Does organizing a company into multiple business units inherently reduce velocity and visibility compared to a unified team structure?
11 recorded positions from 6 people, first said Oct 12, 2022. They do not agree — the readings below are what each one actually argued.
Multi business unit structure naturally reduces velocity and visibility without being a failure
Ryan Petersen · Nov 13, 2023
He has changed his mind: autonomous business units running fast in isolation don't work at Flexport, which needs an end-to-end view and more process
Their industry requires end-to-end delivery, and most true bureaucracy actually comes from a lack of process where people can't get things done or understand how things work; a healthy amount of process and cross-team planning is necessary
Scope: the isolated fast-moving business unit model works for a lot of other companies; only a little bit more process, not heavy bureaucracy
49:50 20VC: Flexport's Ryan Petersen: Reflections on Leadership from 13 Years Leading Flexport, Why Velocity not Speed is Most Important in Company Building, How Money Creates Inefficiencies in Scaling, The Future of Trade with China & Why Remote Work is so Cha
Alex Taussig · Feb 12, 2024
Vinted's velocity is actually good, and the slower pace and lower visibility are an expected consequence of shifting to a multi-business-unit model rather than a failure
When you run several separate businesses you should expect different velocity and visibility, and M&A has already helped accelerate and bootstrap those units
Scope: says it can always be better
67:08 20VC: The Ultimate Guide to Scaling Marketplaces, Why Rule of 40 and EBITDA Optimisation is BS, How Founders & VCs Should Approach Market Sizing and Outcome Scenario Planning and Why Europe is Failing with Vinted CEO, Thomas Plantenga & Alex Taussig
Gm structure is the root cause of big company downsides like hr overhead and decision friction
TJ Parker · Jun 5, 2023
The moment a company moves to a GM-focused org it stops being a startup, and most of the big-company downsides people bemoan — HR volume, decision-making overhead, over-processedness — are a byproduct of that single decision.
GM structures create pockets of every function under each single-threaded P&L owner, which generates the process and overhead characteristic of large companies.
Scope: TJ flags this as somewhat controversial
28:16 20VC: The Largest Venture Backed D2C Consumer Exit; PillPack: $0-$300M Revenues in 5 Years & The Biggest Lessons Scaling the B2B Business to $300M in 2.5 Years with TJ Parker, Co-Founder @ PillPack
TJ Parker · Jun 5, 2023
Founders should stay functional until it is so obvious the structure cannot possibly work that GM becomes excusable; going GM will definitely make the business worse over the long term.
GM structures force leveling parity, compensation consistency across many businesses and transfer equivalence work, creating enormous HR and coordination overhead that comes free in a single functional org.
Scope: TJ acknowledges he is far on this extreme; with a manageable number of product lines he wouldn't even have the conversation; damage may not show up immediately
29:51 20VC: The Largest Venture Backed D2C Consumer Exit; PillPack: $0-$300M Revenues in 5 Years & The Biggest Lessons Scaling the B2B Business to $300M in 2.5 Years with TJ Parker, Co-Founder @ PillPack
Also on the record
Ryan Petersen · Nov 13, 2023
Silos are dangerous in logistics and no org chart can solve the coordination problem, because the only person with the whole network reporting to them becomes a bottleneck
An end-to-end shipment spans factories, trucks, ports, carriers, customs and warehouses across 100+ countries; a delay at origin forces replanning downstream, so teams must attend to what is upstream and downstream of them
23:30 End to end network complexity cannot be solved by org chart since full visibility creates a bottleneck
Gustav Söderström · Oct 12, 2022
Most executives split their org into as many independent units as possible for parallelization, but the right approach is the opposite — deliberately synchronizing the organization
At the end of the day one user uses one application and doesn't care who built which part; the experience must feel as if it was built by a single developer for a single user, which requires synchronization rather than swim lanes
36:49 Org should be deliberately synchronized rather than split into independent units since users experience one product
Harry Stebbings · Jun 5, 2023
GM-led orgs create dysfunctional decision making because the GM running a segment is not the ultimate decision maker and can be overridden by the CEO or CPO above them, producing a chasm around where power actually sits.
You end up with CEOs on top of CEOs.
30:50 Gm structure creates decision making dysfunction since gms can be overridden by executives above them
TJ Parker · Jun 5, 2023
Functional orgs produce better products because your single best product, engineering and design leaders work on every product rather than being split across business units.
GM structures mean you don't have your best people on every product.
31:10 Functional orgs produce better products since best leaders arent split across business units
Glen Coates · Mar 8, 2023
Shopify's checkout problems were caused largely by its org structure: 12-13 fully-staffed product-line divisions each owning a slice of the product made it very hard to coordinate fixes that spanned divisions
Checkout bridges everything in Shopify, so its fundamental problems required multiple divisions to collaborate, and an org cut along those lines makes cross-cutting solutions hard — org chart shipping / Conway's Law
7:37 Product line divisions fragment ownership blocking cross cutting fixes
Glen Coates · Mar 8, 2023
Merging misaligned teams into one large team is not itself a fix for misalignment — you still have to split it back into a set of workably sized teams
You need an org chart: a way of organizing people into groups of a size that can actually work together
23:22 Merging misaligned teams does not fix misalignment still need workable org chart
Glen Coates · Mar 8, 2023
The real defect in the old structure was that the first point where all teams joined was the CEO, making him the first stop in every escalation — an unreasonable ask of a public company CEO and a cause of six-month stalemates
With no shared leadership layer beneath the CEO, disagreeing teams had nowhere to escalate and simply sat frustrated; a cross-functional leadership team above the teams can absorb escalations and make clear calls
24:18 Missing cross functional leadership layer forces ceo into every escalation
Your assistant can query this graph directly — 11 positions here, 19,646 across the corpus. Add 996.fm over MCP.