Marketing Malpractice
The Malpractice
Christensen, Cook and Hall open with a number that should stop every product manager cold: roughly 30,000 new products are launched every year. Over 90% fail. The conventional explanation is poor execution — bad advertising, wrong pricing, inadequate distribution. Christensen's argument is that the failure happens upstream, in the way companies think about markets. The malpractice is in the diagnosis, not the treatment.
The root cause: companies segment markets around products and customers, not around jobs. They build demographic profiles — 35–45 year old males, urban, income above ₹15 lakh — and then optimise products for the average of that profile. The result is products that are mediocre for everyone and ideal for no one.
The core claim. When a customer buys a product, they are not buying a thing — they are hiring something to get a job done. The job is the fundamental unit of analysis. If you understand the job, you can build the product that does it better than any alternative. If you understand the demographic profile of the person hiring the product, you have learned almost nothing useful about what the product should do.
This is not a philosophical point. It is a practical one. Demographic data tells you who bought. It tells you almost nothing about why they bought — and why is the only data point that lets you improve the product, sharpen the positioning, and predict what would cause someone to switch away.
The Milkshake Study
The most famous example in the paper — and one of the most cited anecdotes in all of marketing — is the McDonald's milkshake study. It is worth understanding in full, because it does more explanatory work than any abstract definition of the framework.
What Job Is This Milkshake Being Hired to Do?
McDonald's wanted to increase milkshake sales. They did what any company would do: profiled milkshake buyers (demographic data), asked them what they wanted from a milkshake (survey data), and improved the product accordingly — thicker, chunkier, more flavours. Sales did not move.
Christensen's team took a different approach. They spent a day in a McDonald's observing — not just who bought milkshakes, but when, under what circumstances, and what the buyer did before and after. What they found was counterintuitive: nearly half of all milkshakes were sold before 9am, to people who were alone, who got in their car immediately after buying, and who ordered nothing else.
When they interviewed these morning buyers, the job became clear. The commute was long and boring. They needed something to do with their free hand. They wanted something that would keep them full until lunch. They had tried bananas (too fast, gone in 2 minutes), doughnuts (messy, made fingers sticky for the keyboard), and bagels (dry, required both hands and condiments). The milkshake was thick enough to last the whole drive, fit the cupholder, required one hand, and kept them satisfied. They weren't buying a beverage. They were hiring a commute companion.
The afternoon milkshake buyers were a completely different customer hiring the product for a completely different job — parents buying a treat for children after school, who wanted it to not last too long. Same product, same store, two jobs with opposite product requirements. A thicker milkshake is right for the morning job and wrong for the afternoon job. No single demographic profile captures both.
The implication for product development. McDonald's could not improve the milkshake for "milkshake buyers." There is no such person. But they could improve it for "morning commuters hiring a drive companion" — make it even thicker, put in chunks that appear intermittently to extend the experience, make it available at the drive-through without waiting. And separately, they could improve it for "parents hiring an after-school treat" — smaller sizes, faster to finish, fewer calories. Same category, two products, two jobs, two entirely different improvement paths.
The Jobs-to-be-Done Framework
The milkshake study illustrates the principle. The framework makes it operational. Understanding JTBD requires getting three things right: what a job is, what dimensions jobs have, and how circumstances shape which jobs are active.
What Is a Job?
A job is not a task. It is not a need in the Maslow sense. It is the progress a person is trying to make in a particular circumstance. The circumstance is everything — the same person wants to make different progress at different moments. The morning commuter and the afternoon parent are hiring products to make different progress, even if they are the same person on the same day.
The practical, measurable outcome the customer is trying to achieve. Get to the destination faster. Stay full until lunch. Keep the laptop running through the flight. Most product development focuses here because functional requirements are easy to specify, measure, and test.
The trap: optimising purely for functional outcomes often misses why customers actually make the switch to a new product — which is rarely the functional improvement alone.
How the customer wants to feel while doing the job, and how they want to be perceived by others while doing it. The Harley-Davidson buyer is not hiring a vehicle to get from A to B — they are hiring a feeling of freedom and a social identity as a certain kind of person.
