Engineering allocation measures how many full-time engineers work on different categories: bugs, infrastructure, features, roadmap. In one August example, a team split 9 FTEs across these buckets: 2.5 on bugs, 1.6 on infrastructure, 2.0 on unplanned features, 2.9 on roadmap. The same 2.5 FTE could mean five people half-time or twenty-five people at one-tenth time.
Chapters
What is allocation? Great question. Allocations account for how many people worked on which things. It's hard capturing a human's time and attention in a quantified way. For Jellyfish, the unit we use are FTEs, full-time equivalents, or full-time engineers. Think of it this way: how much time and attention went into one thing versus another. So based on Jellyfish insights, I know that 2.5 FTEs worked on bugs in August. Let's break down that 2.5 FDEs to see how it applies across a team. So 2.5 FDEs being spent on bugs really could be: 5 people spending half of August on bugs, or 25 people spending a tenth of August, or approximately 3 days each on bugs, or 2 people working full-time on nothing but bugs in August, with a third person spending half their time on bugs. But it's most likely less simple than that. In reality, it's probably a gradient. This spread gives a great depiction of what a typical gradient looks like. The superpower of jellyfish is the ability to quantify the gradient. What's relevant about jellyfish is we can account for all observable work we can see. However we classify investment categories, we can use the power of measuring how much human time and attention went in across all of them. 2.5 FTEs for bugs, 1.6 FTEs for infrastructure, 2.0 FTEs for unplanned features, and 2.9 FTEs for roadmap. For this example, we can account for all 9 human beings. One way or another, that will add up to 9 FTEs during the whole month of August. This investment allocation time series expresses the whole organization's focus. But what's really behind it is someone who spent most of their time on bugs, someone else who spent a little bit of time on bugs but mostly features, and so on. The jellyfish allocation model is doing the math to express that in a digestible yet quantifiable way for all end users.
