Tagged: Embrace Toggle Comment Threads | Keyboard Shortcuts

  • admin 9:47 am on February 19, 2017 Permalink
    Tags: , , Embrace,   

    Teradata Customers Embrace the Cloud 

    Teradata Press Mentions

     
  • admin 9:51 am on August 18, 2016 Permalink
    Tags: , Athleticism, Embrace   

    Embrace Analytic Athleticism 

    Latest imported feed items on Analytics Matters

     
  • admin 9:46 am on November 19, 2015 Permalink
    Tags: , Embrace, , sucks, , tightens   

    Teradata sucks up IoT data tightens Hadoop embrace 

    Teradata Press Mentions

     
  • admin 9:54 am on August 27, 2015 Permalink
    Tags: , , Embrace, , Pathway,   

    Run Before You Walk – Only Pathway to Embrace Analytics 

    A large Australian Government department was recently looking to set up their internal analytics capabilities. I was invited to present at their kick-off meeting where their IT team proposed the approach of first bedding down a reporting infrastructure and then afterwards move into complex data analysis – the phrase used was “we need to walk before we can run”. Based on my experience in successfully setting up analytics in many organisations, I suggested a different approach.

    This approach takes into account the highly unstructured nature of analytics projects and the possibility that all analytics ideas may not be successful. It enables quick testing of analytics opportunities with the ability to move rapidly into production if needed. The approach speedily searches the data landscape to conceive, test and deploy analytics ideas and supports business in fact-based decision making. Important enablers of this approach which may be implemented through business processes or software include the following.

    1. Access to operational data: In many organisations, access to data is limited to IT teams with the data scientists restricted to accessing cubes and reports. This regime slows momentum forcing data scientists to waste time foraging for data, building their own duplicate data marts on Excel and then having no means to operationalize ideas that result in good business outcomes. Direct read access to operational data for data scientists will remove these blockers.
    2. Space and tools to prototype: Business value of operational data increases when it is linked with other external data. For instance, my retail analytics team correlated sales with demographics from census using a regression model to develop a prototype for identifying new store locations. The space and modelling tools enabled the development of the prototype as well as identification of the data feeds needed for the production application.
    3. Support for rapid deployment: Once an analytics prototype demonstrates ongoing business value, it is important to leverage the prototype to move the analysis into production without a long and protracted IT engagement. This means:
      1. Using the prototype as the specification rather than have to write the requirements again;
      2. Restricting data feeds to those identified by the prototype; and
      3. Continuing to publish the analytics insights to business while the application is being developed and using that engagement to define the user interface.
    4. Strong governance: Given that not all of the analytics projects may be successful, a strong governance process is needed to ensure that databases created by abandoned projects do not hang around indefinitely, and that production applications do not access any of the prototype databases.

    Using these enablers, analytics teams can easily build momentum with the following.

    1. Analytics project pipeline: Data scientists need to have a pipeline of projects with multiple stake holders and business functions in order to ensure that analytics exposure is as broad as possible and mitigate the risk of some analytics opportunities failing or not being adopted by business.
    2. Early wins: When I embarked on setting up analytics capabilities at a manufacturing company, I was given 3 months to show the value of advanced analytics. This constraint provided the impetus to speedily identify and prototype an analytics opportunity. The insights from this prototype funded both the production and further investment into analytics. Analytics ideas need to move from conception to prototype in 6-12 weeks, with unprofitable ideas abandoned much earlier than that.

    And that brings me to the title. No baby ever learnt to walk by taking slow deliberate steps. Similarly no organisation is likely to succeed in setting up a brand new analytics capability by laboring over a couple of years to deliver an analytics infrastructure — except in the unlikely scenario of a patient management willing to continue to invest in unproven potential; unchanging market landscape and competitors at a standstill!! Even then what is delivered at the end may not be what was originally envisaged and the organization gives up analytics as a failure.

    So, my advice to organisations embarking on the analytics journey is: rather than walk first with business intelligence and reporting and then attempt to run with analytics on a reporting infrastructure, run first by exploring multiple analytics opportunities to yield quick business value, and then walk by deploying the reporting of the analytics so as to get ongoing value from the insights. This run before you walk” approach will set up the organization for data driven decision-making using both advanced analytics techniques and regular business intelligence reporting.

    Bhavani Raskutti is the Domain Lead for Advanced Analytics Teradata ANZ . She is responsible for identifying and developing analytics opportunities using Teradata Aster and Teradata’s analytics partner solutions. She is internationally recognised as a data mining thought leader and is regularly invited to present at international conferences on Mining Big Data. She is passionate about transforming businesses to make better decisions using their data capital.

    The post Run Before You Walk – Only Pathway to Embrace Analytics appeared first on International Blog.

    Teradata Blogs Feed

     
c
Compose new post
j
Next post/Next comment
k
Previous post/Previous comment
r
Reply
e
Edit
o
Show/Hide comments
t
Go to top
l
Go to login
h
Show/Hide help
shift + esc
Cancel