Companies routinely amass much raw organizational data; amid this information abyss are solutions that could solve the firm’s problems. Data experts from Deloitte explain how businesses can unlock insights from such data that could spark changes in the company for the better.

By Nan-Hie In

Information is power. Harness it to solve business problems and such revelations can transform a company. The potential benefits vary, such as improved efficiency that translates to millions in cost savings. Robert Rady-Pentek, an associate director at Deloitte Consulting, recalls a success story of one of his former clients, a gas canisters company in the US, which achieved those outcomes exactly courtesy of data analytics.

Along with Jez Heath, a partner at Deloitte Consulting, the two veteran data experts gave a talk at AmCham on how solutions can be extracted from company data to address business dilemmas that would prompt changes in the organization.

Too Early, Too Late

Robert Rady-Pentek
Robert Rady-Pentek

The gas canister distributor had a supply chain problem. To purchase gas tanks – used to fuel barbeque grills – customers had to visit a store or gas station where the products were stocked in metal cages. “The clients had [tens of thousands of] locations across the US for these cages and from a distribution perspective, it was a challenge to deliver the product to these locations all the time,” says Rady-Pentek.

If the delivery truck arrived too early, as evidenced by the lack of gas canister sales, it was a futile trop as the cage did not need product replacements. If the delivery truck arrived too late, the cages were out of stock which posed risks, such as making it difficult to assess whether the stockout occurred the night before or weeks prior.

“The problem was that sales of gas tanks in these cages are highly volatile, driven by factors such as weather, holidays, the type of store that housed these cages, and more,” he says.

The company needed a better way to forecast sales to deliver gas canisters to these locations at the right time.

Enter a team of data specialists.

Their initial task was to grasp key drivers behind product sales in various locations. They dug up internal and external data (including weather forecasts) and qualitative and quantitative information. Many revelations arose – for instance, holidays were a key driver to gas tank sales.

Essentially a forecasting equation was calculated based on the data. “When this equation was put in their system, it improved their forecasting accuracy by about four percent, which significantly reduced stockouts and early deliveries; the company saved several million dollars a year in delivery cost,” explains Rady-Pentek.

Both Rad-Pentek and Heath recounted various case studies that illustrate the successful implementation of solutions generated from data. But this approach cannot fix every company quandary.

Besides forecasting, there are other ways to use and apply data analytics to achieve success in one’s organization. The duo outlined some guidelines to help companies get started in this process.

Analytics Defined

Jez HealthJez Heath

There are many definitions under the umbrella term of ‘analytics.’ Descriptive analytics is gathering a myriad of information that has already occurred, such as recent sales transactions to generate reports. Descriptive analytics reveal a snapshot of the company but it has limited forecasting ability as it is difficult to make informed decisions about the company’s future based on what just happened, says Heath.

Diagnostic analytics use information to extrapolate trends or other insights to better understand the data. It explains why the results have occurred. “It’s bringing more meaning to the numbers that you have captured and presented,” explains Heath.

Predictive analytics on the other hand, is the key technique in generating insights that could potentially transform a company. Heath defines it as using predictive and prescriptive statistical-type modeling that takes historical data based on a key set of conditions that have happened before, to determine a defined event that is going to happen in future.

“This is the essence of analytics: to take action to address a situation thanks to insight generated from data analytics,” he says. The gas canister distributor’s journey is a case in point.

The Fundamental Steps

According to Heath, many companies harbor much unprocessed data. However, hiring data scientists to scour through the data to uncover insights for the company to take action is a mistaken approach, he says.

“The whole point of doing all this is to solve your business issues. Ask yourself: what is preventing you from achieving your business strategy? Start from [the business issue] then work backwards,” advises Heath.

Once the most glaring business issue has been identified, investigate what actions and pieces of information are needed to address the problem. Then gather and analyze the data, the most time-consuming aspect of the entire process, according to Rady- Pentek.

One reason many companies do not embark on this process is the misconception that their data quality is not good enough for analysis. However, Rady-Pentek says the data does not have to be 100 percent in quality to apply analysis.

“Every company has an enormous amount of data. If you’re waiting for the time for all your data to be cleaned, you will never get there.” He recommends starting on a small-scale analytics project to begin to have clarity around which data needs to be cleaned.

He suggests embarking on a “proof of concept” process, which is conducting a scaled-down version of a full-scale analytics work. For example, instead of performing a data analysis proficiently throughout the company across its outposts worldwide, focus on one department in one city.

“If it works, this proof of concept suddenly shows that the data is important and you know what kind of data needs to be cleaned and what type of tools are needed.” The data expert says once the company stakeholders learn the value of the data, it can be institutionalized across the company.

Interpreting the Data

In the process, the company will be confronted with an avalanche of data, which can be overwhelming. How can key insights be sifted through the wealth of information and be presented in an understandable way? Rady-Pentek recommends visualization tools to extract pertinent insights from the data to help business users make decisions.

That includes geospatial analysis tools, software that lets one process and visualize data in the format of maps. Patterns can be revealed through this tool as data is sorted by location and its relationship with your chosen parameters.

For instance, if a retailer needs information to decide where to place its new store, this tool can help. The geospatial data is showcased in a map with all the company’s stores plus indicators presented in various colors or styles that reflect the data sorted through your chosen parameter, such as sales transactions. This tool can help viewers visually see patterns such as areas where stores generate most customer traffic or sales, information that is useful for retail executives to decide where to place its next store.

Organizational Considerations

Finally, integrate the solution into business practice. Rady-Pentek says all the key players in the process – from the data specialists team to the company stakeholders – need to work in collaboration to enact these insights into daily business operations.

From the data analytics perspective, the teams need to understand the process of how analytics works for it to be done successfully. A range of specialized skills are required including various specialists with technology, statistics, data science and testing skills.

Additionally, the business people need to be on board too. “You cannot get anything out of these insights unless you can bring it back to the business; [the process starts with] the business and understanding the business problem,” he says. Various stakeholders, including the groups that help communicate to others the insights culled from the data, all play an essential role in the overall scheme of things.

Most of all, one must stand firm behind the revelations generated from the data.

“The whole notion of analytics is to bring insight to the business and sometimes it might counter existing thought,” Rady-Pentek elaborates. “If you believe in your data and analytics, then stand firm. This is why your company is doing analytics, to bring new ways of thinking to the company.”