So far in my series of articles on Big Data I’ve explored what Big Data is, some of the ways businesses are using it, and what it can do to help grow your business.
In this article we’ll take a look at how a company should organise itself to make best use of the opportunities from Big Data and the processes it should follow to deliver Actionable Intelligence.
Business leaders around the world are recognising the strategic edge that Big Data brings. At the recent Big Data World conference in Singapore, many of the executives that I spoke to saw the value of Big Data but getting to the stage of turning that into action was proving difficult.
|More from Think Business about Big Data|
|Big Data: What’s the big deal?|
|Using Big Data to amplify your customer experience|
One strategy we employed successfully when I led Supply Chain Intelligence at a global luxury cosmetics firm in the US was to establish a shared services group on Actionable Intelligence reporting to a senior leader in the business unit that was most in need of performance improvement.
This group reported to the senior vice president in the Supply Chain Centre of Excellence to improve inventory management and customer service. Although the group’s home was in Supply Chain, the shared services supported all functions including finance, marketing, quality assurance, and customer service.
Your organisation may need help in other areas such as sales and marketing or finance. To get the best impact, begin by choosing an area with the following characteristics:
- One with leaders with a vision for Actionable Intelligence
- Where critical business needs are clear, but the facts to improve the situation are lacking
- With resources and time to establish a unit to deliver the capabilities
After identifying the organisation, location and resources, the next step is to build the right process.
The diagram below illustrates a simple yet effective cyclical framework to help business leaders translate Big Data into Actionable Intelligence.
The SWAT process breaks down into four stages:
S – Strategic Business Questions
Every business has its priorities and overall strategy and any Big Data strategy should not be implemented without giving these thorough thought.
As business leaders, it is important to understand that Big Data is intertwined with your overall strategy.
Depending on the strategic priority of your company, direct your Big Data application accordingly. Don’t try to source all the data; focus just on the “burning platform”, using the strategic business question as a guide.
W – Wrangle data
While many people have been advocating the mass collection of data, I beg to differ. First, the data must be focused solely on your key strategic question. This enables the acquisition team to collect and review the data quality.
The quality of your insights is only as good as the quality of your data sets. So before collecting any data, pause for a moment to consider the following:
- What kind of data do I really need to answer the strategic business questions?
- Where can I find them?
- What are the IT infrastructure, tools and policies needed?
- Who are the people in my team with the required skillset to analyse the data?
- How can I do all these in a cost-effective manner?
A – Answer with Visualisation
To quote Napoleon: “a good sketch is better than a long speech”. Raw data, unprocessed and displayed on an Excel spreadsheet, is difficult to use to garner insights.
Fortunately, innovation in the field of data visualisation has created business discovery tools such as Qlikview to provide user-friendly yet concise executive dashboard and report capabilities.
Such tools enable the result, but the process is critical to success so it is important to have regular design sessions with your organisation’s business users asking:
- How will the answer be used?
- How will the standard operating procedures change?
- Are the right critical issues being highlighted for review?
T –Take Action
With a visual answer supported by good data, we can then move into the Take Action phase. Review the answer with the project sponsor and determine the short-term and long-term actions to be taken, including:
- What actions can we take right now?
- Can we capture the value saved or earned?
- Does the current process and organisation need to change to start making fact-based decisions?
Inevitably, the first answers will generate more and deeper questions. Start the SWAT Framework again and quickly provide answers again. This is how repeated use of the SWAT Framework enables fast continuous delivery of Actionable Intelligence.
So, does this work? “Yes!” says Jack Levi, Director of Process Management at courier firm UPS and creator of the amazing ORION (On-Road Integrated Optimization and Navigation) project.
Levi manages a team of mathematicians who build algorithms that know better than any human how UPS drivers ought to plan their routes, helping the firm shave millions of miles off delivery distances every year.
When he first brought up the project, UPS management was incredulous, and so were the drivers. So Levi set up competitions between the drivers and the system to refine the model and to acquire more data. Following this approach and engaging the drivers paid off.
“Starting small shows the project is feasible, optimal, and more importantly implementable. So we quickly tested lots of prototypes,” he says. “Most importantly, we showed that the pilot project actually cut-down miles which meant we delivered real monetary value.”
All of the data points focused on one outcome: more efficient delivery.
At the Big Data World conference, more than 50 per cent of the participants said they felt challenged by a lack of people capability, Big Data tools, and resources. By the end of the session their key takeaways were to start small and focus on answering strategic business questions. One executive said that he would go back and convince his managers by starting with small projects that showed results.
At another recent presentation I gave in Tokyo, few of the 80 participants arrived knowing what Actionable Intelligence was – they had tried big data tools, but had not delivered the expected results.
By sharing the SWAT Framework, companies streamlined their strategic business questions and wrote down that they were going to focus on answering just one question. This reduced the complexity of their projects and the amount of data needed, leading to speedier results.
For example, a large Japanese electronics manufacturer started using the SWAT framework to deliver the benefits of Actionable Intelligence to their supply chain. They wanted to become more responsive to demand changes and did so by bringing key bits of information: forecasts, inventory, supplier performance, end to end lead time, macro impact (such as commodities and currency).
They delivered results, but aren’t stopping. With proven monetary benefits they have deeper strategic questions and are answering them quickly and effectively.
The key lesson here is not to be lured into thinking that you need all the data on everything and massive projects in order to deliver results.
Instead, start small, cycle through the framework repeatedly, and deliver quick wins to show management that the Actionable Intelligence you extract is worth the investment because you can seize game changing opportunities with facts.