Data has become critical to achieving competitive advantage in business. But with the growing industry noise around big data and a plethora of data management tools to consider, enterprises should develop a clear strategy to invest in the right platforms and capabilities.

“Managing data is not for the faint of heart,” says John Akred, chief technology officer at data science consulting company Silicon Valley Data Science (SVDS).

A key reason for this is that it takes a lot of hard work to piece together data from various sources within and outside an organisation.

And even when the work is done, it could be foiled by changes in executive strategies, mergers and acquisitions, as well as new technologies and regulations.

The data flood is another bugbear that could put many companies off course, before they even start to develop an enterprise data strategy.

With more data created than ever – from customer transactions to emails, text messages and Twitter feeds – many organisations are just trying to get all this data faster, so they can act more quickly to tap new business opportunities and prevent problems from arising.

Faced with such challenges, it’s hardly surprising that only one in 10 organisations have an enterprise data strategy. But those who do have one continue to reap the benefits – Singtel, for one, has been leveraging mobile data to optimise its mobile networks and to better understand its customers’ preferences.

 

What makes a good data strategy?

Many initial implementations of big data and analytics fail because they are not aligned with a company’s day-to-day processes and decision-making norms, according to management consulting firm McKinsey.

Thus, a good data strategy should be one that is business-driven, not technology-driven. It should directly support the goals of your business. It should be oriented to outcomes, to reliably deliver more of whatever your business needs.

Take Zillow, an online real estate database company founded in 2006 by former Microsoft executives, for example. The company formed its data strategy around its strategic imperative to provide products and services to help consumers with every stage of home ownership – buying, selling, renting, borrowing and remodelling.

In realising this strategy, it set out with several business objectives, one of which was to increase the completeness of its data by including data obtained directly from real estate boards, as well as broker-sourced, user-submitted and public data sets such as construction listings, foreclosure listings and market context. By using this data, Zillow was able to build and maintain the algorithms for Zestimate, a tool that customers can use to estimate the market value of some 100 million homes in the United States.

A successful data strategy should also identify the data most critical to an organisation. Examples include financial, product and customer data, which typically has long-term value and is used across multiple business processes. By combining inventory data with customer relationship management (CRM) data, for instance, a fashion retailer would be able to better optimise its supply chain and predict which items are more well-received by customers than others.

Organisations that want to get the most of out of their data strategies should also establish a data value chain that spans capabilities, such as capturing and organising data, as well as gleaning insights from the data to drive business outcomes.

 

Getting started 

Devising a data strategy starts with identifying the strategic imperatives of your business before deciding on the data, technology and analytics techniques required.

If you are a fashion retailer looking to better understand your customer’s preferences, you could begin with a proof-of-concept experiment that combines data on product attributes with actual sales data. And while doing so, you should identify priorities and constraints for technical workloads, and ascertain if the data you need is available.

A modern data strategy would vary based on the size of your organisation. Compared to start-ups and small and medium-sized enterprises (SMEs), enterprises have to maintain day-to-day operations, which must not be affected by experimental projects. As such, only the necessary analytics techniques that fits into an organisation’s IT infrastructure should be chosen.

For example, while behavioural analytics could deliver better recommendations for online customers, it may be less of an architectural fit compared to, say, predictive modelling. Furthermore, in behavioural analytics, the data such as a customer’s personality traits may not be fully available, while predictive modelling can be easily undertaken with actual sales data.

When it comes to technology choices, always go for those that meet the needs of your technical use cases – often, it’s not about choosing one technology over another.

An e-commerce company, for instance, might need several types of databases that perform different functions, such as a graph database for social network analysis, a key-value database for session data and shopping-cart functions, and a columnar database for capturing sensor data from its delivery fleet.

All these technology decisions, along with your strategic imperatives and business objectives, should be detailed in a road map and action plan that can be easily communicated and followed throughout your business.

 

Experimenting to drive business value

But as with most business plans, nothing is cast in stone. Organisations should consider the possibilities of emerging technology, understand new data assets that become available along with their business needs, and even explore working with partners experienced in delivering business value from data analytics.

To this end, experimentation is a key part of a successful data strategy. Once your business case has been proven to stakeholders, it will be easier to scale up your big data deployment across the enterprise. Meanwhile, continue to explore technology options even if they don’t fit into your current infrastructure – you never know when things might change with the rapid pace of advancement in the technology industry.

For more insights on how you can utilise big data and analytics strategically for business advantage, get in touch with DataSpark today.

 

Get in contact

  • This field is for validation purposes and should be left unchanged.