As a business owner in the modern world, you've probably heard a lot about Big Data in recent years. Maybe you've even started using it to inform your business decisions. But because of the enormous volume of data being generated every day, it's difficult to know if you're really using it effectively.
"Every industry vertical today is opening up to the Big Data world," said Anil Kaul, the CEO of intelligent analytics company Absolutdata. "Small businesses ... have started leveraging a combination of in-house and third-party technologies for developing a 360-degree view of their customers using data coming from multiple sources. However, the main challenge ... is to determine which data to really focus on and how they can extract the real value from that data."
Vladik Rikhter, the CEO of mobile workforce and task management software company Zenput, said there's a lot of "noise" around Big Data, and businesses need to get to the core of how they can use that data to achieve their goals.
"Businesses should only be paying attention to one or two key metrics to get a good sense of their customer health," Rikhter told Business News Daily. "The rest of the data should be used to refine the approach."
But how do you decide which metrics to focus on? Data and business experts weighed in on the different types of Big Data you can analyze, and how to find and use the ones that matter most for your company. [Big Data 'Escapes the Lab': Tips for Small Businesses]
What does Big Data encompass?
If you're not sure where to begin with Big Data analytics, it's helpful to first study and understand the different types of data available to you. Kaul provided an outline of several categories of Big Data and their uses for small business:
Typically this type of information is in the form of human experiences, Kaul said. It is entirely digitized and stored everywhere from personal computers to social networks. Data are loosely structured and often ungoverned. Examples include:
- Social networks (Facebook, Twitter, Tumblr, etc.)
- Blogs and comments
- Pictures (Instagram, Flickr, Picasa, etc.)
- Videos (YouTube, Vimeo, Vine, etc.)
- Internet searches
- Mobile data content (text messages)
- User-generated maps
Traditional business systems
This type of data is highly structured and includes transactions, reference tables and relationships, as well as the metadata that sets its context. Kaul noted that this data is the vast majority of what IT manages and processes, and it is usually structured and stored in relational database systems. Examples include:
- Data produced by public agencies
- Medical records
- Commercial transactions (including e-commerce)
- Banking/stock records
Internet of Things
This machine-generated data is derived from devices and sensors used to measure and record the events and situations in the physical world. From simple sensor records to complex computer logs, this data is well-structured, Kaul said. Examples include:
- Data from sensors
- Fixed sensors
- Weather/pollution sensors
- Traffic sensors/webcam
- Security/surveillance videos/images
- Mobile sensors (tracking)
Which data do you really need?
Though you may be collecting and storing many of the above-named types of data, you don't have the time, resources or need to sift through each of them. Instead, you should figure out which ones are essential to informing your business decisions and only hone in on those data sets.
Sara Vera, a data scientist at CRM and project management software company Insightly, advised small businesses to focus on metrics that illuminate customer behavior.
"Use Big Data to get the most information about the customer base; who these customers are, what they like and dislike about the product and how they are using the product," Vera said. "This information can radiate outward in terms of continued growth and development of the product."
However, depending on the type of business you run, there may be another area that is more crucial for you to focus on. Charles Silver, the CEO of advanced analytics company Algebraix Data, listed five major concerns that most businesses have: revenue growth, profitability, customer management, operational efficiency and risk/fraud. Each of these big topics can be broken down into smaller areas where specific analytics can produce useful insights.
The first step is to decide which one of these five concerns is your business's top priority. Then, you can rank smaller related tasks in order of importance. For example, Silver said if operational efficiency is your top priority, your specific analytics can focus on areas like demand forecasting, labor scheduling or transportation optimization.
"It depends on the individual business and its current challenges," Silver said. "Owners of restaurant and retail operations will benefit from analytics that focus on 'customer segmentation' and 'menu/inventory optimization.' By contrast, a group of local insurance agencies or a midsize health care company might want to focus on 'detecting fraudulent claims.' And many businesses would profit from analytics that predict the lifetime value of a customer, so they can gauge retention efforts accordingly."
You may also want to look at analytics that will identify your top collaborators and initiatives, as well as the driving forces behind them, said Jeff Boehm, vice president of marketing at DataGravity, a provider of data-aware storage solutions.
"With more details about the efforts that are providing a high return on investment for your company, you can more easily repeat these situations in the future," Boehm said.
Using your key data
When you know what data sets you're going to look at, you then need to determine how to put it to work. Rikhter noted that Big Data analysis should always start with a question. What do you want to achieve in analyzing this data? Once the problem is identified, then you can focus on how the data will solve it, Rikhter said.
To begin the process of analyzing Big Data, Aaron Rallo, the CEO of IT efficiency software company TSO Logic, said businesses need an automated system or program that allows them to collect the data and convert it into direct actions.
"Without automation and intelligence you will be left with an overwhelming pile of data that does nothing but cost money to store," Rallo said. "Find a partner that has experience in solving the problems that you are trying to solve. Also, be sure that you trust the data and can rely on it when making decisions. If you, or your team members are questioning the source then it will produce more questions than answers."
Giles House, chief marketing officer of CallidusCloud, a providers of cloud-based sales, marketing and learning solutions, said seeking out the correlations between your data sets can help you determine what to do from there.
"There's value in a single, specific critical metric, but if you want real value look for correlations between data sets, like average deal size versus average quote size, or deal size compared to the amount of content downloaded," House said. "There are many more correlations to examine — pick the ones most vital to your success, and then use the insight they provide to take action."
An important thing that businesses should keep in mind — but often overlook — is the fact that, like any other data, your Big Data and any subsequent analysis on it is at risk of being hacked or stolen. Rallo said that in many cases, data that is stored is confidential and needs protection. The level of protection required is based on the sensitivity of the data.
"In some cases the data will need to be kept on premise and in other cases it can be trusted to third parties or stored in the cloud. But in all cases, security and data privacy should be considered," Rallo said.
Boehm agreed, and reminded business owners that internally generated data such as emails, notes, text documents and presentation decks are also part of your company's Big Data, and also needs to be managed properly and securely.
"If you don't know exactly where you're storing critical assets and private information, it's hard to stop them from becoming exposed," Boehm said. "You need to apply analytics and visualization to fully understand this information. Where is it? Who is accessing it? What type of information is it? How old is it? Sorting through this dark data can point out potential security risks, and help identify which files, folders or servers you might be paying to maintain, but are rarely using."
For just about every company, the ultimate goal of Big Data analytics is to make better business decisions that will lead to higher profits. Kaul said the key to monetizing your data is to look at the economic questions it can help answer.
"Often the data can help answer questions about the value, use, risk or future value or risk of a specific asset," Kaul said. "To derive value from Big Data, the data must be converted to a form or product that answers a fundamental market or asset question. Such data products can be sold or traded to clients. It can also be that giving away data products, derived from Big Data, will drive other related monetization strategies."
Lloyd Marino, CEO of Avetta Global, a technology strategy and applications development firm, said that following what he calls the three A's of Big Data — automation, analytics and action —will help you reach your intended purpose and boost your ROI.
"Automate the collection of your data, apply the analytics to create insight into your data and take action on the results continuing to enhance your algorithms, then rinse and repeat," Marino said.
Most importantly, Marino advised business leaders to create a plan for their data analysis to keep themselves and their teams on track.
"A small business can get caught up in Big Data," Marino said. "You risk getting buried under it unless you incorporate a plan to properly manage and leverage it. Without a solid plan aligned with your business objectives ... you may miss out on an elegant solution with a solid return on investment."