Case Study DatabaseUSA Relies on Alteryx to Ensure Clients Get

Case Study DatabaseUSA Relies on Alteryx to Ensure Clients Get
Case Study
DatabaseUSA Relies on Alteryx to
Ensure Clients Get the Most Accurate
Lists and Data Possible
CHALLENGE
DatabaseUSA wanted a flexible ETL
solution that would increase the
quality and accuracy of its data
while reducing the time and money
spent on data processing tasks and
related training.
SOLUTION
Using Alteryx in a unique self-hosted
environment, DatabaseUSA created an
elegant workflow that easily blends
hundreds of data sources and ensures
industry-leading accuracy with little
to no coding required.
RESULTS
• Deeper Insights: Blended hundreds
of data sources, including call
center data, online data, and
utility data, into a clean, accurate
database upon which customers
can depend
• Hours vs. Weeks: Compressed
complex processes from several
weeks into hours, enabling the
delivery of ad hoc requests as the
DatabaseUSA team tests
new business rules
• Intuitive Workflow: Replaced script
and low-level code with a simple
drag-and-drop visual workflow
Introduction
DatabaseUSA.com is a leading provider of business and personal mailing lists, email
lists, sales leads, and related database products serving most of the Fortune 1000.
The company’s lists and online products are used by customers large and small
for everything from marketing, analytics, and customer profiling to background
checks, business verifications, and employee queries—and have been highlighted
on the television news program “60 Minutes.” One of the few database providers
that still compiles its own complete databases from original sources, DatabaseUSA
performs rigorous verifications online, via its call center, and by way of proprietary
algorithms, resulting in greater than 95 percent data accuracy.
Situation
In the competitive landscape of database marketing, offering the “biggest lists”
alone is no longer enough. “Even though DatabaseUSA is only four-and-a-half years
old, many of us are industry veterans with 10 to 20 years of experience,” relates
Jerry Dailey, Vice President of Product and Content at DatabaseUSA. “That’s a lot
of accumulated knowledge. We know firsthand that more data is not better—
higher quality is better. To that end, we’ve sought to provide more features,
greater depth, and better data accuracy than any of our competitors.”
“Prior to Alteryx, raising that bar of data quality was not easy,” recalls Erich Kaiser,
Vice President of Data Processing and Special Projects at DatabaseUSA. “Our
consumer database has details on 238 million individuals and 164,000 households,
and our business database details approximately 14 million businesses. That’s a
lot of data to process, check for accuracy, and update on a monthly basis.”
With so much raw data as part of the monthly build process, hitting deadlines for
fresh builds was a Herculean task. To deal with this reality, DatabaseUSA sought
individuals with high-level data expertise—but that brought high-level cost.
Remembers Kaiser, “When we started out, we had to hire people familiar with a
Linux environment and shell scripting to support our database builds. Unfortunately,
those people don’t come cheaply; each hire cost us $100,000 to $150,000 and forced
us to put all our eggs in one basket because of the intense specialization required.”
Because the company was working only with flat files from hundreds of different
sources and multiple directories, including data from utility companies and call
“We never really had a lockeddown process before. But now,
with Alteryx, we can string these
components together, create
a build process, and lock down
that process. And if there is any
anomaly during quality control,
we know exactly what component
to go back to and correct based
on what we are seeing—and we
don’t miss deadlines.”
—Erich Kaiser, Vice President of
Data Processing and Special Projects
at DatabaseUSA
data from its call center, it needed to
develop a significant amount of new,
low-level code to get the data into its
database successfully—and accurately.
“Every time we introduced new data we
had to custom-code solutions to make it
fit,” continues Kaiser. “Not only was our
initial platform expensive, but also the
knowledge required to build and maintain
it was prohibitive and difficult to scale.”
environment. “Although most customers
access Alteryx remotely, we needed to
run it on our hosted servers due to data
sensitivity issues,” says Daily. “The Alteryx
team understood our needs and helped
us make this a reality. As of now, we have
two rack servers, two application hosted
servers, and a data analytics database
server with Alteryx in the mix.”
Solution
Since deploying Alteryx, Kaiser has been
impressed with how well the solution
has risen to new challenges. “There are
several tools in the ETL realm,” Kaiser
notes, “but none near as strong for data
quality, and that’s key for us. All of our
ETL work is now done by Alteryx, which
blends and cleanses all our external
sources and call center data.”
Driven by the need for data accuracy
and scalability and to simplify and codify
its processes, DatabaseUSA actively
searched for solutions before landing
on Alteryx Analytics. “We did an extensive
review and comparison of databases,
platforms, and tools, including SQL Server,
SSIS, Clover, and Melissa,” recalls Dailey.
“But, ultimately, Alteryx did a better job
of scrubbing and enhancing the data
and providing the ETL we needed.”
While enthusiastic about Alteryx,
Dailey and his team weren’t certain how
much of the heavy lifting Alteryx would
ultimately be able to provide them in the
areas of ETL and overall data blending.
“Initially, we weren’t sure how far we
could take Alteryx and thought we might
have to use other tools to help us with
our needs. But, so far, with the exception
of our platform and server environment,
we’ve now used Alteryx to do just about
everything. Alteryx helps us blend, clean,
input, and update the data, test its
accuracy, maintain the required database
relationships, and deliver the data the
way we need it, when we need it.”
Kaiser agrees, elaborating that
Alteryx’s successful integration into
DatabaseUSA’s methodology stems
from how well it handles tasks like
standardization and duplicates.
“We use Alteryx where all of our data
sources come together. For example,
address data can be represented
in a variety of different ways in terms
layouts, formats, values, and fields,”
explains Kaiser. “After we use Alteryx
to standardize, clean, and de-dupe
the data from the different sources,
we bring it into our own database to
identify which records are truly new.”
3345 Michelson Dr., Ste. 400, Irvine, CA 92612
+1 714 516 2400
www.alteryx.com
DatabaseUSA appreciated how amenable
the Alteryx team was in supporting
its unconventional deployment
requirements—namely, to a hosted server
Alteryx is a registered trademark of Alteryx, Inc. 6/16
Results
Dailey agrees, adding, “Before Alteryx,
I had to do a lot of low-level complex and
custom coding. But now, with Alteryx,
using fuzzy matching we can apply
several different types of matching
algorithms on our data by simply
changing a configuration setting.”
Prior to Alteryx, DatabaseUSA dealt
with each step in its build process as a
discrete code issue. “Each step, including
transforming, merging, identifying
duplicates, removing duplicates, and
assigning address standardization
enhancements, was its own challenge
and deadlines were easy to miss,” recalls
Kaiser. “But now, with Alteryx, we can
string these components together,
create a build process, and lock down
that process. And if there is any anomaly
during quality control, we know exactly
what component to go back to and
correct based on what we are seeing—
and we don’t miss deadlines.”
“Likewise, with our process locked down
and simplified, database build processes
have easily shortened from a month to
a week,” asserts Dailey. Simplification
of the workflow has also relieved the
pressures and costs of hiring high-level
expertise. “By using Alteryx, I can now
recruit new, young computer engineering
students, train them, and get them up to
speed applying business rules in a very
short time. We can accomplish more in
two months now with Alteryx than we
could in six months with our former
environment, custom coding,
and expensive hires.”
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement