. It is the work. And the market eventually prices it that way.
The Framework Behind the Revelation
Jobs to Be Done is a theory of innovation developed and popularized by Clayton Christensen, the late Harvard Business School professor and author of The Innovator’s Dilemma and Competing Against Luck. Christensen’s core argument: customers don’t simply buy products. They “hire” them to accomplish a specific job.
The analogy that crystallized it for me came from HBS marketing professor Theodore Levitt:
"People don't want to buy a quarter-inch drill. They want a quarter-inch hole."
I had spent decades selling drills. So had nearly every executive I knew.
Strategyn, the firm we’d engaged at ThomasNet, had built their entire research and consulting methodology around this framework, operationalizing it into a rigorous process for uncovering and prioritizing the full landscape of customer needs. Not what customers say they want. Not what your product team believes they need. What they are actually trying to get done, in the specific circumstances they face, with the social and emotional dimensions that feature checklists never capture.
The Milkshake Story: JTBD in Action
Christensen’s most famous illustration involves something far more ordinary than industrial sourcing, a fast-food milkshake. A major quick-service chain was puzzled: the bulk of its milkshake sales occurred before 8:30 in the morning, purchased by solo customers who ordered nothing else and drove away alone. Traditional market research offered no useful explanation.
Christensen’s team asked a different question. Not “what do you think of the milkshake?” but “what job did you hire it to do?”
The answer was immediate: these customers faced a long, tedious commute. They needed something to keep them occupied and stave off hunger until mid-morning. Something consumed one-handed while driving, lasting long enough to make the drive bearable, without the guilt of a donut. They’d tried bananas (gone too fast), bagels (messy), and coffee (too hot, over too quickly). The milkshake was nearly perfect. It took almost thirty minutes to finish, fit the cupholder, and kept them full.
The chain had been asking how to make a better milkshake. The real question was: what is a 7am commuter actually trying to get done?
Two entirely different questions. Two entirely different answers. And only one of them leads to a decision that actually moves the business.
What the Research Did to My Thinking
After the ThomasNet experience, I took Christensen’s Harvard online course on innovation. The aha moment wasn’t just intellectual, it was visceral. I realized that I, like most business leaders, had been building and positioning products based on my understanding of features and benefits. Not my customers’ experience of them. Two entirely different things.
Christensen observed that firms have never known more about their customers, yet their innovation processes remain hit-or-miss. The reason: product developers focus too much on building customer profiles and looking for correlations in data rather than on the job the customer is trying to get done.
The evidence is brutal at scale: somewhere between 75 and 85 percent of all new products launched don’t succeed financially. Not because of poor execution. Because of a fundamental misread of what the customer actually needed.
Jobs are multifaceted. They’re never simply about function. They have powerful social and emotional dimensions. And the circumstances in which customers try to do them are more critical than any buyer characteristic.
This is why demographic segmentation so often leads companies astray. You can know everything about who your customer is and still have no idea what they’re actually trying to accomplish.
Here’s the Christensen point I think is most underappreciated: if you frame your business in terms of products you’re trying to sell, you get supplanted as products and technologies change. But if you’re organized around delivering a job that’s genuinely done well, you can absorb new technologies as they emerge rather than being displaced by them.
Jobs are stable. Products are not.Print that out. Put it on the wall of every product team you run.
The Agentic AI Accelerant
Here’s why all of this has become urgent in a new way.
We are entering the agentic AI era: where autonomous AI systems don’t just assist human buyers, they act on their behalf. They search, evaluate, compare, negotiate, and in some cases complete purchases without a human ever visiting your platform, reading your copy, or clicking through your feature list.
Analysts project that 90% of B2B buying will be AI agent-intermediated by 2028, driving over $15 trillion of B2B spend through AI agent exchanges.
When I first encountered that number, I sat with it for a while. I've watched several major platform transitions up close: print to digital, trade directories to industrial marketplaces, on-premise software to SaaS. Each one scrambled the competitive map for companies that didn't see it coming. This one is different in speed and scope. And it has a specific implication for JTBD that I don't think enough leaders are tracking.
When a human buyer visits your platform, you have the opportunity to guide, persuade, and educate. You can tell your story. A skilled sales team can build relationships and reframe a conversation.
