Stop rolling the dice: an approach to effort estimation in software development
summary
Claims a pragmatic process that turns estimates into documented decisions, not guesses, and shares an Excel estimation tool that embeds an Amdahl-like model so teams size and parallelism drive schedule and cost. Uses a 0.6 parallelization coefficient example, shows four FTEs cut elapsed time roughly in half but raise total man‑days from 50 to 112 (paying ~\$200k more in the example), warns that tripling speed may need ~50 people, and recommends aiming for ~5% variance through requirements mapping, skill matrices and shared estimation.
AI-written summary · about
chapters (7)
- 0:00Effort estimation challenges
- 3:14Why traditional methods fail
- 8:14Estimation as decision-making
- 14:19Common estimation mistakes
- 21:48Cost of velocity and team size
- 25:17Introducing estimation tool
- 33:05Q&A
related talks
01h 45mEstimates or NoEstimates
03h 00mBeyond Estimates (Estimates and NoEstimates ) - Let's Explore the Possibilities
47mCreating efficient, accurate, software estimations
30mSoftware Estimation Made Easy
47m[Workshop] NoEstimates: Forecasting with Less Effort and More Accuracy
48mRespect estimates
45mBeyond Estimates
42mBeyond Estimates (Estimates and "NoEstimates") - Let's Explore the Possibilities