Perficient Enterprise Information Solutions Blog

Blog Categories


Posts by this author: RSS

Listening Deeply: Using Text Mining on Social Media Data

Social media means big business for a lot of brands.  Many companies jump into social media outlets with great gusto and litter the place with ethusiastic diatribes about how great they are.  They release tweets with no greater value proposition than “try our great product now”, and then they wonder why their social media efforts are not returning an impressive ROI.

What these brands fail to realize is that in social media listening and engaging are the most potent ingredients.   Social media begs brands to be part of the conversation around their brands and around those topics that are the most important to their target market.

 This concept was broken down quite well in “The 5 W’s of Social Media Listening“, which highlights the questions an organization must know so that they can listen and engage appropriately.  They are: 

  • Who is talking about you or your products?
  • What are your stakeholders saying about you?
  • Where are the conversations happening about you and your products?
  • When are they talking about you?
  • Why are they talking about you?

 Sound easy enough?  Not so much.  Let’s take some random brands and gauge their recent impressions via Twitter:

  • Hershey: 21,494 impressions
  • Sony: 29,853 impressions
  •  Ford: 34,907 impressions
  • Whole Foods: 87,902 impressions
  • Verizon: 115,765 impressions

When you add in impressions gained from the likes of Facebook, LinkedIn, etc., it is easy to see that no mere mortal could be responsible for deeply listening and engaging this mass of impressions.  However, they can do so quite easily with help from analytics.

The Art of Listening Deeply

The power of analyzing big masses of data can kick a company’s social media ROI into overdrive.  This is now possible through business intelligence tools that dive into big data through text analytics.   Text analytics is the art of deriving relevant information from unstructured text.  In the healthcare industry text analytics is used to diagnose disease .  In the music industry, text analytics are being use to find the next hot artist among social media buzz. 

If this technology can be used as physician’s assistant and talent scout, then  just think about what it can do for your brand.  End users can view social data through a dashboard that tracks intelligence in near real-time along with key social media metrics.  This social media data can also be integrated with existing marketing data to improve marketing efforts, target specific markets, or improve customer service efforts.

Customer-Centric Business Intelligence: Predicting the Future

Faster analysis and decision making can create a powerful strategic advantage.  Enterprise Performance Management (EPM) can be used as a powerful diagnostic tool through the use of automated scorecards and dashboards.  A company can identify potential flaws and short-comings in the cause-and-effect relationships that drive company performance.  Reaction times can be cut from hours and days down to minutes.  As the balanced scorecard is used and feedback is received, a company can begin to test the hypothesis of their original strategy through formal testing, such as statistical regression analysis.  EPM, through Business Intelligence reporting tools, provides a powerful platform for this activity.

EPM can also predict future behavior by linking past financial performance to a simulated future performance. Essentially, managers can view the company’s pro forma income statement by customer segment and profit center to evaluate future trends. These pro-forma income statements can calculate sales per customer, cost of goods sold as a percent of sales, acquisition costs per customer, marketing and operations costs per customer, and projected changes in these measures over time.  Past financial information can be used as a “base case” to validate the information provided in the pro forma statements. 

The Ever-Evolving Customer Relationship

This base case can then be used to compare against projected marketing investment allocations to determine worthwhile future investment or necessary changes in strategy.  Time trend analysis should show customer attrition levels decreasing among the most profitable customers and movement of less profitable customers to more profitable products or less costly distribution channels, which will further drive profits to the bottom line.  Valuable information gleaned from this analysis can be incorporated into the next Strategize process in the closed loop EPM system.

Successful companies realize that strategy is an effective way to grow with market conditions.  Strategy is most effective when it can evolve with the competitive landscape, economic conditions, changing customer data, and correlations found in every business environment.  Most of these same companies utilize the balanced scorecard methodology to meet these objectives.  Overall, EPM provides the balanced scorecard approach with a nimbleness that it does not have on its own merit.  Traditionally, users of the balanced scorecard approach revise their strategy map on a yearly, or perhaps quarterly, basis, which may or may not be adequate given business conditions.  Using EPM, the strategy map can be tested, supplemented, and cascaded down throughout the company efficiently to drive firm value through important customer relationships.  Using this Business Intelligence enabled strategy platform to communicate and evaluate employee performance will provide a powerful catalyst for company growth.