When an AI agent evaluates your platform on behalf of a buyer, none of that happens. It doesn’t read your positioning statement. It doesn’t respond to your brand narrative. It evaluates your product against the specific job the buyer has tasked it to complete. Either your product does that job — demonstrably, specifically — or the agent moves on.
The AI agent is a pure JTBD machine.
These agents are goal-oriented, context-aware, and ruthlessly efficient at matching products to jobs. Feature lists won’t matter. Brand stories won’t matter. What will matter is whether your product demonstrably accomplishes the job the buyer agent has been assigned.
This is why JTBD isn’t just a useful framework anymore. In the agentic era, it becomes the foundation on which competitive survival is built.
If you don’t know the precise jobs your customers are trying to get done — down to granular specificity — AI-mediated buying will route around you. Not maliciously. Automatically.
What This Means If You’re Running a Business
The implications run across the entire organization.
Product: Most roadmaps are answers to questions nobody asked, built on internal assumptions dressed up as strategy. In an agentic buying environment, the gap between what you built and what the buyer needs becomes fatal. Not because a competitor beat you. Because an AI agent simply moved on.
Marketing: Messaging built around your product’s attributes will increasingly fail to reach AI agents making autonomous decisions. The language of outcomes, specific, measurable jobs your product enables, is the only language agentic systems will evaluate. If your content strategy doesn’t speak to jobs, it’s speaking to no one.
Sales: The human selling motion isn’t going away, but it’s narrowing to where relationships and judgment genuinely differentiate. Everywhere else, AI agents will mediate. Sales teams need to understand the jobs their customers are executing at a depth most have never been asked to develop.
Strategy: The businesses that will thrive in the agentic era are those that can map their capabilities directly to the jobs of their market, with specificity and evidence. Not “we help engineers source suppliers.” But: “we help an engineer with aerospace certifications find a titanium CNC machining partner in North America within a 48-hour sourcing window, with documented quality controls for FAA compliance.” That’s a job. That’s a product. That’s a business.
Do the Research
When our team at ThomasNet discovered 350+ distinct jobs to be done in what we thought was a well-understood market, the response could have been paralysis. Instead, it became a forcing function for discipline. Not every job could be served equally. We had to prioritize, sequence, and build a roadmap anchored in actual user needs, not internal assumptions.
That’s hard work. It requires real investment in understanding your market at a level of depth most companies never attempt. Christensen observed that for most companies, innovation remains a flawed business process — yielding failure rates consistently over 80% — because companies fail to define their customers’ needs with the rigor, precision, and discipline required to discover, prioritize, and capitalize on growth opportunities.
We are not in an era where that failure rate is acceptable. The competitive margin is too thin. The pace of AI adoption is too fast. And the buyers, increasingly assisted or replaced by AI agents that will simply go find a better match, are too empowered to wait for you to figure it out.
The quarter-inch drill metaphor has never been more relevant. And the urgency to understand the hole has never been more acute.
We rebuilt ThomasNet around that understanding. The market noticed. So did the acquirers.
The question is whether you move before the agents do.
The views expressed in Uphoff on Media are entirely my own. They don’t represent the opinions of any company I’ve led, any board I’ve sat on, or any investor who’s had the pleasure of debating strategy with me over the years. If something I write here sounds brilliant, I’ll take full credit. If it turns out to be wrong, I was clearly misquoted by myself.
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit tonyuphoff.substack.com
View all episodes
By Tony UphoffUphoff on Media | Field-Tested Frameworks
When I became President and CEO of ThomasNet.com, I inherited a research project already underway with a firm called Strategyn. They were conducting over 300 individual interviews with Engineers, Procurement professionals, and MROs, all focused on the jobs to be done around custom manufacturing sourcing.
The core value proposition of ThomasNet was supplier discovery and custom manufacturing sourcing. The goal of the research was simple: give us user insights to inform our product roadmap, and surface intelligence we could use when marketing to our customers.
We thought the number of distinct jobs would be relatively small. Maybe a dozen. Perhaps two dozen if we stretched it.
We were stunned.
Strategyn’s research surfaced over 200 primary jobs to be done, and another 150 secondary ones. They ranged from the mission-critical: an Engineer searching for a certified manufacturer to produce a key component for a medical device, where the wrong supplier choice could cost lives. To the granular and human: a Procurement professional simply trying to “get my global team all on the same page” about a potential new supplier.
