SA Delivers Portfolio Case Study for the World's Largest Computing Project
Signals, Winter 2001
Strategic Analytics recently completed a comprehensive portfolio study
for the SETI@home project, the world's largest computing endeavor. Because
SETI@home is a voluntary undertaking and is composed of millions of active
accounts, the project managers face many of the same issues faced by managers
of consumer financial product portfolios.
Like usage of a credit card, a subscriber's access by SETI@home
is voluntary. Attrition and unit return rates for SETI@home are
analogous to voluntary attrition and credit utilization rates
for a credit card. In both cases, user behavior matures with the
age of the account, but is affected by external impacts. And like
retail bank customers impacted by the economic environment, competition
and policy changes, SETI@home subscribers are dramatically impacted by
server failures, CPU upgrades, and software upgrades. Our challenge: quantify,
through the use of advanced portfolio analytics, forward-looking measures
of 'customer' value, and identify strategies for optimal portfolio growth and high-value
customer retention.
At the heart our approach to the problem is the use of SA's unique
and proprietary Dual-time-Dynamics (DtD) technology. DtD, unlike typical
scoring and forecasting methods, analyzes intrinsic behavior separately
from behavior influenced by environmental influences. For financial services
companies, this capability allows portfolio managers the capability to
develop and test scenarios that faithfully reflect the response of a portfolio
to various environmental factors. Designed to work at the strategic level,
our technology enables portfolio managers to better control, understand
implement their decisions.
For SETI, the approach entailed analysis of project growth projections
recruitment and attrition rates, and the critical variables affecting
the future of the project, such as usage impacts from hardware and software
upgrades, new policies and software versions, and electric power concerns. Using
26 months of historical data, and an emphasis on forecasted behavior,
we produced accurate models of customer "net present value" and
in the process uncovering which groups will drive long-run value—apart
from their near-term behavior.
The results revealed several crucial management strategies for growth
and management of the portfolio of SETI@home users. For instance,
the quantified net present value (based on long-run behavior) of users
in every major country enables clear recruitment strategies. With
a projection-based NPV by segment calculated, management can now use acquisition
and other anticipated costs to make decisions about originations and optimal
portfolio construction in varying environments.
The analysis also showed which computing platforms could be counted on
for portfolio growth, and which users should be targeted for retention
work. We also quantified cross-sell opportunities amongst the different
user platforms—identifying the precise stages (by customer segment)
when the segment is optimally receptive to changing to higher value versions
of the software. This analysis is an important input for building
account-level models to optimize cross-sell and illustrates our focus
on integrating strategic decision-making with existing tactical tools.
The deployment of SA's advanced analytics at SETI@home illustrates
their ability to enable management to make sound assessments
of long-term value, and weigh costs and benefits of alternative actions
rapidly—within a comprehensive and consistent analytical framework. As applicable
to financial services and other consumer-based portfolio issues
as they are to SETI, these capabilities translate to better, faster portfolio
decisions and significantly enhanced bottom-line results.
|
Top of page