Originally published on CMSWire
There is irony behind a discussion of the business value and metrics related to big data analytics, since there are a lot of people in the business world who still don't understand much about big data, including people who think they do have a handle on big data. Business analysts, IT, Marketing, Business Intelligence, Data Scientists, Upper Management – each of these roles can have a meaningful stake in big data analytics – and it's likely that each role has a different idea about the impact it can have.
Numerous misconceptions about big data – what it is, why it matters – make it difficult for organizations to know what to do with it and to understand the business value that can be acquired. Some organizations have been too quick to declare that big data is only hype or a big flop. Frankly, the same can be said for any business technology initiative - if the right approach and effort aren't undertaken, if participants don't understand why they're working with it, if beneficial strategies aren't in place.
Big data is not just one thing and there's not just one application for it. If big data analytics aren't producing good results, it's not the fault of the big data. Just like working with any data – you have to know what you want to do and why; you have to experiment and learn from the approaches that don't work; you have to adhere to continuous improvement, identify other needed data sources, and so on.
It's a very good idea for organizations to pursue a good understanding of big data and where it might produce value for current circumstances and for future direction. Initially an organization may decide that big data doesn't have a place in current strategies – but strategic thinking should also stretch into when big data analytics should come into play.
"Big Data" Isn't Always Big
A good many people think of extremely large datasets and high velocity data when "big data" is mentioned. But there are other kinds of big data sources that provide value to organizations, which means big data comes in disparate shapes and sizes.
Many big data sources are messy and complex, usually dubbed as multi-structured data (some erroneously say non-structured). These are the kinds of data that cannot be processed by traditional technologies and methods – data that doesn't fit neatly into a relational database. Multi-structured data and content comprise highly variable formats and semantics, such as log files, e-mail, social media content. Lots of big data isn't data but text or content that requires the additional dimensions of context and sentiment to attach meaning to the words mined from the original source.
The largest chunk of big data originates with machine-generated data. This data is continuously produced by an ever-proliferating tsunami of physical objects in every industry that are embedded with sensors interconnected through every kind of network, and that often operate with little or no human touch. In other words, the "Internet of Things". Machine-generated data is definitely big, but also complex and messy.
Tools and Techniques – Hard and Soft
The exciting news for organizations is that newer ways to work with difficult data and content text mining continue to appear on the scene. These technologies vary from expensive and complex to lower cost cloud platforms, as well as the Hadoop ecosystem of technologies. The lesser expensive technologies allow for open-ended experimentation with data, content mining and analytics, especially to fine tune approaches and find new and sometimes unexpected sources of insight.
Usable, relevant and insightful analytics output frequently doesn't come from big data alone, which can be fragmented and vague. Big data needs further qualification through the context and relationships that come from core enterprise data, to make sense of what big data may be communicating. Other sources of data and information, including master data, extend the accuracy and value of big data and provide essential context to align big data with guideposts like customer identity, products, and locations or channels of interaction.
Pivotal tools for big data analytics are not always technologies. Critical thinking is still one of the most important tools for any analytics initiatives, and must be undertaken by many roles in an organization.
Critical thinking must be used to determine what business problem you want to solve or which questions you would like to answer. For those involved in big data analytics certain questions should hover in the background: Why are we doing this? What are we trying to achieve?
Further critical thinking is needed to correlate results with other intelligence sources, with the experience of domain experts – and with common sense. Analytics results have to be reconciled with decision making processes and action plans. Then teams must review, improve, change analytics processes – or move on. If the focus is put solely on data, technology and methodology won't magically conjure relevant and accurate insights.
Business Value: Subjective and Different for Every Organization
Since a lot of big data work is still in early stages, establishing value is also just beginning. However there are an increasing number of real world successes published by a wide variety of businesses, from large enterprises to midsized and smaller organizations.
Obviously there are hard and soft benefits that can be attained through big data analytics. Trying to calculate an actual dollar value is tricky but not impossible. It requires mapping data and business processes to desired business outcomes, and then measuring the impact or success of the outcomes. But there is also gray area involved in such calculations.
Big data analytics are becoming a significant piece of the business performance puzzle for many organizations. Accenture Research took a look at the impact of analytics (including big data analytics) on high performance companies and found that:
- High performers are five times more likely to aggressively use information and analytics to improve decision making and business performance than lower performers
- Companies that invest heavily in advanced analytical capabilities outperform the S&P 500 on average by 64%
- Companies that invest heavily in developing analytical skills and adopting an analytical mindset recover quicker from economic downturns
Selecting Big Data Metrics
As with any strategic initiative, the right metrics are the ones that measure actions and events that relate to key business context and goals. But with the volatility of business and markets, such metrics will need constant assessment for relevance and effectiveness. It's likely that metrics will change continuously, hopefully to reflect actions taken in response to analytics results, evaluations of decision-making processes, the emergence of new markets, and company innovations.
To determine the right metrics, one approach is to start with what you want to achieve (goals, outcomes, effects) and then "reverse engineer" the process of attaining the end results. Identify the parameters of key steps or milestones that can be measured and that map back to big data analytics results. This approach also introduces a discipline of connecting data and analytics to events that matter to the business.
Former Forrester analyst and current IBMer James Kobielus offers these guidelines for measuring the benefits of big data analytics, in this case related to better customer intelligence and improved customer relationships:
Volume-based value: The more comprehensive your 360-degree view of customers and the more historical data you have on them, the more insight you can extract from it all and, all things considered, the better decisions you can make in the process of acquiring, retaining, growing and managing those customer relationships.
Velocity-based value: The more customer data you can ingest rapidly into your big-data platform and the more questions that a user can pose more rapidly against that data (via queries, reports, dashboards, etc.) within a given time period prior, the more likely you are to make the right decision at the right time to achieve your customer relationship management objectives.
Variety-based value: The more varied customer data you have – from the CRM system, social media, call-center logs, etc. – the more nuanced portrait you have on customer profiles, desires and so on, hence the better-informed decisions you can make in engaging with them.
Veracity-based value: The more consolidated, conformed, cleansed, consistent current the data you have on customers, the more likely you are to make the right decisions based on the most accurate data.
What Comes Next?
It's still very early in the big data story. Still lots to learn, lots of change coming to how analytics are done. Even for organizations with early starts and lots of resources for big data analytics, a great deal of change will come. And around the world, many organizations have barely begun to tap into big data, if at all.
The tools of big data also must evolve. Yes, a lot of the work of big data analytics is highly technical and mathematical, requiring sophisticated tools, algorithms and the right experts. But more tools need to be developed for the involvement of business roles in analytics processes and for assessing the validity of the results and the right direction for applications.
Image source: moniquestoner.com
About the author: Julie is an independent consultant and industry analyst for B2B software solutions, providing services to vendors to improve strategies for customers and target markets, products and solutions, and future direction. Julie has expertise in several solution spaces including: data integration and data quality; analytics and BI; business process, workflow and collaboration; digital marketing, WCM and social media; and the pivotal importance of user and customer experiences. Julie shares her takes on the software industry via her blog Highly Competitiveand on Twitter: @juliebhunt For more information: Julie Hunt Consulting – Strategies for B2B Software Solutions: Working from the Customer Perspective