What Kind of Institution Are You Building?
Most founders answer 'what are you building?' with a product description. The deeper question, what kind of institution are you building, gets deferred until it's too late to design the answer deliberately.
Founder, Majhi Group & Majhi OS
The question every founder is eventually asked, "what are you building?", has two answers. The surface answer describes a product. The deeper answer describes what kind of institution you are trying to create. Most founders spend years on the first answer and very little time on the second, usually until the decisions that define institutional character have already been made by default.
I built two things simultaneously: Majhi Group, a retained executive search firm, and Majhi OS, the operational infrastructure for hiring systems. For the first few years, my answer to "what are you building?" described the product: what each did, who it was for, what problem it solved. That answer was accurate and incomplete.
The more honest answer required naming something about institutional intent: what kind of organisation was I trying to build, and what would it need to look like to exist beyond my personal effort and attention? That question is harder, less glamorous, and more consequential than any product decision.
The question most founders defer
Every building decision is also an institutional design decision, whether or not the founder recognises it as one.
Choosing to run every client relationship personally is an institutional design decision, one that embeds the founder as the product. Choosing to write down search criteria rather than carrying them in your head is an institutional design decision, one that begins distributing the organisation's operating logic beyond any single person. Choosing to celebrate a fast close in a team meeting or a high-quality outcome is an institutional design decision, one that begins shaping what the organisation rewards.
None of these feel like institutional design when you're making them. They feel like operational decisions. But the accumulation of operational decisions made without institutional intent is itself an institutional design, just not a deliberate one.
Jim Collins and Jerry Porras, studying visionary companies across 70 years, found that the companies built to last were distinguished not by exceptional product decisions but by clarity of institutional purpose combined with disciplined institutional design. They knew what they were building, not just the product, but the organisation, and they made decisions accordingly.
Two kinds of institutional intent
There are, broadly, two kinds of institutions a founder can build.
The first is a company built around the founder's capability. The product is the founder's judgment. The brand is the founder's reputation. The quality standard is what the founder would accept. This kind of institution can be excellent, even exceptional. But it scales with the founder and it is fragile to the founder's departure. The clients who came for the founder's judgment leave with the founder. The culture that formed around the founder's presence degrades in their absence.
This kind of institution is harder to build in the early years, when the founder's personal involvement is what generates quality, and more durable in every year that follows.
The distinction matters because the decisions that determine which one you end up with are made in the early years, when they feel like operational details rather than architectural choices.
The Majhi Group version of this question
For the first three years of Majhi Group, I was the product. Clients chose us because I was personally involved in every search. The quality was real and it was dependent on my direct attention.
At some point the question became unavoidable: was I building a boutique that would be excellent while I was running it and irrelevant the moment I wasn't, or was I building a firm that would outlast my personal involvement in every search?
The answer required making institutional design decisions I had been deferring. Writing down the criteria for candidate advancement rather than carrying them as intuition. Building a quality gate that existed as a process rather than as my personal review. Developing the search infrastructure that would allow a second search lead to run a mandate with the same standards I would hold.
None of this eliminated my involvement. It changed the nature of it: from being the primary executor to being the person who designed and maintained the system the execution ran through. That is a different job. It is also a much more scalable one.
The Majhi OS version
Majhi OS started as an answer to the same question applied to operational infrastructure. The observation that led to it was simple: the process of running a hiring operation has repeatable, instrumentable components, candidate sourcing, pipeline health, response rates, recruiter load, that should be visible in real time rather than reconstructed after something has gone wrong.
The institutional design question was: was I building a tool that would require constant attention from its creator, or was I building infrastructure that carried its own logic, that could detect what was wrong, surface the right information to the right person, and initiate recovery without requiring me to interpret every situation personally?
Autonomous execution is not just a product feature. It is an institutional design goal: the system should function well enough without its designer that the designer can be building the next level of capability rather than managing the current one.
The question worth asking early
What kind of institution are you building?
Not what product. Not what market. What kind of organisation, with what relationship to your personal involvement, with what design for transmitting its culture, with what accountability architecture, with what decision systems, are you trying to create?
The answer should be specific enough to guide real decisions. If it's only "something great" or "something that matters," it's not specific enough to be useful when the operational choice presents itself.
The founders who build durable institutions are the ones who were asking this question before the answers were expensive, when the organisation was small enough that the design was still clean and the choices were still reversible.
After that window, the institution you're building is mostly determined by the decisions you already made. And most of them will have felt, at the time, like operational details.
See also: Building as a Response to a Place, Building Through Uncertainty, From Service to System
Sources
Jim Collins and Jerry Porras, Built to Last: Successful Habits of Visionary Companies
Harvard Business Review, Building Your Company's Vision (Collins and Porras, 1996)
McKinsey, Corporate Longevity: Turbulence Ahead for Large Organizations
Frequently Asked Questions
What is the difference between building a company and building an institution?
A company is built around a product and a market. An institution is built around a purpose, a set of operating norms, and a design intended to outlast any individual founder or market cycle. Most founders are building companies and calling them institutions. The distinction matters because the decisions you make in the early years, how you structure authority, what you reward, how you transmit culture, whether you design for legibility or efficiency, determine which one you end up with. These decisions are made whether or not you're making them deliberately.
When should a founder start thinking about institutional design?
Earlier than it feels necessary, ideally from the beginning. The decisions that make institutions durable are much easier to build correctly from the start than to retrofit once the company has grown around a different architecture. The authority structure that works at ten people sets the pattern for fifty. The culture that forms in the first year becomes the baseline everything else is measured against. The founders who build durable institutions are typically the ones who were asking institutional design questions when the answers were still cheap to change.
Does institutional thinking conflict with the speed and adaptability early-stage companies need?
Not if it's done right. Institutional design is not about adding bureaucracy; it is about making the right things explicit early so the organisation can move faster later without the founder at every decision node. Written decision criteria allow faster decisions. Clear accountability produces faster problem-surfacing. Deliberately transmitted culture reduces management overhead. The organisations that mistake formalism for institutional thinking are confusing the map for the territory. The institutions that last are not the most rigid ones. They are the ones that preserved their core while continuously renewing their methods.
Did this land? Push back? Add something I missed?
Reply to Manas →Continue Reading
Related writing
EntrepreneurshipBuilding as a Response to a Place
The mission of an institution is shaped by the problem it came from. Kalahandi taught me something specific about gaps between what exists and what's possible, and building has been my answer to that ever since.
EntrepreneurshipHow Incentive Systems Shape Culture
Every organisation has a stated culture and a real culture. The gap between them is not a communication problem. It is an incentive design problem. What you reward is what you get, regardless of what you say you value.
EntrepreneurshipHow to Build a Culture That Doesn't Depend on You
The cultures of founder-led companies feel coherent while the founder is present. Most of them don't survive the founder's departure, or even the founder's distraction. The culture wasn't built. It was performed.