Two hundred primary jobs. One platform. One team that had been building features based on what we thought mattered.
That research didn’t just inform our roadmap. It demolished it. We went back to first principles and redesigned and reengineered the entire ThomasNet platform around the actual jobs of our audience. It was clarifying in the way that only being genuinely wrong can be.
And here’s what happened next: active registrations grew. Platform usage grew. Revenue followed. The JTBD research became the north star of every product decision we made, not a one-time exercise but a living operating framework. It ultimately led to an unsolicited offer to buy the company at an eye-popping multiple.
I’m not telling you this to brag about an exit. I’m telling you because the causal chain matters. Understanding what your customers are actually trying to accomplish — with rigor, specificity, and humility about how wrong you probably are — is not a soft strategic exercise. It is the work. And the market eventually prices it that way.
The Framework Behind the Revelation
Jobs to Be Done is a theory of innovation developed and popularized by Clayton Christensen, the late Harvard Business School professor and author of The Innovator’s Dilemma and Competing Against Luck. Christensen’s core argument: customers don’t simply buy products. They “hire” them to accomplish a specific job.
The analogy that crystallized it for me came from HBS marketing professor Theodore Levitt:
"People don't want to buy a quarter-inch drill. They want a quarter-inch hole."
I had spent decades selling drills. So had nearly every executive I knew.
Strategyn, the firm we’d engaged at ThomasNet, had built their entire research and consulting methodology around this framework, operationalizing it into a rigorous process for uncovering and prioritizing the full landscape of customer needs. Not what customers say they want. Not what your product team believes they need. What they are actually trying to get done, in the specific circumstances they face, with the social and emotional dimensions that feature checklists never capture.
The Milkshake Story: JTBD in Action
Christensen’s most famous illustration involves something far more ordinary than industrial sourcing, a fast-food milkshake. A major quick-service chain was puzzled: the bulk of its milkshake sales occurred before 8:30 in the morning, purchased by solo customers who ordered nothing else and drove away alone. Traditional market research offered no useful explanation.
Christensen’s team asked a different question. Not “what do you think of the milkshake?” but “what job did you hire it to do?”
The answer was immediate: these customers faced a long, tedious commute. They needed something to keep them occupied and stave off hunger until mid-morning. Something consumed one-handed while driving, lasting long enough to make the drive bearable, without the guilt of a donut. They’d tried bananas (gone too fast), bagels (messy), and coffee (too hot, over too quickly). The milkshake was nearly perfect. It took almost thirty minutes to finish, fit the cupholder, and kept them full.
The chain had been asking how to make a better milkshake. The real question was: what is a 7am commuter actually trying to get done?
Two entirely different questions. Two entirely different answers. And only one of them leads to a decision that actually moves the business.
What the Research Did to My Thinking
After the ThomasNet experience, I took Christensen’s Harvard online course on innovation. The aha moment wasn’t just intellectual, it was visceral. I realized that I, like most business leaders, had been building and positioning products based on my understanding of features and benefits. Not my customers’ experience of them. Two entirely different things.
Christensen observed that firms have never known more about their customers, yet their innovation processes remain hit-or-miss. The reason: product developers focus too much on building customer profiles and looking for correlations in data rather than on the job the customer is trying to get done.
The evidence is brutal at scale: somewhere between 75 and 85 percent of all new products launched don’t succeed financially. Not because of poor execution. Because of a fundamental misread of what the customer actually needed.
Jobs are multifaceted. They’re never simply about function. They have powerful social and emotional dimensions. And the circumstances in which customers try to do them are more critical than any buyer characteristic.
This is why demographic segmentation so often leads companies astray. You can know everything about who your customer is and still have no idea what they’re actually trying to accomplish.
Here’s the Christensen point I think is most underappreciated: if you frame your business in terms of products you’re trying to sell, you get supplanted as products and technologies change. But if you’re organized around delivering a job that’s genuinely done well, you can absorb new technologies as they emerge rather than being displaced by them.
Jobs are stable. Products are not.Print that out. Put it on the wall of every product team you run.
The Agentic AI Accelerant
Here’s why all of this has become urgent in a new way.
