The value of big data consulting is measured in some form of tangible ROI. Just as marketers can have trouble justifying the fiscal returns of social media campaigns, it can be difficult proving ROI from a big data consulting engagement. To demonstrate ROI you need to agree on the scope and objective of the project, and the metrics to be applied to measure success.
Demonstrating big data ROI is challenging. In 2013, IDG and Kapow Software asked 200 IT professionals about big data. More than 85 percent said that big data can help businesses make better informed decisions, but only 23 percent said they had seen big data success, and 52 percent termed their big data projects “somewhat successful.” The conclusion drawn was that big data projects take too long, cost too much, and you can’t pinpoint the insights without expensive big data consulting.
Big Data Is Not the Product
You need to have a common understanding of how to define big data ROI. The data itself has no intrinsic value; it is the insight derived from the data that creates value. Unfortunately, you don’t know what you don’t know, so the potential value from any big data project is unknowable because the possible insights are uncertain. Conventional methods of measuring ROI for IT projects don’t apply.
Since the returns come from big data insight, or more specifically the actions recommended as part of that insight, ROI has to be viewed in the context of long-term deliverables. It’s not the return on the big data insight, but the return on the results from the action that is the result of the big data insight. So as part of big data consulting, you want to align ROI with appropriate expectations then work toward a positive outcome by converting insight into recommended action.
The first step in any big data consulting engagement is identifying the business problem and converting them into well-architected use cases. If you focus on the business problem rather than the technology to deliver the project, you will have a greater chance of success.
Use Case Dictates ROI
Big data use cases typically fall into different categories, and each with a different ROI:
- Supporting business decision-making – Big data projects allow you to create analytical proofs using transactional data stored in SQL data warehouses and combine it with external data or real-time data streams to perform “what if” scenarios based on different performance variables. To gauge ROI, you need to assess the cost of the big data engagement against the accuracy of the insight, and how that insight converts decision-making into dollars and cents.
- Identifying new business opportunities – Big data use cases can be used to test new market and new product strategies by analyzing product and customer data and modeling likely market acceptance. If the data tells you that a new initiative is likely to be a bust, the ROI is measured in the money saved on a project that is likely to fail. You also measure ROI from the returns yielded by retooling the strategy, retesting the new strategy using big data analytics, and identifying a more promising strategy.
- Improving business performance – By breaking down big data silos you create transparency into business processes, including inefficiencies, so you get some ROI from the outset by making data accessible throughout the organization. When you apply use cases to assess business performance, you should be able to identify flawed processes and ways to improve performance. ROI is measured in fiscal snapshots of “before” and “after.” For example, by using big data to analyze delivery patterns, UPS was able to save 8.4 million gallons of fuel and $30 million.
- Replacing human decision-making – Big data algorithms can automate routine processes, eliminating human error and reducing staffing costs. For example, harnessing the Internet of Things (IoT) to use machine data to power big data analytics can automate manufacturing. Returns are measured in saving and revenue generated by automated processes. How much is saved in time, resources, and money by automating workflows? Is production increased or productivity enhanced?
- Improving customer targeting – Big data analytics can tell you a lot about consumer behavior, buying patterns, likes, and dislikes. Analytics can help shape marketing programs, retail strategies, advertising, and fulfillment. ROI is determined by increased sales, improved customer loyalty, and related performance metrics.
As part of ROI you want to make sure your big data initiatives are repeatable, provide ongoing insight, and lasting value. You also want to track the cost of big data technologies such as data storage, servers, software, redundancy, backup, licenses, etc., and map that to the returns from big data insight. Bear in mind that the cost of supporting technology, such as storage, will change with time.
How do you sell big data ROI to your customers?