Examining the ongoing challenges of delivering high-quality, value-added ERP services in Higher Education.
Thursday, November 1, 2012
“There are eight million stories
in the Naked City.” The Naked City, 1948.
How many stories can we write for
data analysis and reporting at this institution? That is the key question, soon
to be answered, as we formally launch our public efforts to transform business
intelligence at Harvard.
The project began a year ago when
a litany of complaints and concerns was articulated by campus financial leaders
and two dozen eyes turned toward me in an unsettling moment of “and what are you
doing about it?” The ocular mandate to take action on this long-festering
problem – a dramatic disconnect between business operations and information
technology – has driven my personal work agenda for the last year. And throughout
that time I have spoken melodramatically about that day and how I instinctively
curl into a fetal ball when I step foot on the Divinity School campus.
Today I return to the same room
with the same group (more or less) with a plan. (I wish it didn’t take so long,
but so it goes when you are also upgrading all your ERP systems and
implementing a few new modules and securing funding and negotiating various and
sundry other details). We now have a team in place, a platform that is
available and partly configured, and funding for the next few months. We have
defined a vision for improving tools and technology and a rubric for making
sense of root causes beyond IT. We are embracing a new (for us) methodology – Scrum
– and investing in resources (full-time Scrum “advisor”), training (CSPO, CSM,
Agile Bootcamp), and tools (GreenHopper) to support that transition. It is an
The next step is to build a
backlog. Despite the long list of concerns with the current state of business
intelligence / reporting / metrics, we do not have an inventory of gaps and
pain points. We need that list, preferably prioritized, in order to do
anything. So for the next three months we will be interviewing anyone who will
talk to us and writing user stories. Lots and lots and lots of user stories. How
many? How about 5,000 – that is my working target. That seems like a realistic
number based on what I know – wearing my hat as a consumer of current-state
reporting, I know that I can come up with a hundred, and my needs are fairly
simple. 5,000 stories would be a good problem to have – talk about an
unambiguous metric regarding the scope of unmet business needs! But it could
also prove overwhelming. How can we possibly manage a backlog of that scale? As
good as GreenHopper is for managing sprints, I think it might choke on that chicken
bone. We will have to think creatively, that’s for sure! And my full-time job might be backlog grooming!