Kaz Nejatian on Shopify: Building Complete Products and a Company for the Long Term

Shopify COO and product leader Kaz Nejatian tells Harry Stebbings that product teams must build the right thing, the right way, at the right time. He argues for complete early products, technical fluency and a clear vision, and says remote work requires substantial supporting systems.

Nejatian admires Keith Rabois and Max Levchin for combining kindness with aggressive execution and ambition. Working with them has taught him to remain hungry, he says.

Stebbings contrasts a director-led vision with customer testing and iteration. Nejatian argues that The Lean Startup has been widely misunderstood and that shipping software known to be bad simply to elicit reactions can turn users away. He is not advocating waiting three years to release a product.

An early version can be constrained and still complete, he says. Looking back, creators will naturally see flaws in their past work; what matters is that the version was usable when shipped. He compares incomplete software to a chair with three legs.

Stebbings asks whether a vision can also constrain a founder. Nejatian agrees that visions can change: Shopify founder Tobi Lütke initially wanted to build a snowboard store. But, he argues, a good product begins with a belief about a problem and a desired outcome, even when that belief later changes.

Nejatian identifies two lessons from Lütke. Shopify prioritizes building great products, making money, and using that money to build more great products, without reversing the first two. It also maintains a high quality bar, holding back products that other companies might ship.

That approach can look inefficient over a one-year horizon, Nejatian admits. If the aim is to build something that lasts a century, he believes more investment before shipping can make sense. The answer depends on the time horizon and how future value is discounted.

Stebbings challenges Shopify’s practice of building internal tools across functions, given the cost of maintaining and updating them. Nejatian replies that tools shape how people think and operate. He sees nothing inherently wrong with buying software, but believes a software company benefits from understanding the assumptions embodied in its own tools.

Shopify built an internal prioritization and project-management tool called GSD as well as HR software. Nejatian criticizes annual headcount plans built in spreadsheets for presenting uncertain projections as fixed facts. He says Shopify’s tools instead use internal data and priorities to model the organization continuously.

He praises Rippling as a startup tool that addresses some related problems. In his view, business software often leaves data scattered across products, so teams export CSV files and assemble spreadsheets before making decisions. Shopify prefers an integrated system that reflects its own operating preferences.

Nejatian initially thought building internal software to support remote culture was a terrible use of engineering resources. Lütke overruled him; six months later, he says, the tool had become indispensable. He presents the episode as an example of colleagues disagreeing openly while trying to find the truth.

A learner’s mindset is more than a slogan, Nejatian argues. He believes companies often choose established experience over learning from scratch, and says Shopify accepts different ways of doing things in exchange for people’s ability to learn. Stebbings pushes back that experience and a learning mindset can coexist.

Nejatian concedes that experience can help people learn in fields such as investing. But he argues that many software companies rely heavily on accumulated expertise and unquestioned conventions. In his view, growth companies need to keep questioning their assumptions.

Early in his Shopify tenure, Nejatian proposed putting the careers page behind an API so applicants would need technical skill to access it. He accepts that this was arrogant and is glad the company did not do it. His underlying goal, he says, is to recruit people who are highly capable and aligned with the mission.

Shopify’s marketers create landing pages in GitHub and learn some code, Nejatian says; many salespeople also write SQL. The company is willing to teach those skills, but looks for the desire to learn them. He compares the hiring ambition to assembling a professional sports team.

From his time at Meta, Nejatian says Mark Zuckerberg is underappreciated for understanding and taking risk. He argues that risk-taking should be judged by the quality of the process, not solely by the outcome, and points to major changes in Facebook’s product as examples of willingness to make bold decisions.

Nejatian also defends Meta’s long investment in augmented and virtual reality. He believes Ray-Ban smart glasses show the direction of a future computing platform and predicts future hardware will impress users. These are his judgments and expectations, not established outcomes.

Asked about Meta’s worst decision, he argues that the company did too little to explain and defend itself and Zuckerberg publicly, allowing a damaging narrative to take hold. He also claims former Meta employees have founded an unusually large number of valuable companies; the interview does not independently establish his comparison.

For Nejatian, product management means representing the user’s best outcome in the market. Shopify defines the job as building the right thing, the right way, at the right time. If users are not better off, he says, product managers should take responsibility.

Asked which element he would sacrifice, he chooses the specific “what,” arguing that teams underestimate how and where in the technology stack a product is built. Good architectural choices can make future work possible. Product managers need not all write code, he says, but should understand how it is written.

Stebbings questions whether extensive discussion is useful. Nejatian says talk is expensive because colleagues must spend time listening; Shopify calculates an estimated cost for meetings and includes it in invitations. He favors writing decisions and code down rather than relying on live meetings.

Written discussions also create a record that could be useful for AI systems, Nejatian argues. Stebbings notes that some executives see in-person work as more effective. Nejatian responds that remote work is a poor idea for most companies, in his opinion, and that Shopify makes it work through extensive investment in tools, teams and processes.

He sees this as part of Shopify’s willingness to pursue approaches that may be unsuitable elsewhere. Its business was built around small merchants, a group investors once considered highly prone to churn. He describes two company values as thriving on change and learning rather than relying solely on prior experience.

