Re: Anyone have stats on attrition rates after agile adoption?
- Thanks everyone, I appreciate your responses! I'll try to answer some
of your questions, below.
--- In email@example.com, "dnicolet99" <dnicolet@...>
>I'm trying to write an article on how using Scrum mitigates personnel
> I have to join Mike in saying that I don't have a direct answer to
> your question, but I do find the question itself intriguing. More to
> the point, I'm curious about the reason why you're asking.
loss (one of the five core risks defined by Lister and DeMarco in
their book Waltzing with Bears).
> But why look for a study to confirm or deny a cause-and-effectI believe that it's a direct cause-and-effect relationship, but I'm
> relationship between a methodology and some secondary factor the
> methodology was not designed to address?
no scientist. I think it would be cool to read about and learn from.
Indeed, maybe the real reason for the reduction in turnover that I've
observed is due to the relationship with the ScrumMaster and not
Scrum at all! (or did Scrum force better leadership?) There's a
study by Florida State University that shows how bad bosses affect
turnover: "They say that employees don't leave their job or company;
they leave their boss." Read more here:
How likely is it that anyone
> would undertake such a study?I don't know. But I'm pleased to see that the Agile Alliance has an
Academic Research Program that is funding studies - it will be great
to see what things turn up as a result.
> Your "quest for a study" is interesting because it seems as if manyNo, I don't do drugs, and I just had my annual physical, so I'm sure
> people believe in statistical information resulting from controlled
> studies, which may have been filtered and massaged in any number of
> ways for any number of reasons, while they disbelieve what they
> derisively call "anecdotal evidence and hearsay." Yet, anecdotal
> evidence and hearsay are nothing less than direct reports of reality
> coming straight from the horse's mouth. Why do you distrust your own
> observations? Do you think you might have been hallucinating?
I wasn't hallucinating! I certainly trust my observations, and it's
always nice to have confirmation by others, particularly when you're
in Scrum up to your eyeballs every day. :)
And I agree that statistics can be manipulated and can be misleading.
I'm really just interested in reading the study - how the problem was
approached, how they measured, their findings. It's a geeky interest.
It also helps those who aren't currently using agile approaches to
remove some skepticism. If I as a "known agilist" write that 'this is
true because I've observed it' it is likely to have less credibility
with the non-agile skeptical reader than 'this is true because I've
observed it and here are studies that back up my opinion.' I want to
be objective, and I feel outside corroboration (case studies, other's
observations, statistical studies, etc.) is helpful in achieving this
And for those who were trying to remember Ken's quoted numbers -
I did find a quote on a blog here: http://www.gunjandoshi.com/?p=88
that said Ken's quote on turnover was:
"Ken mentioned another interesting metric. He mentioned that there is
usually 20% turnover in development and 40% turnover in management
due to adoption of agile processes."
- They aren't definitive numbers, but Pete Deemer and Gabrielle
Benefield's great summary of Scrum at Yahoo! Concludes with a couple
of stats that indicate happier and therefore lower attrition amongst
"85% of team-members stated that they would continue using Scrum if
the decision were solely up to them."
"Team Morale: 52% of respondents reported Scrum is better or much
better; 9% reported Scrum is worse or much worse; 39% reported Scrum
is about the same."
I hope that helps a little.
--- In firstname.lastname@example.org, "michele_sliger"
> Hi all,
> I've looked high and low and can not find any statistical studies done
> on attrition rates of established agile teams. I've found plenty of
> anecdotal evidence, but nothing from a study. Has anyone seen any
> definitive numbers anywhere?