Sridhar Vembu, founder, Zoho, has raised questions about the way companies define workplace discipline, arguing that an overly structured culture can unintentionally discourage employees from taking initiative.
In a recent post on X, Vembu outlined two contrasting organisational models. In one, employees largely follow goals and directions set by senior leadership. In the other, employees are encouraged to make decisions independently, take ownership and challenge one another’s ideas.
According to Vembu, organisations that depend heavily on metrics and clearly-defined targets can gradually create a workforce that waits for instructions from the top. While this approach can make operations more predictable, it can also leave the organisation vulnerable if senior leaders make incorrect decisions.
An initiative-led culture presents a different challenge. Employees are more likely to question decisions and challenge colleagues, which can create disagreements and uncertainty. Vembu illustrated this through software development, noting that programmers often assess each other’s work critically and believe they can improve it.
He argued that companies therefore, face a trade-off between stability and initiative. A highly- structured workplace can provide clarity and certainty but may limit independent decision-making. A culture that strongly encourages ownership can generate more debate and uncertainty.
Vembu cautioned against trying to combine the two approaches without recognising their differences.
His comments come as Zoho itself faces business pressures linked to rising technology costs. Vembu recently highlighted sharp increases in memory prices and AI token costs, saying these pressures were making it increasingly difficult for the company to operate without raising prices.
He has also argued that higher memory costs could force software developers to rethink programming practices. With memory no longer effectively treated as an inexpensive resource, Vembu believes developers and compiler designers will need to place greater emphasis on memory efficiency, including when building AI systems.