That philosophy can make ordinary processes slower. When Nejatian joined, he says, Shopify lacked a standard process for requesting a hire, so managers had to work out how to recruit themselves. He sees the extra ownership as useful at Shopify, although it may be difficult for other companies.

The transition to remote work was nevertheless painful. Nejatian says the first six to twelve months did not work well: an intensely in-person organization had lost the cultural practices that made its teams effective. The company deliberately made reversal difficult by relinquishing leases and hiring people far from offices.

Another difficult change gave employees a pool of compensation they could allocate among salary, equity and other choices using software. Nejatian says implementing it required changes to equity arrangements and substantial coding, including work by Lütke. Employees understandably worried during the transition; he believes the result works well.

Shopify can be difficult for analysts to model, Nejatian says, because its growth comes from merchants starting businesses, some of which grow considerably while many fail or remain small. He compares the resulting variation to a portfolio, while noting that the company earns money from different kinds of merchants.

On enterprise customers, he says Shopify previously turned large retailers away rather than selling features that were not generally ready. Its rule against “party tricks” means a demonstrated feature should work beyond a narrow demonstration. He describes the later enterprise expansion as accepting demand after the product matured.

Nejatian distinguishes building software enterprises can use from creating bespoke enterprise software. As merchants grew, Shopify had to meet their needs, including compliance for a company preparing to go public. He says starting with a real merchant problem led to different product choices than starting with a compliance checklist alone.

Stebbings asks why Shopify relies on Stripe while building so much internally. Nejatian says the choice depends on whether external partners share Shopify’s standards and views. Shopify has integrations with more than 700 payment providers, he says, and Stripe is its largest partner.

Nejatian calls the Stripe–Shopify relationship exceptionally valuable and credits both companies with helping each other grow. He cites Affirm, billing partner Milo and Global-e as other partnerships. His assessment comes from a Shopify executive with a direct interest in these relationships.

He argues that software companies often pick unnecessary fights over adjacent markets. Shopify prefers deep partnerships where each side has a clear job, he says, giving banking relationships as an area handled by payment partners. He also describes Shopify’s competition with earlier store software as a conflict of visions for internet commerce.

Nejatian says his most serious disagreement with Lütke concerned going remote, and believes Lütke was right. He does not keep score of whose opinion wins: having an opinion matters, but he says he is wrong often and places less weight on owning the final answer.

He thinks observers underestimate the breadth of Shopify’s product. Rather than a single point solution, he describes it as closer to an operating system and its teams as toolmakers. That breadth, he says, also makes the company harder for outside analysts to model.

On controversial beliefs, Nejatian says he is comfortable acknowledging when he is wrong and regularly shares his mistakes inside the company. His largest mistakes, he says, have involved information flowing through the organizational chart instead of reaching the people who needed it.

In a conventional hierarchy, an idea passes from employee to manager and up the chain before a decision returns. Nejatian prefers people to share information across the company, sometimes allowing a manager to learn about it at the same time as everyone else. That gives opposing views a chance to surface.

Stebbings argues that some messages and morale require more careful control. Nejatian accepts there are exceptions, including legally sensitive matters, but says openness should be the default so employees see both progress and setbacks. He notes that Apple’s greater secrecy is a different approach used by a successful company.

He compares open feedback to teammates communicating during a game. If people know what others are building, they can coordinate instead of duplicating work. To him, values should describe choices where reasonable people can take the opposite position.

Stebbings raises high-profile divorces among very wealthy founders. Nejatian cautions against using such unusual individuals as a template for most people. He also claims commonly cited divorce rates can be misleading because some people divorce more than once; the interview does not establish a specific population rate.

Nejatian argues that research associates marriage, particularly for men, with better outcomes including health and wealth. These are claims made in the conversation; associations alone do not establish that marriage causes those outcomes, and the interview does not provide the underlying studies.

He rejects the idea that marriage necessarily limits professional success, while acknowledging that bad marriages can be very harmful. Relationships take time, he says, but so do other valuable activities. He believes popular culture often makes jokes about marriage that obscure its benefits.

Nejatian adds further claims about married and single men’s finances and personal lives, without presenting the underlying data in the interview. For his own marriage, he says the key is to care for three things: each partner and the relationship itself. With four young children, he and his wife try to have a date night almost every week.

In the quick-fire round, he says hires from Meta have generally worked well in his experience, while switching industries can be difficult for finance and marketing leaders. He has also changed his mind about the gold standard, which he once mocked.

A self-described hacker, Nejatian jokes that he was briefly proud when his young son found a way around an iPad password, before realizing the consequences. He advises early-career product leaders to maximize opportunities to ship products rather than optimizing primarily for promotion.

Asked about the US election, Nejatian says he is not American and prefers checking political developments less frequently rather than treating politics as a daily horse race. He describes himself as a believer in free markets and liberty.

He admires Microsoft and OpenAI’s approach to search and cites ExxonMobil as another example. He attributes what he describes as strong profit growth to production optimization and positioning for a post-pandemic travel rebound; these are his assessments and figures, not independently verified in the discussion.

Nejatian praises Canva’s product and its way of building. He also admires Figma for continuing to innovate through a proposed acquisition that later fell apart. In his telling, maintaining that pace was risky because a completed deal could have required the company to undo some of the work.

Stebbings thanks Nejatian for a conversation spanning product, family and company design, and Nejatian closes by thanking him.