VPEC-T and Outcomes

3 views
Skip to first unread message

Roy Grubb

unread,
Sep 24, 2009, 3:49:00 AM9/24/09
to VPEC-T
Adrian Apthorp has mentioned Business Capability Maps to me.
http://www.forrester.com/Research/Document/Excerpt/0,7211,43049,00.html
They are closely related to some work I did with him for DHL a few
years ago.

From the description, it made me wonder if VPEC-T should include
outcomes and become VPEC-TO. OK, Nigel, desired Outcomes are another
way of saying Values, but Values are from the wide variety of
perspectives of all the different players involved. That was, I
believe, one of the major insights of VPEC-T. There, the business
seems to just be jostling with all the other players.

But the outcomes in Business Capability Maps appear to be Values to
the business specifically. Might it be worth having business outcomes
as an explicitly recognized sub-set of VPEC-T's Values?

O would clearly be just a subset of V (all the intended outcomes), so
perhaps should not be promoted to appear in the name in caps. VoPEC-
T? Nah! Even more geeky.

There may be advantages in demonstrating to the business that business
outcomes have a special place in the pecking order.

I haven't yet found a description of Business Capability Maps (as I'm
not about to stump up $499 to Forrester) to see how closely they are
related to Business Domain Models, so perhaps I'm running down a wild
path into the for(r)est.

Roy

nigelp...@googlemail.com

unread,
Sep 25, 2009, 2:46:54 AM9/25/09
to VPEC-T
Hi Roy,
See my post on Richard's thread
n

On Sep 24, 8:49 am, Roy Grubb <roy.gr...@gmail.com> wrote:
> Adrian Apthorp has mentioned Business Capability Maps to me.http://www.forrester.com/Research/Document/Excerpt/0,7211,43049,00.html

aapthorp

unread,
Sep 25, 2009, 2:09:00 PM9/25/09
to VPEC-T
This and the discussion on what's missing reminded me of this comment
I posted on the LiT site http://www.lithandbook.com/?p=33#comment-68

When I posted this I was thinking about how domain models (which tend
to focus on two dimensions; object and function) would extend based on
the VPEC-T dimenisons.

P.S. aren't outcomes events?

On Sep 25, 8:46 am, "nigelpsgr...@googlemail.com"
Reply all
Reply to author
Forward
0 new messages