The barriers to extracting business value from big data can seem daunting, but they can be overcome through a systematic plan; one that breaks down the challenge into a series of nine sequential steps that will enable organisations to take advantage of this valuable and growing asset.
1. Define responsibilities
The onus of collecting data should be shared by the IT and analytics teams, but analysis must be the sole responsibility of analytics professionals. Similarly, only functional leaders, such as the chief marketing officer, chief financial officer and chief procurement officer, should be responsible for identifying areas within their respective functions where big data could drive value. The team may want to appoint a big data programme sponsor for each function, and work closely with them to discover and locate the types of information that would improve business outcomes. Most importantly, however, the programme sponsor should try to get functional buy-in and identify big data opportunities.
2. Identify the right questions
Senior executives will have an easier time winning buy-in from business functions if they demonstrate how big data might be valuable to them. Simple questions such as “What would you like to know about your business, and how can data help you with it?” are a good place to start. Such questions can spur the functional experts themselves to start asking the more fundamental questions that can unlock the value of data. For instance, marketing professionals could ask, “What is the value of a ‘tweet’ or a ‘like’?” or “What is the optimal price for our product right now?” The ability to ask the right questions is the key to succeeding with big data.
3. Identify data worth analysing
Valuable business insight can come from many sources, including social media feeds, “dark data”, which is currently unused but has already been captured, machine instrumentation, and operational technology feeds. It is important to explore these sources and to experiment with new ways of capturing information, such as complex-event processing, video search, and text analytics.
>See also: Staying afloat in a sea of big data
4. Select business functions to lead
It is wise to launch big data initiatives in business functions that are most ready to collect and analyse data and for which the potential payback is high. Functions such as marketing, customer service, supply chain management and finance are poised for maximum growth. If system readiness is not an issue, these are usually the right places to direct initial investments.
5. Match big data with business functions
Some big data programmes can be implemented in a variety of settings, but most are suited to specific functions. For example, customer functions such as marketing and customer service can use big data for targeted advertising and loyalty management. Finance functions can use big data for tasks such as improving credit risk assessment, through multiple big data-supported credit risk assessments that factor in hundreds or even thousands of indicators. Supply chain and procurement can use big data for dynamic route optimisation.
6. Determine whether big data will yield valuable insights
Making the business case for a big data initiative will be easier if it can be shown that it creates new value. For instance, if a marketing department is currently segmenting customer profiles using standard demographic indicators, would there be additional benefit in analysing attitudes and preferences through text and speech analysis? When comparing traditional business intelligence to big data, businesses should consider what the limitations of the structured data they are capturing today are and what extra value will be obtained by collecting external, context-specific and unstructured data. They should also consider how data will be found and collected, whether the business would act upon the insights gained, and, ultimately, whether the extra business value is worth the investment of time, energy and money.
7. Assess complexities and prioritise
An organisation should begin its big data experimentation with an initiative that is not too demanding. It is helpful to keep in mind the complexity of both the type of data and the type of analysis it will require. Different types of analysis present varying degrees of complexity. Generally speaking, descriptive analytics like social media sentiment analysis, which provides insight into what has happened, are relatively easy to do. However, diagnostic analytics, which provides insights into why it happened, predictive analytics, which provides insights into what will happen, and prescriptive analytics, which provides insights into how businesses can make something happen, are increasingly complex to conduct.
>See also: How to build a big data infrastructure
8. Assess the IT architecture
An organisation’s traditional information architecture may not accommodate massive, high-speed, variable data flows. Many traditional and even state-of-the-art technologies were not designed for today’s or tomorrow’s level of data volume, velocity and variety. As such, organisations need to consider a variety of methods to upgrade their infrastructure to support big data, which is no small feat. Building everything from scratch takes time, and buying everything is expensive. Therefore, finding the right combination of insourcing and outsourcing requires careful consideration.
9. Start building a team
Big data initiatives require multidisciplinary teams of business and technology experts. Every team member – business analyst, programmer, data scientist and data visualiser – will need to have cross-functional familiarity. To build this team, businesses first need to break down their talent needs and then scan the internal landscape for those skills. They can then hire people with necessary or related skills if they are not available or cannot be acquired by cross-training existing employees.
Big data analytics is not a passing fad. It will be a central means of creating value for the organisation of tomorrow. However, it represents a major change in the way that businesses and other organisations operate and will require a new mind-set and capabilities. Given that, many organisations are struggling to know where to start in the realm of big data. A step-by-step approach can make the transition seem less daunting, helping to overcome the barriers and minimise the difficulties that are bound to occur along the way.