The Last Mile Problem - Data Fluency: Empowering Your Organization with Effective Data Communication - page 16

and retention metrics at local and regional levels. First, data on the 250,000
employees is collected from many sources (employee, employer, and a variety
of healthcare and wellness sources: hospitals, clinics, counseling centers, thera-
pists, antitobacco programs, online programs, and even gyms). Some parts
of the wellness program might be done in-house, whereas other campaigns
like quitting tobacco and fighting depression were handled externally. Data
collected on the 250,000 employees is then merged with that of 40 million
other lives covered by this large insurance company—a large pool of data
from which statistically significant comparisons could be made between
those participating in a wellness program and those not. Data points came in
daily—sometimes overnight via data sync between the employer's employee
database and the insurance company's secure servers, and at other times via
snail mail as in the case of gym reimbursement forms sent in by individual
employees. Different departments collected different data points into 52
different databases. Extracting, analyzing, and presenting this data back to
an employer in a useful and meaningful way is a challenge—demonstrating
lowering healthcare costs, improving employee productivity and retention,
and improving employee health and morale. Showing gains in these three
areas would have to be contextualized for local, regional, and national trends.

Although this seems like an overwhelming task, large health insurance com-
panies often have some of the most complete and comprehensive databases
of employee claims and historical trends. Combined with top notch teams of
statisticians and data scientists, it would seem pulling appropriate data for a
large client presentation would be easy.

However, there are many breakdowns in the data ecosystem that prevent
the systematic development and sharing of data presentations that can tell a
compelling story. Perhaps these four challenges sound familiar:

Data is stored in many separate, disjointed databases. Data is often
collected from dozens of separate reports. Getting these reports may
require a “ticket,” queuing it up for a database administrator (DBA).
Data retrieved might spark interest in another thread of reasoning and
require another “ticket.” And if reports are readily accessible without a
DBA, significant time might still have to be spent logging into different
data systems to collect the needed information.

⬛

Consolidating all this data into a single PowerPoint presentation
requires discretion. Synthesizing data and trying to make it look clean,
professional, and visually engaging isn't a simple feat. Different reports

⬛