close

Big Data for Beginners

Why having access to  large data sets and tools to analyze them is an absolute must

Many will be familiar with the term “Big Data”. It’s the tag given to describe a large amount of data gathered from a variety of sources that are dynamic and fast moving. The data that a company like Google or Amazon hold would be considered big data. Previously considered a resource only available to large corporations with expensive technology to hold the information and a large staff to analyze it, due to technological advances and dropping prices, big data is now available to small and medium-sized businesses.

Driven by the growth of online transactions, companies now have the ability to capture far more information about their customers than they could in a brick and mortar storefront. An in-store purchase may gain a business inventory information, such as sales volumes, at what price point the product sells, and volumes of sales per geographic location. If the customer has a loyalty card, the potential to collect individual customer information exists, depending on what information is stored as part of the loyalty program. An online transaction collects all of those data as well as the ability to see what else a customer looked at, how long they spent on the site, how long they spent on each page, and how spending is affected by different promotions.

Having access to these large datasets and the tools to analyze them is an absolute must when looking for an advantage in a competitive industry. This belief in the strength of data is seen in the newer school of thought of Evidence-Based Management which, like the Evidence-Based Medicine or Evidence-Based Policy it is based on, uses the best available evidence to make decisions rather than subjective feelings of what is the “best” decision. Having large datasets at your disposal to query will help give the best evidence. Often it is better for smaller businesses to use big data than larger ones for Evidence-Based Management because their smaller size makes them more agile and able to adapt to needed changes.

For small business owners who have never used big data, or even any data at all, to examine their information and make use of what they have, there may be questions. Some of the most popular follow.

How does a small business begin using big data? Where do you find it?

Chances are, small businesses already have a fairly large amount of data collected, particularly if they have been in business for at least a year. Even if the business is older and had not begun in the digital age, and does not have many electronic records, the paper records still contain data. Sales slips, time cards, order forms, all of these have data worth analyzing. Perhaps the records are a mix of paper and electronic records. Maybe more recent inventory records are recorded in a spreadsheet, while the older information is kept in a hand-written ledger. It would be worth the business owner’s while to digitize the paper records. This will require an initial output of resources, but the time spent scanning images or entering data into a database program will be paid back in the time saved by the staff not having to dig through paper files looking for information in addition to gaining the ability to query these records.

Once all of the information is entered into a system that can be queried, data can be found from other sources that can be linked to the business’ records, often from free sources such as the US Government’s open data portal or Google Trends. In addition to these, there are many other publicly available large data sets to be found that can provide context and support to the data that has already been collected by the business. Population trends, weather, most popular baby names, etc. – all of these are data sets that are available for incorporation into a data set.

Joining the dots

Small business owners just beginning to use big data should start with small projects. Look for solutions to specific problems or answers to specific questions. If the business has a website, Google Analytics, offers free data analysis. IBM Watson Analytics offers a full-service range of big data analytics ranging from the price points of a free trial to monthly licensing fees at two different pricing levels. There are other analytic services and products available that may be better suited to specific businesses. All businesses who purchase a solution should look for a company that offers scalable licenses. If the business does not need the data analysis capabilities of a large enterprise, they should be sure that they are not paying for it. As the business grows and their needs increase, they can scale up.

Many small business owners may find it difficult to see why big data should take time from the day to day operations of the business. The idea of using data to analyze operations may sound like a good idea in theory, but there are only 24 hours in the day and those hours need to be dedicated to the running of the business. But wouldn’t those same small business owners like to know that 73% of their customers are men for a product that is not specifically marketed to them? Or that 84% of their online orders are shipped to only three States? Of course, by just doing their day to day work, they may have suspected those things were happening, but to have those figures in black and white will show what a huge number of potential customers they are not reaching or appealing to and work to figure out why.

As much as sales, marketing, or management, the ability to mine, analyze, and manage data is an essential tool in a business hoping to succeed and grow in any market. Business owners who refuse to use data analysis to grow their business run the risk of being left behind.

What’s next?

As the business grows, potentially adding more staff, product lines, locations, and complexity, owners would like to see what they can learn beyond the basics. The owners may be looking for a way to get more out of the data they hold. One way is to scale up on the licensing they may have with the vendor of their analytic solution. Or potentially, by trading from a free solution to a paid solution. The owner may find it worthwhile to have an in-house analytic resource and choose to hire an employee whose sole or partial responsibility is the analysis of the business’ data.

Sometimes, in order to understand more about a business, it is necessary to “look outside the box.” There may already be analysis that tells the business which of their sales people is earning the most revenue, or which product line may be selling less than expected, but by thinking outside the box and coming up with different ways of querying your data, they may discover that shortly after their latest ad campaign, an advertised product’s sales took a nose dive among women aged 35-39 who live in the Mid-West, but climbed in women aged 20-24 in the same region. That invites an exploration of why that advertisement was agreeable to one age group but not to another.

Start Small

As the business grows and staffing levels increase, data is essential for the performance measurement of the employees, allowing the business to see revenue, hours worked, revenue per hours worked, any trends, such as a habit of taking a sick day on sunny days. Maybe a strong performer in all areas has a product line that seems to be giving him or her trouble based on the lower level of revenue that is associated with it. This can allow the business to see that the employee is struggling and provide more assistance or training before it becomes a job performance issue. Big data can be used to analyze the internal operations of the business just as well as the external performance.

By starting small with big data and increasing the analytic capabilities as the business grows, a small business owner can ensure that his or her business is always making the best evidence-based decision for every stage in the business’ growth. And by making data analysis a day to day part of the operations of the business, they will be well-positioned to adapt to the changes in a fast-moving industry.

Tags : Apache HadoopBig DataBig Data Sparkfeatured
Tweet
Share
Share
+1
Email
WhatsApp