Timeline for Task estimation for PMs who are not developers
Current License: CC BY-SA 4.0
11 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Mar 5, 2019 at 11:36 | vote | accept | Anna Galstyan | ||
| Mar 5, 2019 at 11:33 | vote | accept | Anna Galstyan | ||
| Mar 5, 2019 at 11:33 | |||||
| Mar 5, 2019 at 11:33 | vote | accept | Anna Galstyan | ||
| Mar 5, 2019 at 11:33 | |||||
| Feb 18, 2019 at 10:31 | comment | added | kpollock | To be clear, x3 is my personal experience for teams of 2-6, in which I have worked over 25 years. This has been at many different companies (I was a contract and agency developer) I arrived at this by comparing actual to (best honest developer) estimates. Obvously, YMMV. But if you have no historical data to go on, I suggest it as a good starting point. | |
| Feb 9, 2019 at 20:50 | comment | added | Peteris | @kpollock simply assuming 3 seems to be based on superstition. In our team, we're using a more science-focused approach and multiplying by π instead, which gives slightly more accurate estimates. | |
| Feb 8, 2019 at 16:30 | comment | added | user1675016 | @kpollock yes that pattern that holds out on many projects, not just software: take your worst case estimate and multiply it by 3. Some developers learn to do that, some don't. I would recommend the article joelonsoftware.com/2007/10/26/evidence-based-scheduling | |
| Feb 8, 2019 at 9:11 | comment | added | kpollock | After 25 years as a developer, looking retrospectively at estimates vs actuals - I find the "Magic Multiplier" is consistently 3. I work usually in teams of 2-6. | |
| Feb 8, 2019 at 7:52 | comment | added | par | As a long-term developer who has worked with PMs for years, I suggest for a team of 4 developers or fewer you should multiply developer estimates by two for your magic number and you're likely to find your project estimates are reasonable. For each developer beyond 4 in a given team add 0.3 to 0.5 to that multiplier. I'm 100% serious as complexity increases as you assign personnel. | |
| Feb 8, 2019 at 0:18 | comment | added | Quaternion | From personal experience I've found the estimates provided by developers to be sound, if and only if you multiply against a magic number. I would recommend setting up a table of initial estimates vs actual time to delivery. As your sprints go on you'll be able to better estimate time. | |
| Feb 7, 2019 at 14:12 | comment | added | MCW | Bounty!! Well said. | |
| Feb 7, 2019 at 14:11 | history | answered | Sarov | CC BY-SA 4.0 |