Enabling big data analytics initiatives is much more than just collecting all the data from as many sources as you can and then throwing a data scientist in the mix to solve the problem. A very critical component is taking a business-first approach to your big data initiatives and focusing on the information that will help you meet your business objectives.
A big data analytics initiative is usually (but not always) started due to the company’s eagerness to begin a path toward having a more digitally centered business. However, before starting on this route, questions need to be asked and answered, and challenges must be identified and overcome. There’s no point in starting down the path of transforming your big data environment if you have not identified the factors that will determine whether the initiative is a success or failure.
Here are a few questions to start your conversation:
- What is your business trying to achieve? This is the number-one question that always should be asked. Without an understanding of the business use case and requirements, a big data initiative quickly will become bogged down in internal use-case requirements.
- Who is the stakeholder for the project? If a clear and defined stakeholder is not determined, there cannot be a drive behind the requirements to bring business units and IT together to create the correct atmosphere to deliver success.
- Will the project have the ROI you want? In this age of digital transformation, all initiatives should be able to show a return on investment.
- Will the outcome be successful? You can’t just throw data and data scientists together and wave a magic wand. You need to understand that you are not doing a big data analytics initiative just for the sake of doing big data. You are driving toward outcomes.
When you have answered these questions, you will have a better understanding of some of the challenges that you may encounter in trying to be successful with your initiative. These can be as simple as the fact that the enterprise is not storing or does not have access to the correct types of data, to the more complex questions of identifying factors that indicate the success of the project.
Every industry faces big data challenges: Healthcare has security, manufacturing has data velocity and IoT devices, retail needs to overcome the challenge of the proliferation of social media data, and finance struggles to keep up with the volume of data required for fraud detection. There are many approaches to dealing with these challenges, but I like to develop a strategic plan at the outset to understand the trade-offs or critical decisions that may have to be made as an enterprise moving forward. This plan should consider things like whether speed of growth or strong performance is important, and how capabilities can support these objectives.
The plan should include stakeholders, data requirements – what, where, and when – along with the methodology and tools that will be used to integrate this data. In the plan, key milestones should be identified – the initiative should not be open ended, it should have a fixed timescale. Finally, the plan should consider your internal talent pool – do you have the resources to implement the plan? This type of planning cannot be done in a vacuum and requires IT and business alignment at a scale not currently practiced in many organizations.
Some of these challenges can be offset by partnering with innovative technology partners and organizations that can work with your enterprise to assist you in your big data initiative.
When you have your plan and possibly a partner to assist you, it is then possible to move forward from the planning stage to the implementation stage. This stage brings further challenges, and not only technologically in the form of legacy systems that can’t be integrated into the data flow. When you have the data, you are faced with the question of how and where to store it for use in your analytical model. This can be overcome in a number of ways, and I always advise clients that big data should not always be about just Hadoop.
Most enterprises already have an established infrastructure in place that provides the ability to store data as well as deliver standardized reports and ad hoc queries for lines of business – there should be no reason to rip and replace this infrastructure just to enable a big data analytics initiative. Looking back to our planning stage, we can infer what type of data we will be required to use to achieve our aims, and this should enable us to build out a Hybrid Data Eco System (HDES) that allows data to be integrated, interrogated, and analyzed by any number of departments. This HDES allows you to use an existing infrastructure and to add Hadoop or any number of data stores or analytical platforms that suit your industry vertical and, more importantly, tailor the solution to fit your enterprise.
There is no one size fits all when it comes to big data.
Finally, how do we identify and measure factors that we can use to judge if our implementation has been successful? Determining whether a project has been successful can be difficult, but if you use the plan methodology and ask the above questions, you can identify one or two critical identifiers of the implementations success or failure:
- Adoption. Has the outcome of the initiative’s analytics been adopted by key business units? How many users are utilizing the data, and what are they using it for?
- Business Process. Since the enablement of the initiative, has the process that the question was asked about improved? For example, are customers more satisfied? Have you stopped wasting materials in your manufacturing processes? Have you seen a drop in fraud cases?
- Cost. Was the process cost efficient? Do the benefits outweigh the costs? This is a very difficult metric because once a big data initiative is started, it tends to spawn more use cases and builds a life of its own, making it difficult to measure against its initial delivery model.
Recently, we launched our annual big data and Advanced Analytics adoption report in partnership with IIA. The survey results give a good look into what top management sees as barriers to digital adoption, and the research gives insight into the state of big data adoption in the United States. Read more about the researching findings here.
As part of your enterprise’s digital transformation, a big data analytics initiative can lead to insights that enable you to change your companies’ image, performance, or profitability. However, it is only one step on the road to becoming a completely digitally driven, data-centric organization.
Raman Sapra is the vice president and global head of Dell Digital Business Services. His team leads a business-first approach to digital, enabling organizations to create new business models, drive enhanced customer engagement and loyalty, empower a mobile work force, and deliver operational efficiencies. Digital Business Services uses a robust consulting methodology to define digital roadmaps with customers and delivers comprehensive solutions including advanced analytics, business intelligence, omni-commerce, mobility, customer engagement, digital marketing, digital application development, and cloud, Internet of Things (IoT), and digital process orchestration. Raman has close to 20 years of experience in the information technology industry.
Subscribe to Data Informed for the latest information and news on big data and analytics for the enterprise, plus get instant access to more than 20 eBooks.