Internal And External Data Strategies Indispensable To A Cohesive Analytics Program

Hermann Wimmer, President - EMEAAP, Teradata Corporation talks about the value of data and extracting that business value by the right strategic approach to data storage and analyses. In conversation with Biztech2.com, he sheds light on optimising data to make analytics pervasive across the enterprise.

Managing centralised or de-centralised BI – which one do you think is a better approach?

A good approach for most types of data is to store it in a centralised location but analyse it in a de-centralised fashion. The next step is for the company to allow the external customers to utilise the enterprise data warehouse for analytics and procuring information, which will contribute to further de-centralisation of data analytics.

Many enterprises are now investing in BI and Hadoop. How do you see data quality and integration investments in this area?

In the future, data volumes will grow so vast that it will not be possible to store all the data in an enterprise data warehouse. An integrated stack with Hadoop will become necessary. To address this we recently announced the Teradata United Data Architecture (UDA) which is an architecture that integrates Hadoop and Teradata.

More from Biztech

Future Group - Reliance Retail Deal approved by CCI RBI ban on cryptocurrencies takes effect; prohibition could force investors to tap the black market

From a data quality perspective, one of the things that is in favour of Indian enterprises is that they do not have legacy systems that generate data which is very difficult to manage. There isn’t this additional layer of data complexity. Most Indian companies are taking the approach of doing data quality at the ETL (Extract, Transform and Load) stage itself. But, I have not seen many companies do a lot of standardisation of data across the enterprise or work at the metadata level. The data quality initiatives that take place at the ETL level are more along the lines of solving de-deduplication. The CIO now thinks ‘Can I do data integration and cleansing at the very beginning?’ The discussions have moved forward to making data integration a real-time consolidation of data before it becomes siloed in the various departmental data marts.

These initiatives are driven by business value that is obtained out of this data . If organisations do not see business value in investing money in these exercises then they are not ready to make this commitment. In such a case they should stick to web analytics and smaller projects until they have a definite plan to realise ROI for analytics.

What is your pick when it comes to the build vs. buy debate for Hadoop?

It differs from organisation to organisation. The value of data integration to support analytical capabilities is where the key value of these investments lies.

Knowledge from data is a key differentiator for enterprises and these core competencies will not longer be considered as an option but a must-have. Globally we have seen that more or less most large organisations have increased spending in developing these capabilities in-house even in hard economic times because these are vital to competitive edge and businesses would not want this data to reside outside the enterprise. No longer is it a cost discussion but a value discussion.

Enterprises need to decide if they have to do the integration work themselves or whether they hire someone to work with them to build an information infrastructure. What enterprises can also do is rely on one strong partner who will do all of this for them.

SMEs have to keep the lights on and innovating with small budgets is an issue. Do you see cloud analytics as a service becoming available to them in the near future?

Over the next few years I am very positive the market will see a cloud offering of analytics for small organisations . For SMEs, it is not only a cost to buy the technology but also to have people available to manage and utilise that technology to generate value. These analytical resources also need to have a good understanding of the business. If organisations are not willing to invest in long term strategies that include resources like analysts then buying a data warehouse and spending money on that is rather a failed cause.

Outsourcing BI and analytics is already big for digital marketing and CRM initiatives . Smaller enterprises lacking volumes of data may have settled into long-term partnerships with these vendors. Larger enterprises are experimenting with outsourcing vendors as well before taking on these capabilities in-house The mid-sized market will be a major area of growth as some mid-sized enterprises will experiment with these options while some might follow the SMEs into outsourcing partnerships. Outsourcing the social media analytics projects to drive marketing initiatives is just a starting point that we are noticing with enterprises today.