Kayvon Beykpour on Product Reviews, Feedback and Periscope’s Defaults

Kayvon Beykpour recalls meeting Scott Belsky through an investor who declined to back Periscope. At the time the founders had prototypes and a fundraising deck but no finished app; Belsky shared their enthusiasm and invested. The relationship is relevant to Beykpour's later praise of him.

The founders imagined live video as the closest available technology to teleportation: a way to see through another person's eyes. Belsky's support helped the team pursue that idea after many early fundraising rejections.

Beykpour says Belsky taught him to move between a product's mechanics and its larger story. Periscope was a live-streaming tool, but the founders also described an ambition to connect people through shared experiences; that framing could help recruiting and product conviction.

He says strong stories can combine familiar emotions with unfamiliar possibilities. A compelling account of what a product stands for can help both a team building it and customers choosing among many alternatives.

Periscope faced skepticism because live-streaming products had existed before. Beykpour's lesson is to distinguish criticism worth testing from opposition a founder should persist through, without becoming so attached to an idea that they ignore evidence of failure.

The team believed low latency was essential to its interactive vision. If a viewer asked a broadcaster to turn left, the response needed to arrive quickly; tens of seconds of delay would weaken the feeling of participating in the scene.

Beykpour recalls the substantial technical work needed to reduce the delay to under five seconds. He says the early experience of typing a comment and seeing a broadcaster respond almost immediately felt magical to users at the time.

Asked whether product is art or science, he resists a numerical split. Understanding customer needs, data and coordination matters, while storytelling, emotion and conviction about the problem are often underappreciated parts of building a product.

He defines a product manager's distinct role as clarifying the product's purpose, strategy, customer value and success measures. The PM should ensure the team knows what it is building and why, rather than allowing different groups to optimize for incompatible outcomes.

At greater scale the role also involves coordinating product, engineering, design, research, legal and other disciplines. A small startup may share these responsibilities informally; a larger organization needs explicit alignment across many participants.

A founder may need a product manager when the team is constantly reacting to late-discovered problems and has little time to think proactively. Another signal is that growing communication costs leave people unclear on priorities.

For an early-stage PM, Beykpour looks for someone deeply curious about customer needs, immersed in the product and willing to gather qualitative and quantitative evidence. The person should be comfortable with ambiguity, multiple roles and making progress without a complete support structure.

To assess comfort with ambiguity, he asks candidates about actual difficult experiences and how they navigated them. Such stories can reveal judgment, although he says interviews remain imperfect and the real test is working together.

His open-ended questions include work a candidate is proud of, a failure and what they learned, and a difficult interpersonal conflict. Asking about products they love or dislike helps reveal how they think, rather than requiring a predetermined answer.

He believes a common hiring mistake is taking too long to recognize that a person is not in the right role. He has made that mistake himself and says delay can be costly for the manager, the employee and the rest of the team.

Beykpour describes leadership-level product reviews held weekly, with up to three 45-minute sessions and a 15-minute leadership debrief after each. Other product leaders can have their own review cadence; this is the practice he describes at the time, not a universal template.

A review may focus on an early hypothesis and strategy, a product being built or prepared for launch, or a product already in market. In the last case, the team examines qualitative and quantitative learning and proposed next steps.

He sees two audiences for each review. The product team gets a fresh cross-functional perspective, challenges and help with a difficult decision; leadership learns how the product and team are progressing, including ambition, velocity and working dynamics.

Warning signs include repeated reviews with little progress and a team that merely tries to guess what leaders want to hear. Beykpour prefers a team with a clear point of view that can explain its reasoning and challenge his own assumptions.

Leaders contribute to the problem if passionate feedback sounds like an instruction. Beykpour says he sometimes falls into that trap and must distinguish a question or suggestion from a decision that the team is expected to follow.

He deliberately prefaces comments with phrases that make clear whether he is asking a question or giving a directive. He also repeats the principle in broader team communications, because new colleagues may not have heard earlier explanations.

His stated default is that product-review comments are feedback unless leadership explicitly says otherwise. He regards overruling the team as an exceptional event and wants teams to be free to consider, reject or defer input with good reason.

Prioritization belongs to the team and the product manager. They must sort customer research, data and leadership input, act on some of it, defer other items and explicitly ignore some suggestions; responding to everything is neither possible nor desirable.

Leaders must also make feedback usable. After a review, Beykpour says his leadership group spends 15 minutes distilling the discussion into a few written points, reducing confusion when several leaders have offered different or contradictory views.

On customer feedback, he distinguishes listening to the underlying problem from automatically accepting a customer's proposed solution. A team should understand the difficulty and then consider creative alternatives, including options the customer may not have named.

He believes intuition can help a team act when exhaustive validation would take too long or is impossible, provided it is balanced with data, research and subsequent validation. The correct balance is context-dependent and teams can get it wrong.

Beykpour identifies a major Periscope mistake: changing recordings from a short-lived default to indefinite retention. He says he personally pushed for the change because the team saw valuable live events disappear and wanted to preserve them.

Users could still delete a recording, but the default became permanent. In hindsight, he believes the team optimized for exceptional broadcasts such as breaking news while overlooking a more common use: people casually talking and spending time together.

Those everyday conversations did not necessarily need to remain online forever. He would instead have chosen a default suited to the primary behavior while offering an elegant way to preserve exceptional broadcasts; this is his retrospective product judgment.

In the quick-fire round, Beykpour names Belsky as a product leader he admires, citing both Belsky's support for Periscope and his view of Adobe's transitions. Belsky was also an early investor in Beykpour's company, a relationship relevant to the endorsement.

His advice to a new product leader is to work on something they care about deeply. He sees conviction and passion as important not only for product managers but also for the designers and engineers doing the work.

As an angel investor, he tries to remember how little context he has compared with a company's team. He aims to offer product opinions when invited and to provide the kind of support founders actually need, rather than treating an investment like a product review.

Asked about other product strategies, he praises Adobe's transformation and SpaceX's ambition, pace and public development. These are his 2022 assessments; he draws a parallel with sharing experimental work publicly and soliciting feedback at Twitter.

Stebbings thanks Beykpour and closes the conversation.