Customer-Centric Business Intelligence: Viewing Customers through the Dashboard

The customer-centric scorecard provides managers with the vital customer data that drives performance.  Using the scorecard and dashboards, managers are not only able to watch the long term trends in customer data, but they can also be alerted when shifts and changes in the customer base require action.  Enterprise Performance Management (EPM) creates additional value to the balanced scorecard approach by contributing reporting, alerting, and analytic tools that can help a company monitor performance against strategy.  Executives can drill down into historic and real-time data to analyze performance.

The Business Intelligence dashboard contains gauges that measure and display specific parameters around customer data in a timely manner.  Traditionally, managers have spent valuable time wading through numerous reports trying to decipher the correlation between the data as presented and the goals by which their performance is measured.  Using EPM, companies can create these dashboards based on strategy, which provide a view towards progress on objectives under their direct control.  The dashboard also allows users to drill down to uncover the data that underlie the visual scorecard dashboard display.  Automated scorecards can also be used to drill down into the data that underlie the strategy map.

Customer-Centric Dashboard Requirements

Smart organizations can build an information system used to sense and respond in their business environment using dashboard technology.  Four of the key requirements of a customer-centric dashboard program are:

1) Provides graphical, real-time views of key performance indicators to assess the health of customer loyalty

2) Integrates alert management when key measures fall below the target performance level and variance setting established for the particular metric

3) Provide staff with a way to drill down into the data that underlie each key performance indicator to locate the source of potential problems

4) Issues alerts to individuals responsible for the duties underlying each key performance indicator to initiate damage control.

Customer-Centric Business Intelligence: Capturing Customer Data

In order to create a customer-centric organization that capitalizes on the power of Business Intelligence, companies must identify the goals they would like to achieve and incorporate them into a formal business strategy.  The strategy map found in the balanced scorecard approach is a widely-used method for achieving this challenging task.

Ultimately, understanding your customers is key to building strategy.  In order to drive firm value, companies must track customer performance based on a single view of each customer as opposed to reviewing a random collective of siloed data points across the organization.  Using Business Intelligence tools, customer groups can be segmented and compared with regard to their ability to drive value. When attempting to acquire new customers, a company must identify prospects that will generate positive customer lifetime values that exceed acquisition costs.  Likewise, maintaining relationships with current customers translates into identifying those customers whose predicted future value is higher than the investment necessary to maintain that relationship.  Doing otherwise would essentially translate into paying your customers for doing business with you.  This seems so wrong, yet so many companies do it.  Why?  Because they are not tracking customer data correctly.

Tracking this level of detail in your traditional “spreadsheet mart” is a challenge.  Tracking this data using Business Intelligence tools will allow you to truly compete with analytics.  Ultimately, defining customers in this way provides a company with a valuable financial tool.  When an organization fulfills unmet needs in a growing market of profitable customers, then financial returns, and increases in market share, will result. 

Finding Your Most Profitable Customers

At the most basic level, customer transactional data can be used to identify profitable customers via the whale curve (read: that graph that tells you 80% of your profit comes from 20% of your customers).  However, a customer-centric Business Intelligence scorecard takes this important practice a step further by creating a company culture that revolves around tracking and growing these vital relationships with profitable customers.  In order to develop a customer-centric scorecard a company must have the ability to access and give meaning to high quality data.  This is done by:

  1. Assessing the size and quality of the current customer mix 
  2. Determining if more high value customers are entering the customer mix as compared to prior periods
  3. Using this data to determine any baseline trends with regard to customer mix, size, and quality over time. 

Using this actionable data, companies can make more informed decisions with their valuable resources.  Many companies sacrifice investment in existing customers in order to generate new business.  However, new customer acquisition is notoriously expensive.  It costs six times more to attract a new customer than to maintain and expand a relationship with an existing customer.  Many companies, such as Kroger, Panera, and Shell are increasing firm value by increasing loyalty among existing customers instead of expending resources on developing new customer relationships.  They realized this value through Business Intelligence.

Predictive modeling is one of the main tools companies can use to identify customers with the most profit potential and customers that are most likely to end their relationship with the company in the near future.  This is done by comparing current customer data patterns with past customer behavior.  This comprehensive understanding of customer behavior can be integrated with the supply chain to determine its full impact on the organization well before it becomes a reality.  A company can route shipments and modify processes around potential problems, as opposed to scrambling to deal with a problem once it works its way into the supply chain.  Additionally, prices can be established in real time using predictive modeling so that a company can yield the most profit from each customer transaction.

