Vaibhav Sahgal on Connecting Users to Value and Building Growth Teams

Harry Stebbings welcomes Vaibhav Sahgal, Reddit’s consumer product leader and former head of growth, to discuss product and growth work. Sahgal describes his path as partly accidental: he began as a software engineer at the social network hi5 before moving into product management at Zynga and later joining Reddit.

At hi5, an email showed users pictures of people they might know. Sahgal made the pictures smaller so more could fit and ran what he later recognized as an A/B test. He recalls that click-through rose 20%. Seeing an idea tested against actual behavior made him eager to keep learning, although he says growth is broader than testing alone.

Zynga became a kind of product-management school for him. His most important lesson was to understand users in depth: their lives, motivations and reasons for returning. While working on Words With Friends he learned that many users were older women who saw the game as exercise for their minds, not people with his own interests.

The team used a representative persona, “Mary from Ohio,” to question whether a proposed feature would serve the actual audience. Sahgal says research was only the starting point. He also found users in the product, spoke with them regularly and learned how the game fit into their lives; that made the persona more useful than a generic demographic label.

One habit he wanted to leave behind was burning out distribution channels. In Zynga’s early Facebook era, developers could send huge numbers of notifications, and he says the team spammed users. Facebook eventually changed the platform rules. His account illustrates how short-term usage can damage a channel’s long-term value.

At Reddit, he tried to prevent a repeat by imposing a notification guardrail. In his account, a push needed at least a 10% click-through rate before the team could increase its volume; otherwise it had to improve targeting or value. This was a specific historical operating rule, not a universal benchmark or a complete measure of user well-being.

Users have different channel preferences and tolerance for messages. Sahgal argues for maintaining several channels and personalizing both frequency and channel choice rather than assuming every user responds the same way. A useful channel can be exhausted if the team treats its current effectiveness as permission to send indefinitely.

He prefers the term “value connection” to “growth.” A product already contains value, and growth work connects that value to the right person so they engage and return. Onboarding, email and push notifications often serve this purpose. He rejects the idea that growth is merely a short-term hack or that only a growth team cares about company growth.

At Reddit, notifying someone about a relevant subreddit connects existing content to a user. Building a new real-time chat feature would instead create a new type of value, in his example. Both can help a product grow, but they involve different questions and responsibilities.

For an early startup, Sahgal generally recommends hiring a dedicated growth leader after product-market fit, while being intentional about distribution and growth loops from the beginning. A founder or adviser may supply the early expertise. He cautions against optimizing onboarding or email so early that the team neglects the product’s underlying value.

A head of growth is fundamentally a product leader focused on value connection. The first hire may need to analyze data, form a strategy, write specifications, work with engineers and build a team. Sahgal favors someone experienced enough to wear those hats, while recognizing that the company’s budget and stage constrain the choice.

Stebbings challenges whether a very senior executive will still want to do hands-on work. Sahgal agrees that a founder might find a person who helped build a successful team and is ready for the next step, rather than hiring its most famous leader. The key is relevant experience combined with willingness to operate.

Looking back on building a team, Sahgal says he did not hire quickly enough. By filling gaps himself with specifications and analysis, he left too little time for recruiting and made the problem worse. He estimates that in the early years he should have devoted 30% to 40% of his time to hiring; that is a retrospective personal judgment.

He also says he waited too long to move people whose working style did not match the team’s needs. For a senior hire he suggests looking for evidence over roughly three to six months, while first asking whether the person had the team and resources needed to succeed. A poor outcome can reflect either the hire or the setup.

His third mistake was failing to build relationships across the existing organization. He arrived with strong ideas about culture and speed, but did not bring colleagues along, leading to an “us versus them” dynamic. He would now establish relationships early and explain what he wants to do, why it matters and how others can contribute.

For an early growth hire, execution is essential and difficult to measure in an interview. Sahgal says he initially hired through his network or trusted references to learn whether candidates could actually ship work. Stebbings notes the diversity limitations of relying on networks; their discussion presents this as a trade-off rather than a complete hiring method.

One interview at Reddit centered on a single question: how would you think about driving growth for Reddit? Sahgal walked through a process—define growth, examine the funnel and its breaks, prioritize opportunities, set goals, generate ideas and review progress. He says the interviewer wanted clarity about the approach, not an invented feature presented as the answer.

Prioritization requires both analysis and judgment in his view. A formula can help compare opportunities, but Sahgal does not treat its ranking as a substitute for intuition. The useful blend depends on the evidence, the product and the consequences of being wrong.

A senior growth leader should also be able to attract talent, collaborate with low ego and act with integrity. Sahgal suggests speaking with people who reported to a candidate as well as former managers, asking whether they would work with the person again. References can be incomplete or embellished, so they still require judgment.

For a case study, he likes an ambiguous question that reveals how a candidate breaks a problem down. One example is choosing three metrics for Airbnb hosts or guests, setting goals and explaining trade-offs. He values a coherent process for solving the problem more than a confident but unsupported answer.

