Saturday, May 1, 2010

The Intelligence of Social Media (Part 2)

In the first part of this blog, I mentioned that sentiment analysis measures the polarity of opinion—positive, negative, or neutral—regarding a subject, a product, a service, etc.

Two main approaches can be used to perform sentiment analysis or text mining: a knowledge-based approach, which uses linguistic models to classify sentiments; and a learning-based approach, which uses machine learning techniques to classify text. The concept of sentiment analysis opens a great number of possibilities and opportunities for introducing BI strategies to analyze the enormous amount of data flowing through the Web.

In fact, some software solutions have been designed to address this type of analysis. These tools are called “social Web analytics.” According to the definition provided by The Social Web Analytics eBook (2008), by Philip Sheldrake, social Web analytics are “the application of search, indexing, semantic analysis and business intelligence technologies to the task of identifying, tracking, listening to and participating in the distributed conversations about a particular brand, product or issue, with emphasis on quantifying the trend in each conversation’s sentiment and influence.”

Many organizations are aware of the importance of measuring this information and analyzing it. Currently, sentiment analysis has a strong potential to be used jointly with BI applications making it possible to apply traditional BI techniques to visualize what a sentiment-based tool has discovered on the Web. Some vendors are already offering analytics services (radian6, Sysomos, BuzzLogic, and Attentio) to measure and analyze social media content.

Now, there is also an existing trend regarding traditional BI providers to address these tasks:

Workforce Analytics – A Blend of Business Intelligence and Human Resources

If you are a HR manager in a company that employs thousands of people, one of your main concerns should be reporting and analytics. Workforce analytics can help your organization determine how efficient its recruiting processes are. It can also help during the hiring process—recruiting the right people at the right costs.

Workforce analytics can give your company a general overview of the activity of the HR department. Depending on the product being used, you can drill down to the next level of detail, build interactive graphs, and export the data to different file formats.

By understanding the demand and supply in HR—as well as the gaps between the two—HR professionals can create and implement better internal procedures for talent management, retention, succession, etc.

This is done by gathering information on your workforce and putting it into a single repository. This information comes from your HR system, enterprise resource planning (ERP) solution, or other business software (e.g. time and attendance, project management, accounting software, etc.).

By using this data, your company can create forecasts and what-if scenarios in order to understand how a change in the activity of the company can impact its HR department and vice-versa. For instance, if you decide to launch a new product line, you can estimate how many people you’ll need and how much it will cost you.

Most of the vendors in this area (HR, BI, or workforce analytics) offer pre-defined key performance indicators (KPIs) that your company can use to measure the efficiency of its workforce, but they can also help build new ones and even implement best practices to improve the way people work.

Quote-to-Order: An Overlooked Software Application

Last year, I met an analyst from another firm, and asked him what he thought about quote-to-order (Q2O) solutions, given the relevance between Q2O and the conference that I was attending. Not quite surprisingly, the answer I got was, “this kind of application doesn’t have a future.” The conversation didn’t go any further due to limited time but I could imagine that his reasoning might have sounded like this: even though activities from quoting to ordering may be taken care of by multiple systems, there’s no need to have another system (if there’s good integration in place), which makes the already complicated enterprise information landscape even more complicated. Certainly, this statement can be true if there is good integration in place. However, the truth is that today’s integration amongst various information systems is far from perfection. Let’s take a look at the reality of many companies’ Q2O process.

In a real working environment, sales professionals may have existing tools (independently or as a part of a customer relationship management [CRM] system) to support their quoting activities. These tools may be quite handy in generating a beautiful quotation document. However, what really counts in the quality of a quotation is the accuracy of the information provided to potential clients. More specifically, a good quotation has to present the correct product/service configuration based on a client’s requirements and what a company is willing to offer. Most of the time, product information is mainly produced by another group of people (usually called a product development department or something similar) using different systems. Given today’s fast-paced product development, relying on printed handbook, spreadsheet, or even batched update as the source of product information risks the inaccuracy in quoting.

Inaccurate product information is not the only problem in the Q2O process. Even though a quotation presents what is “technically perfect” (i.e., correct product configuration), it may not present what is commercially and operationally feasible to be delivered. In theory, production, purchasing, and inventory information should all play a role in generating a deliverable quotation. However, in practice, delivery terms are often determined based on experience or rule of thumb. As a matter of fact, in many organizations, only a few individuals know the so-called “tribal knowledge” of all the rules, constraints, etc., about what can or cannot be manufactured and delivered with what difficulties.

Disconnected data flow between quoting and ordering is another issue in the Q2O process. When quoting and ordering are handled by different groups of people, a finalized quotation often has to be re-entered (or in a better case, imported) to systems that control production and delivery. This non-value-added activity not only consumes resources but also opens the door for errors.

In the case that companies sell via distributors and resellers, the situation can only become more complicated.

Undoubtedly, integrating one system to another pair by pair (e.g., product lifecycle management [PLM] and CRM, CRM and enterprise resource planning [ERP], and PLM and ERP) is a method to address the above mentioned issues. However, there is a small group of software applications (titled Q2O or configure, price, quote [CPQ] solutions) taking a more focused approach (click here to see the list of Q2O solutions). Q2O solutions use the Q2O process as a main thread to integrate all relevant activities and needed information in one place. In addition, some solutions also provide functionality such as quotation documentation, product information management (PIM) (also called master data management [MDM] for product information), and e-commerce capabilities (e.g., shopping carts, checkouts, save for later, etc.) that are often not the case associated with ERP or CRM systems.