TITLE: Paying for Value or Paying for Time
AUTHOR: Eugene Wallingford
DATE: June 05, 2009  3:25 PM
DESC: 
-----
BODY:
Brian Marick tweeted about his mini-blog post
<A HREF="http://marick.tumblr.com/post/118439959/pay-me-until-youre-done">
   Pay me until you're done</A>,
which got me to thinking.  The idea is something
like this:  Many agile consultants work in an
agile way, attacking the highest-value issue
they can in a given situation.  If the value of
the issues to work on decreases with time, there
will come a point at which the consultant's
weekly stipend exceeds the value of the work he
is doing.  Maybe the client should stop buying
services at that point.

My first thought was, "Yes, but."  (I am far too
prone to that!)

First, the "yes":  In the general case of consulting,
as opposed to contract work, the consultant's
run will end as his marginal effect on the company
approaches 0.  Marick is being honest about his
value.  At some point, the value of his marginal
contribution will fall below the price he is
charging that week.  Why not have the client end
the arrangement at that point, or at least have
the option to?  This is a nice twist on our usual
thinking.

Now for the "but".  As I tweeted back this feels
a bit like
<A HREF="http://www.mathacademy.com/pr/prime/articles/zeno_tort/">
   Zeno's Paradox</A>.
Marick the consultant covers not half the distance
from start to finish each week, but the most
valuable piece of ground remaining.  With each
week, he covers increasingly less valuable
distance.  So our consultant, cast in the role of
Achilles, concedes the race and says, okay, so
stop paying me.

This sounds noble, but remember:  Achilles would
win the race.  We unwind Zeno's Paradox when we
realize that the sum of an infinite series can
be a finite number -- and that number may be
just small enough for Achilles to catch the
tortoise.  This works only for infinite series
that behave in a particular way.

Crazy, I know, but this is how the qualification
of the "yes" arose in my mind.  Maybe, the
consultant helps to create a change in his client
that changes the nature of the series of tasks
he is working on.  New ideas might create new
or qualitatively different tasks to do.  The
change created may change the value of an
existing task, or reorder the priorities of
the remaining tasks.  If the nature of the
series changes, it may cause the value of the
series to change, too.  If so, then the client
may well want to keep the consultant around,
but doing something different than the original
set of issues would have called for.

Another thought:  Assume that the conditions
that Marick described do hold.  Should the
compensation model be revised?  He seems to be
assuming that the consultant charges the same
amount for each week of work, with the value
of the tasks performed early being greater than
that amount and the value of the tasks performed
later being less than that amount.  If that is
true,then early on the consultant is bringing
in substantially more value than he costs.  If
the client pulls the plug as soon as the value
proposition turns in its favor, then the
consultant ends up receiving less than the
original contract called for yet providing more
than average value for the time period.  If the
consultant thinks that is fair, great.  What if
not?  Perhaps the consultant should charge more
in the early weeks, when he is providing more
value, than in later week?  Or maybe the client
could pay a fee to "buy out" the rest of the
contract?  (I'm not a professional consultant,
so take that into account when evaluating my
ideas about consultant compensation...)

And another thought:  Does this apply to what
happens when a professor teaches a class?  In
a way, I think it does.  When I introduce a new
area to students, it may well be the case that
the biggest return on the time we spend (and
the biggest bang for the students' tuition
dollars) happens in the first weeks.  If the
course is successful, then most students will
become increasingly self-sufficient in the area
as the semester goes on.  This is more likely
the case for upper-division courses than for
freshmen.  What would it be like for a student
to decide to opt out of the course at the point
where she feels like she has stopped receiving
fair value for the time being spent?  Learning
isn't the same as a business transaction, but
this does have an appealing feel to it.

The university model for courses doesn't
support Marick's opt-out well.  The best students
in a course often reach a point where they are
self-sufficient or nearly so, and they are
"stuck".  The "but" in our teaching model is
that we teach an audience larger than one, and
the students can be at quite different levels
in background and understanding.  Only the best
students reach a point where opting out would
make sense; the rest need more (and a few need a
lot more -- more than one semester can offer!).

The good news is that the unevenness imposed by
our course model doesn't hurt most of those
best students.  They are usually the ones who
are able to make value out of their time in the
class and with the professor regardless of what
is happening in the classroom.  They not only
survive the latency, but thrive by veering off in
their own direction, asking good questions and
doing their own programming, reading, thinking
outside of class.  This way of thinking about
the learning "transaction" of a course may help
to explain another class of students.  We all
know students who are quite bright but end up
struggling through academic courses and programs.
Perhaps these students, despite their intelligence
and aptitude for the discipline, don't have the
skills or aptitude to make value out of the latency
between the point they stop receiving net value
and the end of the course.  This inability creates
a problem for them (among them, boredom and low
grades).  Some instructors are better able to
recognize this situation and address it through
one-on-one engagement.  Some would like to help
but are in a context that limits them.  It's
hard to find time for a lot of one-on-one
instruction when you teach three large sections
and are trying to do research and are expected
to meet all of the other expectations of a
university prof.

Sorry for the digression from Marick's thought
experiment, which is intriguing in its own
setting.  But I have learned a lot from applying
agile development ideas to my running.  I have
found places where the new twist helps me and
others where the analogy fails.  I'm can't shake
the urge to do the same on occasion with how we
teach.
-----