On the discipline's hidden model of the user, and what becomes visible once products are treated as social systems
Product management talks about users as if the category were self-explanatory. In practice, the discipline tends to imagine a very specific kind of user: an individual decision-maker moving through a flow, responding to incentives, evaluating value, and acting more or less alone.
That model is so common it becomes invisible. It shows up in funnels, in onboarding, in activation metrics, in personas, in experimentation, and in the way teams talk about "driving behavior." The user appears as a singular mind in front of an interface, not as a person embedded in social norms, institutions, power structures, communities, or networks.
I think this is one of the deepest blind spots in product management. The field does not only lack sociological tools. It is built on a hidden ontology that makes sociology easy to leave out.
That inheritance makes sense historically. Product management grew at the intersection of engineering and business. Engineering is good at modeling systems. Business is good at modeling markets. Neither was built to model society. Even the social sciences most commonly imported into product work, like cognitive psychology and behavioral economics, tend to focus on individuals making decisions rather than groups producing norms.
Once that foundation is in place, a lot of standard PM language follows naturally. Jobs to Be Done asks what progress an individual is trying to make. Personas simplify social complexity into representative types. A/B testing treats users as if they encounter variants independently. North Star Metrics flatten many experiences into one number and then call the result "value."
None of those tools are useless. They are simply partial. They work best when the problem really is individual. They become misleading when the product is governed by status, legitimacy, audience, network effects, collective meaning, or structural constraint.
You can see the ontology most clearly in the tools PMs trust most. Jobs to Be Done improves on demographics by focusing on progress, but it still centers an individual hiring a product to do a job. Personas compress social life into one representative type. A/B testing assumes independence between users. North Star Metrics pull many unlike experiences into one company legible number and call the result clarity.
Each tool has real use. The problem is what happens when teams forget the conditions under which the tool is truthful. Jobs to Be Done weakens when the product is shaped by identity and structural constraint. Personas weaken when the social world matters more than the fictional average. A/B testing weakens when users affect one another. A North Star weakens when average improvement hides uneven costs or unequal visibility.
This is why I think the issue is not just that PMs need "more empathy." Plenty of product teams care deeply and still misread the system because the toolset itself keeps nudging explanation back toward the individual.
Once you start looking for it, the pattern is hard to miss. A product can be useful and still fail because the act it requires is socially risky. A social platform can look healthy while its norms are quietly collapsing. A recommendation engine can look objective while amplifying power-law dynamics that make visibility more unequal over time. A marketplace can look efficient while pushing most of the uncertainty onto workers.
Standard PM tools often register these problems late and misdiagnose them when they do. The dashboard sees a drop-off in onboarding. The real issue may be audience exposure. The KPI sees engagement. The deeper issue may be community erosion. The product review sees growth. The structural question is who benefits from that growth and who is bearing the cost.
That is why I do not think the gap is simply empathy. Plenty of thoughtful teams care about users and still miss the system. The problem is more foundational than that. If your basic model of the user is individual, then society will keep appearing as edge case, nuance, or externality rather than as part of the product itself.
A sociological lens changes the questions before it changes the answers. In research, instead of only asking what a user wants, you ask what social context shapes what they can plausibly do. What norms govern the behavior? Which audiences matter? What identity is at stake? What power dynamics shape the options available?
There are already promising attempts to widen that frame. The Product Management and Society Playbook, for example, treats product work as inseparable from stakeholders, communities, and public impact rather than as a narrow optimization problem around a single user journey.
In strategy, the questions also widen. What social system is this product entering? What kinds of relationships does it strengthen or weaken? What status signals does it create? What forms of labor does it rely on? What gets easier for some people because someone else is absorbing more uncertainty, effort, or risk?
In measurement, you begin to look for social indicators rather than only individual ones. Are users building connections? Are norms becoming clearer or more unstable? Is trust increasing? Are the same actors capturing all the value and all the visibility? Are there groups for whom the product works differently because their position in the social system is different?
It changes research reviews because the question stops being only "what did the user say?" and becomes "what social conditions made that behavior plausible?" It changes product reviews because the team starts asking whether a drop-off is really about friction or about legitimacy, audience, and risk. It changes growth work because the first meaningful threshold may be social density, not click-through. And it changes strategy because the business model itself becomes part of the product diagnosis rather than something discussed only in finance or policy.
Once teams start asking these questions, some product problems look newly obvious. A community breakdown is no longer just a moderation issue. A recommendation system is no longer just a relevance engine. A marketplace is no longer just a matching flow. The product becomes easier to understand because it is being described at the scale where it actually operates.
The practical change is not that every PM suddenly becomes a sociologist. It is that the discipline stops pretending the individual is the whole story. That alone changes diagnosis. It makes adoption look more social, metrics look more political, and product decisions look more consequential.
It also clarifies why so many modern product failures are so hard to fix with interface work alone. Facebook's sharing problem was not just ranking. Nextdoor's profiling problem was not just wording. Uber's labor model was not just a marketplace UX. In each case, the product was participating in a social system it was not fully naming.
I think of the other Sociology x Product essays as specific consequences of this broader blind spot. Why Useful Products Still Fail looks at social risk. The Product Funnel Starts Too Late looks at the invisible stages before action. Value Flow Is Product Strategy, Not Just Ethics looks at who benefits and who bears the cost. This essay sits underneath all three: the discipline keeps missing those questions because it keeps starting from the wrong model of the human.