Using analytics on clean customer data is a pathway towards success.  Stay tuned to future posts in this series where we will discuss viewing customer trends using Business Intelligence dashboards and adjusting future strategy based on information gleaned from patterns in customer data.

Customer-Centric Business Intelligence

In this competitive business marketplace, organizations face increased pressure to forecast business conditions that allow employees to react and drive value from data.  How a company evaluates employee performance has a strong impact on behavior.  Even a company replete with intrinsically motivated employees can fail to be successful if management does not provide employees with guidance on how their efforts impact the bottom line.  Reliable and actionable data that represent a “single version of the truth” should be the foundation upon which a performance program is built and cascaded down throughout the organization. 

As common sense would dictate, companies should evaluate performance based on increases to firm value. To generate these cash flows, many companies focus on product development.  However, the firm’s value can be more specifically defined as the present value of cash flows generated by customers.  This makes gaining and maintaining relationships with profitable customers the single most important factor in generating firm value.  This relationship between customers and firm value is the fuel behind the competition for customer loyalty within industries.

Enterprise Performance Management (EPM), a Business Intelligence-enabled approach to the balanced scorecard, can create strategic advantage in a competitive business landscape.  Using EPM a company can compete for loyalty using strong analytics.  This is done through creating and managing a customer-centric electronic scorecard that is based upon a company’s important relationship with its most profitable customers.  EPM is used to help create strategic goals through the discovery of key objectives and performance measures based on analytical tools such as predictive modeling, customer relationship management, dashboards, and automated scorecards.  EPM also helps companies manage performance goals by linking those goals to the individual employees responsible for outcomes.

EPM consists of four processes, shown below, that can be interlaced with the balanced scorecard approach that most organizations use to refine organizational focus and create value.  The two initial EPM steps, namely Strategize and Plan, closely parallel the strategy map creation process found in the balanced scorecard approach.  The last two steps, Monitor and Act & Adjust, can be used to create additional value beyond what is provided in the balanced scorecard methodology.  The four EPM processes together form a closed-loop business strategy system that connects strategy to execution. 

Over the coming weeks, we will drill down into the processes needed to build a customer-centric balanced scorecard that will build value for your organization through the use of Business Intelligence.

Eliminating Dirty Data from your Sustainability Report

Today is the last day of the July-June fiscal year, which means many organizations, such as Proctor & Gamble, Coca-Cola, and Ford Motor Company, are winding down the period covered under their annual Sustainability Report.  There are many reasons that these organizations choose to release sustainability reports.  Many believe these reports lend an air of accountability and transparency that engages investors and consumers.  Others feel this annual report helps them drive process efficiencies that lead to tremendous cost savings.  After all, resources cost money.  When you use fewer resources you spend less money. 

While all of this is true, many don’t often realize that technology is a key contributor to sustainability efforts.  IT sustainability involves incorporating energy-efficient technologies and practices into everyday business tasks.  Business Intelligence can be a key contributor to a company’s sustainability efforts.  When it comes to sustainability reporting, business intelligence has two key functions:

Provides useful, trusted, and timely information for sustainability initiatives

To make the right decisions for sustainability initiatives, executives need useful, trusted and timely information.  Business intelligence provides a holistic view of all company assets and can accurately track utilization rates and forecast resource requirements.  This allows for a reduction in resource usage through “rightsizing”.  Business Intelligence technology can also create and maintain scorecards and dashboards that monitor essential sustainability metrics across the organization.  Executives can then drill down into company data to uncover the source of a problem before it gets out of control.   By including green metrics, such as power consumption and carbon footprint, executives can track and tune their eco-performance.

Reduces waste by optimizing data

Many forget that technology can deliver the Triple Bottom Line of profit, human capital, and natural resources on its own merit.  By using Business Intelligence technologies, companies can reduce waste by optimizing data through “data deduplication”.  This is a data quality initiative that reduces the number of times a piece of information exists within the system.  Duplicate data can surface in the form of different spellings or versions of a single customer’s name, for example. 

Optimizing data across an organization not only reduces the overall size of the database  but also reduces the processing requirements (i.e., energy used to keep the system going) as a result.  Organizations also need to consider the cascade effect this efficiency has.  By operating your organization under a “single version of the truth” you do not run the risk of sending multiple mailers out to a single customer or ordering the same supplies from multiple vendors.

Overall, there are a myriad of wasteful practices that can be made more efficient with the use of business intelligence.  How about adding the elimination of dirty data to next year’s report?