We are entering the agentic AI era: where autonomous AI systems don’t just assist human buyers, they act on their behalf. They search, evaluate, compare, negotiate, and in some cases complete purchases without a human ever visiting your platform, reading your copy, or clicking through your feature list.
Analysts project that 90% of B2B buying will be AI agent-intermediated by 2028, driving over $15 trillion of B2B spend through AI agent exchanges.
When I first encountered that number, I sat with it for a while. I've watched several major platform transitions up close: print to digital, trade directories to industrial marketplaces, on-premise software to SaaS. Each one scrambled the competitive map for companies that didn't see it coming. This one is different in speed and scope. And it has a specific implication for JTBD that I don't think enough leaders are tracking.
When a human buyer visits your platform, you have the opportunity to guide, persuade, and educate. You can tell your story. A skilled sales team can build relationships and reframe a conversation.
When an AI agent evaluates your platform on behalf of a buyer, none of that happens. It doesn’t read your positioning statement. It doesn’t respond to your brand narrative. It evaluates your product against the specific job the buyer has tasked it to complete. Either your product does that job — demonstrably, specifically — or the agent moves on.
The AI agent is a pure JTBD machine.
These agents are goal-oriented, context-aware, and ruthlessly efficient at matching products to jobs. Feature lists won’t matter. Brand stories won’t matter. What will matter is whether your product demonstrably accomplishes the job the buyer agent has been assigned.
This is why JTBD isn’t just a useful framework anymore. In the agentic era, it becomes the foundation on which competitive survival is built.
If you don’t know the precise jobs your customers are trying to get done — down to granular specificity — AI-mediated buying will route around you. Not maliciously. Automatically.
What This Means If You’re Running a Business
The implications run across the entire organization.
Product: Most roadmaps are answers to questions nobody asked, built on internal assumptions dressed up as strategy. In an agentic buying environment, the gap between what you built and what the buyer needs becomes fatal. Not because a competitor beat you. Because an AI agent simply moved on.
Marketing: Messaging built around your product’s attributes will increasingly fail to reach AI agents making autonomous decisions. The language of outcomes, specific, measurable jobs your product enables, is the only language agentic systems will evaluate. If your content strategy doesn’t speak to jobs, it’s speaking to no one.
Sales: The human selling motion isn’t going away, but it’s narrowing to where relationships and judgment genuinely differentiate. Everywhere else, AI agents will mediate. Sales teams need to understand the jobs their customers are executing at a depth most have never been asked to develop.
Strategy: The businesses that will thrive in the agentic era are those that can map their capabilities directly to the jobs of their market, with specificity and evidence. Not “we help engineers source suppliers.” But: “we help an engineer with aerospace certifications find a titanium CNC machining partner in North America within a 48-hour sourcing window, with documented quality controls for FAA compliance.” That’s a job. That’s a product. That’s a business.
Do the Research
When our team at ThomasNet discovered 350+ distinct jobs to be done in what we thought was a well-understood market, the response could have been paralysis. Instead, it became a forcing function for discipline. Not every job could be served equally. We had to prioritize, sequence, and build a roadmap anchored in actual user needs, not internal assumptions.
That’s hard work. It requires real investment in understanding your market at a level of depth most companies never attempt. Christensen observed that for most companies, innovation remains a flawed business process — yielding failure rates consistently over 80% — because companies fail to define their customers’ needs with the rigor, precision, and discipline required to discover, prioritize, and capitalize on growth opportunities.
We are not in an era where that failure rate is acceptable. The competitive margin is too thin. The pace of AI adoption is too fast. And the buyers, increasingly assisted or replaced by AI agents that will simply go find a better match, are too empowered to wait for you to figure it out.
The quarter-inch drill metaphor has never been more relevant. And the urgency to understand the hole has never been more acute.
We rebuilt ThomasNet around that understanding. The market noticed. So did the acquirers.
The question is whether you move before the agents do.
The views expressed in Uphoff on Media are entirely my own. They don’t represent the opinions of any company I’ve led, any board I’ve sat on, or any investor who’s had the pleasure of debating strategy with me over the years. If something I write here sounds brilliant, I’ll take full credit. If it turns out to be wrong, I was clearly misquoted by myself.
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit tonyuphoff.substack.com