In part 1 of this article series, tED magazine discussed why those emails, complaint calls, Word files, and social media feedback should be factored into your electrical distributor’s overall data management approach.
In part 2, we’re taking a look at specific ways your distributorship can leverage its unstructured data.
When Kavita Ganesan, Ph.D., founder of Opinosis Analytics, ponders how companies overlook the value that lies in their unstructured data, a high-profile example comes to mind: the Samsung Galaxy Note 7 and its unstable battery. The phone’s irregularly-sized batteries led to the units overheating and in some cases catching fire and/or exploding.
Ganesan says what turned into a $5 billion problem for Samsung could probably have been avoided—or at least minimized—by analyzing the manufacturer’s treasure trove of unstructured data. “People were calling in and saying, ‘Hey, my phone is overheating’ and writing online reviews that said, ‘My Note is really, really hot,’” Ganesan points out.
“Had the product recall happened as soon as that feedback started to come in, there probably would have been fewer accidents,” she continues. “Samsung could have avoided a lot of these issues just by mining and making sense of its unstructured data.”
Here are five ways electrical distributors can get up to speed with their own unstructured data strategies and start leveraging this type of data to their advantage:
- Kick off your strategy by looking at the data that you’re collecting. Is it coming in through your inside sales team? Is it being gathered by phone? Is it mostly email based? “Wherever you’re collecting the most data will be your sources of data advantage,” says Ganesan, who points to email as a good starting point. Customer complaints that come in via email are a veritable goldmine for distributors that want to pinpoint and fix problems, for example. However, many times these communications are handled on a one-off basis and not “linked” to other complaints or comments.
- Initiate a paperless strategy. Any company that’s still working with paper files and notes scribbled on Post-Its should aim for a more paperless approach. Not only will this help streamline the communication process, but it will also allow others to access, retrieve, and use those notes. “In most small businesses, a secretary still jots down a note and then delivers it and/or addresses the issue—then the note is thrown away,” says Ganesan. “Instead of writing things down, have your employees document those notes electronically (i.e., in a customer relationship management [CRM] program, spreadsheet, or a cloud-based team management platform like Basecamp). “Over time,” she adds, “this will turn into a real data powerhouse for your distributorship.”
- Choose an analytics tool. No one wants to sit around all day perusing numbers and looking for patterns in that data, but thanks to data analytics and artificial intelligence, much of that task can now be automated. If your company has just a single data source to analyze (i.e., social media postings for marketing campaign metrics), then choose Web harvesting or social media analytics and sentiment analysis. However, if you want to mine information from a broader set of text-based data, choose tools that analyze across a variety of textual formats. “Whatever tool you choose should clearly tabulate and visualize results for you, and reporting should work on computers, mobile devices, and browser-based clients,” Datamation points out. Here are a few options that it suggests:
- Application-specific analytics: When a single unstructured application contains rich value, look at analytics tools that work specifically for that application such as analysis tools for Salesforce applications.
- Text analysis: This is a large category containing data mining, textual analytics such as metadata or part-of-speech tagging, and natural language processing (NLP). Algorithms use document relevance-based analytics to search a variety of textual data types.
- Web harvesting: These tools search for relevant structured and unstructured data from the Web. Technology connects data with user-generated patterns and filters to harvest related data.
- Hadoop: This is an ecosystem that spawns multiple vendors and products. It’s based on clusters of commodity servers with massively parallel processing that supports structured and unstructured analytics.
- Business intelligence software (BI): BI is a category of analytics that works across both structured and unstructured data. It employs data mining, reporting, and dashboards that present data in the context of informed business decisions.
- Aim for a “complete view” that’s focused on cost-to-serve. Don’t just slice-and-dice the unstructured data in an effort to get specific answers (i.e., how many electrical contractors complained about this specific drill bit during this period of time?). Instead, strive for a holistic view of your company’s unstructured data that incorporates email, social media, text files, videos, and other pieces of content. “Not only will this help you get a complete view of customer purchase patterns, behaviors, and transaction preferences,” says Earl van As, VP of marketing at Vancouver-based Conexiom, “but it will also help you build out very comprehensive customer profile and inform your distributorship’s overall sales strategy.” Additionally, a holistic view will provide some line of sight into the cost to serve your customers—insights that are largely rooted in unstructured data (especially in email side). “Many times, pricing data ends up being attachments or orders that a customer service rep has to enter by hand,” van As points out, “and that’s not something that’s typically factored into a cost to serve analysis.”
- Looking for a place to start? Email is a good first bet. van As says distributors should take one step backward, assess the technology that they’re using, and figure out new ways to gather and analyze the unstructured data that lies in those systems. Estimating that industrial distributors get about 70% of their orders via email, he says companies should put some time and effort into organizing those communications in a way that allows for deeper analysis. Conexiom, for example, will take a batch of emails from a specific date range (say 3-6 months) and run that batch through a data assessment engine that evaluates the emails, filter out purchase orders (where most of the price and order data resides), and then group them accordingly (by customer, SKU, payment discounts, etc.) for further analysis. “By reviewing both the unstructured and traditional structured data sources, companies can not only develop a much better view into customer experiences,” says van As, “but also refine and optimize their sales strategies.”