The Big Data strategy needs to address all five dimensions — intent, data, people, process, technology. A Big Data strategy also needs to be dynamic in the sense that it is frequently updated with new input from a range of stakeholders (IT, the analytics team, business executives, and users) across the organization. Best practices from the most advanced department or business unit should be replicated into new areas, learning from past mistakes.
The Big Data strategy needs to be visibly supported by a C-level business executive. In order to drive interest, impetus, and funding, it should also embrace non-executive management as a key audience to drive broad adoption. One of the characteristics of Big Data Innovators that came out of the survey but is not discussed in this document due to limited space is that they have greater involvement from both executive and non-executive management in this way.
Both IT and lines of business need to be involved in Big Data and analytics strategy and operations. The role of IT is to put the right governance model and integration capabilities in place up front. For example, in a recent discussion with a large bank, it became clear that a successful Big Data analytics project focused on risk-adjusted profitability for large corporate transactions could not be integrated with its existing CRM system because IT had not been involved from the outset. The role of the LOB stakeholder is equally critical; too much IT focus at the expense of LOB often leads to a Big Data system that works perfectly well from an IT perspective but delivers no value to the business. A leading U.K. telco recently admitted that its €27.5 million spend on an information platform had yielded no business value. Although neglecting LOB stakeholders has a different effect to excluding IT, IT and LOB involvement are equally vital.
Having an expected outcome agreed from the outset will shape many decisions going forward during the project. Some projects are justified with a business case detailing what costs are expected to be reduced, or what revenue uplift is expected. For some infrastructure-focused projects, the business outcome may not be expressed in monetary form but could be expressed as faster access to information for the business, or the ability to see two different types of data together. Do not allow scope creep, as this can derail Big Data and analytics projects — there is always more information that business units need, but projects need to remain focused. Become accustomed to evaluating information-related projects in terms that are more than monetary — learning that a particular information source is of little value, for example, is a very useful input for future projects, although it does not yield direct monetary value.
by: Alys Woodward Posted 9 September 2015
Finally in 2015 with the rise of Big Data platforms we have a chance of deploying real-time information pervasively across organizations, not just in specialized, lucrative fields like financial trading, security, and casino risk management, but broadly across all industries. But where is the business value? Is it just a matter of convenience to have insights ready as soon as you need them, or is it more significant than that?
The transformation in retail over the last decade, starting with Amazon, meant that retailers were finally able to put products right in front of the customer in real-time as the customer's wants and buying journey at that particular point in time became clear. Real-time insights are transformational in nature.
There is also a correlation between use of real-time, advanced, and predictive analytics and a view of technology's power to transform businesses – 85% of 1,810 survey respondents agreed with this statement.
Amazon transformed book retail, book publishing, and book publication, in the first instance with their real-time recommendations; this was the secret sauce that gave online retail its advantage over bricks and mortar.
Manufacturing companies generally lagged behind in adopting traditional business analytics and data warehousing, because annual trend data was not hugely valuable. However, giving manufacturing plants real-time insights into the quality of production output as machines move out of calibration over time is transformational.
Once real-time insights are available, true business transformation becomes possible. The value of information delivered online is that it can be personalized and targeted. Analytics can show consumers what they didn't think to ask for but that actually they really want and need – the holy grail of any sales and marketing function!
IDC talks about Big Data as both evolutionary and revolutionary. Practices around cultural change, adoption and acceptance are evolutionary (meaning they follow on the lessons we learned from the traditional business analytics and data warehousing worlds) while the technology and the business transformation that Big Data enables are revolutionary. Real-time data is revolutionary: make sure you are ready for it.
by: Alys Woodward Posted 27 May 2015
Gaining value from insights is a journey. Whether you intend to become a fully data-driven organization, with real-time predictive information at your fingertips, or whether you need to speed up the financial close by a few days, it still takes a number of steps to do this. You can speed the journey up, by setting a clear corporate-level strategy, focusing resources, and allocating budget, but you can't cut out the early steps.
Focus on making progress along all five dimensions of IDC's Big Data & Analytics Maturity Model: People, Process, Technology, Data, and Intent. This will ensure that you are advancing along the journey to insight in every way. Leaving one or more dimensions as immature will slow your progress considerably.
The nature of insight-related projects is that they are not always successful. Sometimes they take too long, and the business requirements shifts, so the ROI that was initially expected is not delivered. Sometimes they come up against a hurdle that derails them: poor data quality, inability of users to use tools, a business change that means the requirements have to shift. Failing in any of these ways does not mean you can't succeed; it just shows you that that Big Data success was somewhat harder to achieve than you initially thought. You are far from the first to make that mistake. Achieving Big Data maturity is a challenge; only 5% of worldwide survey respondents are in the fourth stage, Managed, and none were in the top stage – Optimized.
Pick yourself up, dust yourself off, and claim some value from the failure. It could be in terms of a lesson learnt: for example, we need to conduct a data quality project; or we found out that we need to geocode our transactions in order to support a single customer view. Alternatively it could be a path correction: perhaps we found out there are no new insights for us in Twitter data; or we realized that users need mobile charts, not interactive PC tools. Take these insights into your own organization and use them for your next Big Data initiative. The only true failure would be to stop trying.