Feature Toggles: Practitioner Practices and a Case Study
https://users.encs.concordia.ca/~pcr/paper/Rahman2016MSR.pdf
Method
"We mined the Google Chrome development history from 2010 to 2015, which covers 39 releases from release 5 to 43."
Population
"Across all 39 releases, a total of 2,409 distinct toggles have been used, 70% of which were removed."
What it does not show
Does not measure incidents or defects attributable to toggle combinations, so it prices the maintenance burden but not the combinatorial testing risk. The supplementary 13-company material is drawn from public talks and blogs — self-selected and success-biased. Chrome’s toggle discipline may not generalise to smaller organisations.
Md Tajmilur Rahman, Louis-Philippe Querel, Peter C. Rigby, Bram Adams
Flags accumulate as a maintenance liability even at a well-resourced organisation with a dedicated removal campaign: toggle count grew from 263 to 1,040 across the releases studied, developers made over 5,044 commits introducing or refactoring toggles, and of tracked release toggles only 20% were actually removed while 17% lingered in the source as debt.
Tier III: Longitudinal single-project code-mining case study with counted metrics over 39 releases, supplemented by qualitative analysis of public material from 13 companies; no control group.