Compensation should reflect the importance of the role and give the person meaningful ownership in the outcome, Sahgal says. At Reddit, daily active users was the company’s north-star metric in the period he describes and a leading indicator for revenue. He does not prescribe a universal equity percentage or promise a financial outcome.

A north-star metric may remain stable while leading indicators change. Sahgal says daily active users alone is a downstream outcome; teams also need to examine retention and the behaviors that drive it. Those upstream indicators can change as the team tests ideas and learns.

For onboarding, he wants a self-starting growth leader but says the company must supply access to data, equipment and engineering help. Clean, usable data matters because the leader should be able to analyze and form a strategy rather than first repair the entire data function. Scrappiness does not excuse failing to equip the hire.

In the first month, he looks for someone who examines data, forms a concise strategy and starts working with engineers. He sees endless discussion without action as a warning sign. He suggests early execution, but later clarifies that listening and relationship-building remain important and should occur alongside the work.

A few dedicated engineers can give a new leader a way to execute, he says, ideally with shared ownership of the outcome. Stebbings warns that a leader who acts too aggressively could alienate product and customer teams. Sahgal agrees: explain the plan, seek input and take colleagues along, even when that means moving a little slower.

His most painful lesson came from Words With Friends. The team built a faster-play feature after inferring that users wanted shorter games, spending hundreds of engineering hours. Sahgal recalls that only about 5% of daily active users used it and that it added little value; he initially defended the decision.

Zynga founder Mark Pincus asked how the team had tested demand before building. It had not. Sahgal says a simple prompt asking users whether they wanted a faster board could have supplied an early signal at much lower cost. His later claim about the exact response such a prompt would have produced is hypothetical, not a measured experiment.

When Words With Friends was struggling, Sahgal says morale depended on a culture of shared ownership. Everyone could help find the next idea rather than waiting for the general manager to rescue the product. He recalls that an engineer and a designer contributed the idea that helped the team move forward.

After the fast-play lesson, the team built a repertoire of validation methods: prompts, notifications, user research and small prototypes. Some ideas cannot be validated cheaply and require conviction, he acknowledges. The objective is to select a quick, appropriate test where possible and protect engineering time for work that can create value.

Product reviews serve accountability, shared learning and decisions about what to do next. Sahgal wants leads from engineering, design, research, quality assurance and product at the table because each discipline shares responsibility for the outcome. The discussion should cover what happened, why, and what the team learned.

Review cadence should fit the product. Sahgal recalls weekly reviews at Zynga’s fast-moving game business and roughly monthly reviews at Reddit’s more stable platform. Meetings held too often can become meaningless, he says. At Reddit the team used a short one-page record rather than lengthy slide preparation, with documents kept for others to revisit.

Angel investing has made him place even more emphasis on talent. He says he often backs people he believes can find their way rather than only the initial product. This is his personal investment philosophy; he has financial interests in his angel investments and talent alone cannot guarantee their outcome.

He recalls a disappointing hotel investment that he made partly because owning a small part of a hotel sounded appealing. By his own account, he abandoned the criteria he had just described about the people and value involved. He offers it as a lesson in following one’s process, not a complete analysis of that investment.

Looking at social media in 2022, Sahgal expects broad networks to organize more around interests and communities, while friend-to-friend interaction moves toward private messaging. As a Reddit product leader he sees the company’s longstanding community structure as well positioned. That is his forecast and reflects his professional interest.

Asked how Reddit could go wrong, he worries that it might imitate other platforms instead of strengthening its distinctive depth of discussion. He argues that breadth of topics and depth can coexist, but sees depth as the differentiator. These are the company executive’s views, not an independent comparison of platforms.

He says onboarding, useful notifications and email remained important growth tools, while predicting that manipulative black-hat SEO practices would decline. He distinguishes those practices from technical site structure and making good content discoverable. These are his views at the time, not a current assessment of every search tactic.

The biggest hiring mistake, he says, is selecting a smart person by title alone without matching their experience to the company’s stage, industry and problem. He does not assume his consumer growth background would automatically translate to a B2B growth role. Some organizations need strategy; others have a strategy and need execution.

He admires Slack’s bottom-up software adoption, in which individual teams used a product before companies bought it more broadly. He describes the company’s pivot from a game and the eventual network effects as an example of consumer and B2B growth ideas meeting. His account is an interpretation of that growth story.

For a memorable first-use experience, Sahgal points to TikTok’s immediate video and swipe-based personalization as he experienced it. He notes that the product may have changed since then. He views it as an example of connecting users to value quickly, a subjective observation rather than a verified account of its internal systems.

Stebbings thanks Sahgal for the conversation, and Sahgal thanks him in return. The interview closes after a continuous discussion of users, growth hiring, product reviews and lessons from past work.