After we had aligned on a list of Big Rocks and team two-pagers, the next step in our process was to allocate resources. For us, this represented a classic but challenging “matrix planning” problem that we gradually became quite good at.
The key challenge was that we allocated resources to our teams (i.e. Investment Areas) but Big Rocks tended to cross Investment Areas. For examples, to launch YouTube Music, we needed bits of help from many teams — from the search and discovery team, to new infrastructure for audio-only streaming, to new artist-centered work from the creators team.
We ran this process in a complex spreadsheet, but the below technique roughly replicates the same process. There are 3 primary components of the matrix model:
: This was our tracking for big bets (generated from the exercise). Often these Big Rocks would align closely to an Investment Area team, but sometimes they crossed across multiple teams. Our process was gradually designed for the latter since it was the more complex case. We often called these “hoodie squads” to indicate that these team members should think of their primary team as the Big Rock team instead of their home IA, and it was signified by answering “which hoodie would this person wear if they were asked to wear their team’s hoodie to work.” : The YouTube team was organized into investment area teams (grouped into focus areas). As our primary organization structure and primary unit of planning and allocation, every employee belonged to an Investment Area. These teams presented their requests for headcount in their (with low, medium, high guidance) and also managed the allocation of their team members between “IA Bets” and “Big Rocks.” : A table with a row for every employee, and also a row for every ANS (”allocated not started” - i.e. people who had been hired but hadn’t started yet) and TBH (”to be hired” open roles). At YouTube, this was done as an export from WorkDay and then added to by the recruiting team for the ANS and TBH slots.
The alignment process between these three areas was done cyclically; we would iterate through each until the matrix was “funded.” For example:
Big Rock managers would request what they believed they needed in terms of both aggregate headcount as well as specific teams / skills that they needed. IAs would ask for headcount in their two-pagers with a guess at both what their Big Rock obligations were likely to entail, and also what their IA Bets (initiatives from the investment area itself) required. Team members were allocated to Big Rock teams after recruiting conversations with individuals for those assignments. We compared the requested budgets and allocations to see if we had achieved a balance of funding. If we hadn’t, we would iterate again — perhaps by increasing funding for an IA or by reducing / eliminating the priority for a Big Rock. You can replicate this process by clicking this button to and then this button to in the tables below.
Table 1 | Big Rocks
Add your list of Big Rocks here. As resources are assigned from the they will show up here. The key test was to try to get the # of People Assigned as close to Target Headcount as possible.
Table 2 | Investment Areas
Add your list of key teams. Our structure had two primary levels, Focus Area and Investment Areas, with the latter being our primary unit of planning. A few notes:
This structure presumes that Target Headcount is calculated by the sum of People on the team, including “to be hired” (TBH). This was helpful as it focused on “zero based budgeting” and also prevented baseline discrepancies. It was considered bad form / undesirable to assign TBHs to Big Rocks. This is the heart of the Big Rock philosophy — fund your big bets first. The total TBH count was the primary input for the recruiting team for each area. We watched the Big Rock % closely. In aggregate across the team, we generally landed at ~30%. The allocation below is currently at .
People assigned to Big Rocks
People Assigned to IA Bets
Table 3 | People Roster
This would be a full roster of every employee on the team (usually started with a Workday export) with rows added for every “to be hired” (TBH) called for in the budget allocation.
Every employee’s primary assignment is to an Investment Area. Some employees (though definitely not all) could be assigned to a Big Rock hoodie squad. It was generally considered bad practice for an individual to be assigned to multiple Big Rocks, but it did occasionally happen (e.g. for very specialized skill sets, or for our critical path teams). The table will below will highlight in red if an employee is on more than Big Rock, but it will allow it. It was generally not desirable to assign TBHs to Big Rocks (fund your Big Rocks first), but it was technically supported.
The table is grouped by Investment Area for convenience but feel free to ungroup it.
These are a few key useful views for understanding your allocations.
Resources assigned to each Big Rock.
How many resources are allocated to each IA?
What % of investment area is allocated to Big Rock?
People on multiple Big Rocks
Big Rock hoodie squad details