This is where loyalty and premium pricing live. A product that does the functional job adequately but also satisfies the emotional and social dimensions can charge more, retain longer, and create advocates that no advertising budget can replicate.
Why Demographic Segmentation Fails
Christensen's sharpest line. "A 53-year-old man and a 27-year-old woman both buy a Porsche Boxster. Traditional market research would put them in completely different segments. But they're buying for the same job — to feel young and exciting, to reconnect with a version of themselves. The job is the same. The demographic profile is meaningless." The demographic data tells you who bought. The job tells you why. Only why is actionable.
The deeper problem with demographic segmentation is that it creates a self-reinforcing loop of mediocrity. You research a demographic. You build an average product for that demographic. You advertise to that demographic. You attract the average of that demographic. Your data confirms that your product is bought by that demographic. You conclude your segmentation was correct. Meanwhile, you have no idea why the non-buyers in your target demographic didn't buy — because demographic data cannot tell you that.
| What Demographic Segmentation Tells You | What JTBD Tells You |
|---|---|
| Who bought (age, income, location, gender) | Why they bought — the progress they were trying to make |
| Which features they say they want (stated preferences in surveys) | Which features actually cause hiring and firing decisions |
| How to find more people who look like existing customers | How to find more people who have the same job — regardless of demographic profile |
| What the average customer looks like | What specific circumstance triggers the job — and what the competing alternatives are |
| How to defend existing market share against similar competitors | How to identify non-consumption — people with the job who are currently solving it badly or not at all |
Purpose Brands vs Feature Brands
The second major contribution of the paper is the distinction between purpose brands and feature brands — and why purpose brands are the only sustainable source of brand equity.
Purpose Brand
A brand that owns a job in the customer's mind. When a specific job needs to be done and the customer thinks of your brand first — not because of a product feature but because they trust your brand to get that job done — you have a purpose brand.
FedEx owns the job: "I need this to definitely be there tomorrow morning." Not "I need a courier." The certainty of delivery is the job. FedEx charges a premium not because their planes are faster but because their brand has been trusted with that specific job so many times that hiring FedEx has become the default decision.
Why it's defensible: Competitors can match features. They cannot easily take ownership of a job that a brand has claimed through consistent delivery over years.
Feature Brand
A brand that competes on product attributes and specifications. "Our battery lasts 20% longer." "Our detergent removes 12 types of stains." "Our SUV has the highest safety rating in the category." These are feature claims — they can all be matched, exceeded, or made irrelevant by a competitor's next product cycle.
Feature brands are trapped in a perpetual arms race. Each new feature advantage is temporary. The customer has no particular loyalty — they are evaluating specifications on each purchase. The moment a competitor has better specifications, the feature brand loses the sale.
Why it's fragile: In categories where products are good enough on functional dimensions, feature competition shifts to price competition. Feature brands become commodity brands.
Classic Purpose Brands — What Job They Own
| Brand | Job Owned | Why the Job Creates Loyalty |
|---|---|---|
| Disney | "Take my family somewhere magical where everything goes right and my children are completely happy." | The emotional stakes of the job are high. When Disney reliably delivers, the parent feels competent and loving. The alternative — planning an independent holiday that disappoints — is a real risk. Disney charges a massive premium for eliminating that risk. |
| FedEx | "Make absolutely certain this package arrives on time, no matter what." | The job involves high professional risk. If the package is late, the customer faces consequences. FedEx's premium is willingly paid because the customer is not buying delivery — they are buying certainty. No cheaper alternative offers the same certainty. |
| Starbucks | The job is emotional, not functional. The coffee is the prop; the feeling is the product. Starbucks charges 5x commodity coffee prices for this job. The price premium is willingly paid because no one has a powerful enough alternative for the same emotional job. |
The brand extension implication. Christensen argues that brand extensions fail when they try to carry a feature brand name into a new category — because the name carries no job ownership, just product associations. But purpose brand extensions succeed because the brand owns a job, and the extension is simply a new product that does the same job in a different context. Disney's job is magical family experiences — the job travels to theme parks, movies, cruise lines, and merchandise. Each extension reinforces the same job ownership rather than diluting it.
