Measuring Developer Productivity with the DX Core 4

The authors of SPACE and DevEx answering the question they kept getting asked: which one should we use? Their answer folds all three lineages, including DORA, into four counterbalanced dimensions — speed, effectiveness, quality, and impact — with one key metric each and a short list of secondary metrics beneath it.

Read it with the authorship in view. DX publishes this article and DX sells the platform that implements it, so the framework and the product are the same artifact described twice. That does not make it wrong; the research pedigree is real and the counterbalancing argument is sound. It does mean the reported results — 3-12% efficiency gains across 300-plus organizations — come from self-selected customers with no control group and no published methodology. Directional, not evidence.

The most useful passage is the one that undercuts the sales pitch: DX states plainly that diffs per engineer works only when you counterbalance it, attach no targets or rewards to it, and communicate it carefully. Those are conditions on your organization, not features of the platform.

Resources

• Abi Noda, Laura Tacho, Margaret-Anne Storey, Michaela Greiler, and Nicole Forsgren, “Measuring Developer Productivity with the DX Core 4” (DX)
• DX Core 4 — the collection’s analysis, including where the framework misleads
• Improving Developer Experience — the engagement we run on this framework