You are currently on IBM Systems Media’s archival website. Click here to view our new website.

POWER > Business Strategy > BI and Analytics

Is Big Data a Good Fit for Your Midsized B2B Company?


You’d have to live on a deserted island without electricity or Wi-Fi at this point to not have heard someone promoting big data and analytics as a critical way to set your company on the path to rapid growth and competitive advantage.

Most of the published examples of companies effectively using big data come from large retail and online consumer merchandisers, financial services companies, medical research, and military or government intelligence agencies. In those environments, you can find highly trained data scientists running sophisticated data mining algorithms over massive databases in search of heretofore-unknown patterns, associations, trends and correlations. However, if you’re an executive or IT professional in a midsized business-to-business distribution, manufacturing, logistics or industrial services environment, you may be skeptical about how big data and analytics relate to you.

A traditional transaction processing database can only tell you if and what someone ordered, but online behavioral data can be captured whether or not an order is placed.

It may be useful to think of big data as an evolutionary step from data warehousing. Data warehousing became practical when improvements in data collection and quality, as a result of the widespread adoption of ERP software, came together with more affordable servers, relational and multidimensional databases, and user-friendly query software. Companies designed and developed data marts and operational data stores primarily to make it easier to analyze their transaction processing data. Data marts and data warehouses continue to deliver significant value, but today we can capture and acquire far more than operational data, and we can increasingly afford the high performance servers, databases and data mining tools that can help us put it to good use.

A New Approach

The first hurdle to overcome before considering the potential of big data in your business is the term itself. Depending on your business, what you consider big data might not be all that big by today’s standards. A more meaningful term for many midsized companies might be extended data. By that I mean the introduction of new, frequently unstructured data such as video, Web click-stream logs, email messages, online forum postings, public records databases, call center logs, search engine statistics, traffic flow data, product performance data collected by sensors embedded in products, freight handling data collected via radio-frequency identification, and other sources that can provide a richer, more comprehensive view than our relational databases.

If you have an online catalog where customers can study your products, it could potentially be useful to know how often people click through to a particular product’s page, where on that page they position their cursor, and how long their cursor remains in different positions. A traditional transaction processing database can only tell you if and what someone ordered, but online behavioral data can be captured whether or not an order is placed.

If you know people are spending a lot of time studying a product, but not buying it, you might experiment with changes to your Web page graphics and product description to see if you can increase orders without changing the product or its price. While this kind of analysis can be thought of as a big data project, it’s important to realize that the volume of website usage data required to support this in a company selling a few thousand products to a few hundred customers doesn’t require a large investment in servers, storage or software. It does require savvy business people who are willing to try new approaches.

Let’s say you’re an aftermarket auto parts distributor selling to independent service centers in a major metropolitan area. Your reputation and growth depends on consistently supplying your customers with the parts they need before your competition. Your own historical data is valuable for forecasting customer needs and being prepared for demand. The customer information is matched with public data showing the registered automobiles in your service area by make, model and year, and then analyzed against statistics about the most common mechanical problems of those autos could provide valuable insight on what you should stock.

Supply chain companies are frequently one or more steps removed from the ultimate consumers of their products. Incomplete, anecdotal and slow reporting of product performance and reliability data from the field can mean defects and flawed designs remain in production—raising costs, damaging reputations, or in some cases, increasing liability risks. Today, with the cooperation of your trading partners, low-cost monitoring devices can be placed in products to feed real-time data to the manufacturer. Monitoring devices can be placed on trees and in fields to monitor moisture, fertilizer levels and temperature to help farmers manage water usage. Advertising companies can use traffic flow data to strategically position and price billboards.

Bill Langston is director of marketing for New Generation Software Inc., and a developer of iSeries BI software. He can be reached at Langston@ngsi.com.



Advertisement

Advertisement

2019 Solutions Edition

A Comprehensive Online Buyer's Guide to Solutions, Services and Education.

Picturing Success

Capture insight from real-time operational data with smart analytics systems

Unlocking the Secrets of Success

An introduction to business analytics

Layers of Integration

Components of a complete analytics solution

IBM Systems Magazine Subscribe Box Read Now Link Subscribe Now Link iPad App Google Play Store