Reading India Through JTBD
The Jobs-to-be-Done lens is particularly powerful for Indian markets because India's consumer evolution has produced a series of category-defining brands that succeeded not by being better products but by doing jobs no existing product was doing at all.
Job: "Give every Indian affordable internet — everywhere."
Existing telecom operators saw the market as "mobile subscribers wanting better data plans." Jio saw the job as "hundreds of millions of Indians who cannot afford to participate in the digital economy." The job was being done badly or not at all. Jio didn't compete with Airtel — it hired itself for a job Airtel wasn't even bidding for.
Job: "Let me give my family the best, and feel like a good provider."
Amul is not hired because it is the cheapest dairy brand or the most scientifically formulated. It is hired because Indian mothers and homemakers have trusted it with the emotional job of feeding their families well. The "Taste of India" campaign doesn't sell a product — it affirms a job well done. This is why Amul has survived seven decades of competition from better-resourced rivals.
Job: "Eliminate the mental load of planning grocery runs."
Bigbasket defined the job as "scheduled grocery delivery" — planned, weekly, price-optimised. Zepto identified a different job: the daily kitchen emergency, the sudden dinner plan, the out-of-stock realisation at 8pm. 10-minute delivery isn't a better version of Bigbasket's job. It's a completely different job. Two competing products, same category, zero overlap in the job being hired for.
Job: "Make this joint so strong it will never, ever come apart."
Fevicol is the canonical Indian purpose brand. Its advertising doesn't talk about adhesive chemistry — it dramatises the job with increasingly absurd demonstrations of permanent bonding. A carpenter hiring Fevicol is not evaluating viscosity specs; they are hiring certainty that the joint will hold. Pidilite's Fevicol has 70%+ market share in a category where technically superior products exist. Job ownership beats product superiority.
The non-consumption opportunity. Christensen's most powerful insight for emerging markets: the biggest competition is often not a rival brand but non-consumption — people who have the job but are currently not hiring any product to do it, either because nothing affordable exists or nothing accessible enough exists. India's growth decade is built on converting non-consumption to consumption: first-time credit card users, first-time health insurance buyers, first-time stock market investors. The brand that defines itself around the job of "make this accessible for the first time" owns the largest available market.
Applying JTBD — The Practitioner's Method
The framework is only useful if it can be applied. Christensen's method for discovering jobs is observational and interview-based — you cannot survey your way to a job. Stated preferences in surveys reflect what customers think they want; jobs reflect what actually drives their behaviour.
| Step | What to Do | What to Avoid |
|---|---|---|
| 1. Observe the circumstance | Watch what people do, not what they say. When do they hire the product? What are they doing immediately before? What are they doing immediately after? The McDonald's team spent a day in the restaurant — not in a focus group facility. | Asking "what do you want from a milkshake?" You will get answers about flavour and thickness. You will not get answers about the commute. |
| 2. Identify the competing alternatives | What are customers currently using to get this job done — including "nothing"? The milkshake's competitors were not other milkshakes. They were bananas, bagels, doughnuts, and the decision to skip breakfast. Your real competitive set is defined by the job, not the product category. | Defining competitors as "other products in your category." This produces defensive thinking about market share rather than expansive thinking about job ownership. |
| 3. Map the functional, emotional, and social dimensions | For the job you've identified, explicitly map all three dimensions. What functional outcome does the customer need? How do they want to feel while getting the job done? How do they want to be perceived by others? The product must address all three to fully own the job. | Stopping at the functional dimension. Most engineering-driven product development does this. The result: a product that works but doesn't build loyalty. |
| 4. Build the purpose brand around the job | Name the brand after the job, not the product. Position every communication around the job. Evaluate every product decision by asking "does this help us do the job better?" — not "does this add a feature our competitors don't have?" | Letting feature competition pull the brand away from job ownership. Every feature arms race is a step toward commoditisation. |