Overcoming Legacy Processes to Achieve Big Data Success
Blog: Colin Crofts - Business Process Improvement
For large companies with legacy infrastructure, changing to a Big Data focus isn’t easy — but AMEX proves it’s possible.
be a daunting challenge.
build their businesses and infrastructure from scratch in a “green-field” environment, most large corporations are saddled with disparate and
fragmented operational and analytical environments and processes,
characterized by business and data silos that limit the ability to operate
with agility, flexibility, and insight across customers and lines of business.
provide many advantages to large corporations, these larger firms cansometimes be challenged when it comes to innovation and
responsiveness. Yet it’s still these large corporations that continue to be
the bulwark of the economy, accounting for the bulk of business and
consumer transactional activity. Now, Big Data approaches that were
developed by the new economy firms, which enabled the flexibility and
rapid growth of these firms, are being adopted by mainstream
corporations to overcome legacy challenges and introduce greater
corporate agility and speed.
corporate executives, representing 59 Fortune 1000 companies, showed
that more than two-thirds of executives reported that their organization
had a Big Data initiative in production. These same executives reported
that investments in Big Data are projected to grow dramatically in the
coming years — with 75% of executives reporting investments greater
than $10 million, and 28% reporting investments exceeding $50 million
by 2017.
a huge business impact.
the financial services giant American Express (AMEX), which is
harnessing the power of its data to innovate and to streamline complex
operational processes. AMEX has opened its own new technology hub
in Palo Alto, led by former Google and PayPal executive Nik Sathe.
The new tech center “will focus on innovations in Big Data, cloud
computing, and mobile infrastructure,” according to AMEX CIO Marc
Gordon.
data processes end-to-end by migrating many traditional processes
from legacy mainframe environments to Big Data processing
environments, resulting in dramatic improvements in speed and
performance, significant reductions in cost, and notable increases in
responsiveness to customer needs.
Management for AMEX, comments, “Fact-based analytics have always
been in our company’s DNA, yet I have never seen such an analytical
leap as the one we have made with Big Data.” He cites several areas
where AMEX is focusing its Big Data efforts, notably in addressing
service excellence, generating billings and receivables growth, and risk management. A few examples:
– Using Big Data platforms, AMEX has been able to dramatically reduce
the time it takes to process the thousands of variables and billions of calculations needed to match customer card offers to card member
interests. It previously took years of processing time just to sift through
these massive volumes of data, but now it can be done in hours.
members on personalized offers, taking a 3-day process and reducing
it to 20 minutes on their Big Data platform.
in milliseconds and with greater precision, using machine-learning
models that leverage billions of historical transactions across millions
of card members, resulting in instant fraud alerts for customer protection.
mainstream corporations that are adopting Big Data solutions and
approaches with success. Here are a few suggestions:
mainframe environments to faster and cheaper Big Data approaches.
testing ground to identify suitable applications for Big Data.
data insights, rather than investing in large data warehouse initiatives.
professional, whom are conversant in new Big Data approaches and
analytics techniques.
they must demonstrate a willingness and flexibility to tackle their
complex legacy environments and data infrastructure. For those firms
that make the commitment, as evidenced by the example of American
Express, the payback is likely to be high. Their ability to compete for
customer success in the coming years may depend on it.
Leave a Comment
You must be logged in to post a comment.