The Org Chart Is the Product: Why Your Structure Ships Whether You Like It or Not

Kayvon Beykpour -- former Head of Product and GM of Consumer at Twitter, co-founder of Periscope -- was the longest-tenured head of product at Twitter, credited with transforming the product org from stagnant to shipping regularly (Spaces, Super Follows, Communities, Twitter Blue).

Beykpour lived the failure mode of functional organizations for three years. Twitter had a textbook functional structure: separate heads for consumer product, revenue product, engineering, and design. The structure concentrated all decision-making authority at the top. When CEO Jack Dorsey wasn't actively resolving conflicts between these highly opinionated leaders, nobody was.

"If you're going to have a functional organization, you need to have a GM or a CEO who's extremely leaned-in to tiebreak and resolve conflict and make sure the team is moving quickly. And Jack, for all of his amazing qualities, just wasn't operating that way."

Kayvon Beykpour

▶ 00:12:52

The result was consensus-seeking and deadlock. Beykpour describes his first year as head of product as "walking through mud" -- not because the work was hard, but because it was "politically and bureaucratically exhausting." Having a compelling strategy meant nothing when execution required driving consensus across peer functional leaders.

"It is one thing for me to have some ideas and a plan and a strategy that I felt compelling. But when you have to essentially drive consensus amongst your peers across the other functions, that's a different game. That's not execution. That's politics and consensus building."

Kayvon Beykpour

▶ 00:22:40

The functional org's promise of "specialized excellence" became specialized fiefdoms. When Parag Agrawal became CEO, one of his first moves was shifting toward a GM structure. The proof the bottleneck was structural: the same engineers and designers who couldn't ship at Twitter built extraordinary products elsewhere.

Dhanji R. Prasanna -- CTO of Block, overseeing 3,500+ people -- led Block's transformation into one of the most AI-native large companies, previously built Google Wave and worked on Google+. Drew Houston -- CEO and Co-founder of Dropbox -- navigated the company through multiple product expansions and org transitions from Y Combinator to IPO.

Block's CTO made one of the most counterintuitive org moves in recent tech: dismantling the GM structure and replacing it with a functional one. Square and Cash App had been running as effectively independent companies -- separate engineering, design, marketing, even separate offices. This worked during a mature execution phase. But when Block needed cross-cutting AI investment, the GM walls prevented it.

"We went from a GM structure to a functional org structure, which was I think the key to making our transformation into being more of an AI-native company."

Dhanji R. Prasanna

▶ 00:08:55

Conway's Law was the deciding factor. Under the GM model, shared AI capabilities would have been built twice -- or not at all -- because each business unit had no incentive to invest in shared infrastructure. The transition was "quite painful" and not universally supported. But the alternative was worse.

"Conway's Law can be really, really powerful. It's the law that basically says you ship your org structure. So what you're organized as in terms of teams, in terms of collaborating groups and your operating model matters a lot to what you build."

Dhanji R. Prasanna

▶ 00:10:32

Drew Houston at Dropbox experienced the opposite trajectory. The functional org worked beautifully when Dropbox was one product with one customer type. But expansion into Paper, Mail, and later Dash destroyed accountability because nobody owned the full outcome for any product.

"You try to build multiple products in a functional org, but then you lose accountability. People are like, 'I'm spending a lot. I'm trying to make Dropbox work. I'm trying to make mail work.'"

Drew Houston

▶ 01:15:50

Houston describes the broader pattern: "Everything is just negotiated compromise with everyone on the roadmap." The functional org amplifies the accountability diffusion that plagues growing companies. Together, Block and Dropbox prove the central thesis: org structure is a strategic tool that should change when the strategy changes. Block moved from GM to functional for shared AI infrastructure. Dropbox moved toward distributed ownership for multi-product accountability. Neither was wrong. Both were painful.

Lenny's newsletter underscores this with a tactical framework from Rohini Pandhi and Mary Liu ("General management, functional, and hybrid models"), who identify two decision factors: how close the product's North Star metric is to revenue (closer means more suited to GM), and how critical the holistic customer journey is to growth (more critical means more suited to functional). Most companies, they argue, end up in a hybrid model that skews one direction.

Gustav Soderstrom -- Co-President and CPO of Spotify -- led Spotify's product evolution from the original "squad model" era through the current more centralized structure. Nickey Skarstad -- Director of Product at Duolingo, ex-Airbnb, Etsy, Shopify -- has worked inside both functional and GM models at scale. Jeetu Patel -- Chief Product Officer and President at Cisco, managing 30,000 people -- transformed Cisco from a traditional enterprise into an AI-first platform company.

Soderstrom places companies on a spectrum between two trillion-dollar extremes. On one end: Amazon with two-pizza teams, minimal dependencies, direct user access -- speed and novelty, but you sometimes ship multiple search boxes on the same page. On the other end: Apple, the world's largest functional org, where you never see two search boxes on an iPhone -- but people report waiting seven years to ship.

"They are instead in what is probably the world's biggest largest functional org. They certainly do everything we do. They have music service, podcast service, audiobooks. And yet they build something that feels more like it was built by a single developer for a single user."

Gustav Soderstrom

▶ 00:39:20

Spotify chose centralized because its strategy -- a single application adding multiple content types into one user experience -- demands coherence. Soderstrom learned the cost of the alternative: when Spotify gave maximum autonomy to squads, it produced "a lot of heat" with a hundred teams running a hundred strategies.

Nickey Skarstad offers the practitioner's default: start functional, graduate to GM when distinct business opportunities emerge. At Etsy, functional worked because everything was one marketplace. At Airbnb, Experiences needed a GM structure with its own operations, marketing, and product.

"Had they launched experiences inside of the Etsy style of organizational structure, it never would've succeeded because it had such a unique business need, and it needed its own process and ways of building product."

Nickey Skarstad

▶ 00:23:44

Jeetu Patel at Cisco represents the extreme failure mode of GM culture at scale. The company had made 251 acquisitions, and every leader's career aspiration was to become a general manager -- owning their own silo. The result: a $45 billion holding company fragmented into hundreds of $40 million fiefdoms with incoherent customer experience.

"Everyone wanted to be a GM at Cisco. They wanted to own their own fiefdom, be a general manager. Because they felt like, 'In order for me to move up the ranks, I need to be a general manager, which means I need to have my own sales team. I need to have my own marketing team. I need to have my own product team.'"

Jeetu Patel

▶ 00:12:42

Patel's solution: "We have to become not a holding company of 251 acquisitions and thousands of different products, we have to become a platform company." Loosely coupled, tightly integrated. Matt MacInnis at Rippling echoes this: when Rippling's product function became "locally optimized but globally incoherent," the product followed suit -- because you ship your org chart.