The Verghis Group

How do you mine sentiment on the web?

By Phil Verghis on August 24, 2009

Ah the simpler days of the Internet. Way back in 2002, during my Akamai days, we had team members manually mining public forums for customer comments during live events to ensure that we could pin point and resolve issues close to real time. Of course it would be impossible to scale staff enough to do that smartly today.

That’s why today’s New York Times article  (free registration required) on some of the tools available to mine blogs, Twitter and more caught my interest. Selected quotes from the article:

Scout Labs recently introduced a subscription service that allows customers to monitor blogs, news articles, online forums and social networking sites for trends in opinions about products, services or topics in the news.

Jodange offers a service geared toward online publishers that lets them incorporate opinion data drawn from over 450,000 sources, including mainstream news sources, blogs and Twitter.

Bo Pang, a researcher at Yahoo co-wrote “Opinion Mining and Sentiment Analysis,” one of the first academic books on sentiment analysis.

To get at the true intent of a statement, Ms. Pang developed software that looks at several different filters, including polarity (is the statement positive or negative?), intensity (what is the degree of emotion being expressed?) and subjectivity (how partial or impartial is the source?).

For example, a preponderance of adjectives often signals a high degree of subjectivity, while noun- and verb-heavy statements tend toward a more neutral point of view.

How are you tracking sentiments about your organization? Have you been able to strike a good balance between a rapid response and appropriate response (i.e. not over-reacting) ?

The Butterfly Effect – Part 2

By Phil Verghis on March 17, 2008

(I wrote this opinion piece for IBM, for whom I’m an independent thought leader. See more on the ‘Verghis View’ at IBM’s resource center.

In the last opinion piece, we talked about the ‘Butterfly Effect’ – how small changes early in the lifecycle of a complex system can have profound effects on the eventual outcome.

Here’s a real-world example in which a small error quickly ballooned into a major publicity fiasco, upsetting many customers.

In May 2007, Intel announced the long awaited ‘Santa Rosa’ upgrade to the Centrino platform for laptops. A major computer manufacturer (let’s call it BigBox) was one of the first major manufacturers to announce support for it. Orders flooded in.

The initial problem:
Enter the ‘Butterfly Effect.’ Due to a data-entry error, the bezel – the frame that surrounds the LCD panel – was omitted from the bill of materials on certain orders. This
caused a majority of the orders taken between May 9th and May 24th to be rejected during the manufacturing process. The missing-part problem was discovered, and BigBox
manufacturing began to fix and fulfill the rejected orders.

Unintended consequences:
Unfortunately, the fix caused a cascading series of errors. Re-entering orders caused automated emails to be sent to customers listing new shipping dates, often a month or so
later than their original estimated shipping date. The updated orders also showed increased prices, as there was apparently no way to honor the initial promotional prices in the system. Upset customers flooded the phone lines, often waiting over 40 minutes before they could get through to support. Those who did get through were told to check their order status on the Web. Unfortunately,BigBox’s order information page seemed to have only two statuses: ‘In Process’ and ‘Shipped’ – the latter rarely seen.

But that wasn’t all. There were also shortages of key components which caused more delays and more ‘shipment delayed’notices. The cycle of uncertainty for customers continued. BigBox’s post-sales team had no more information than what was on the Web, with no way to get an updated status from manufacturing. There seemed to have been a complete
breakdown of real-time information between post-sales, manufacturing and logistics.

Mechanism for updating:
With no way to get reliable or consistent information about what was going on, customers flocked to a popular blog authored by some BigBox marketing managers. Interestingly, the post-sales team at BigBoxeven started referring everyone to the blog for the latest updates. Over 1,000 comments were posted in the span of a month and a half — very few of them positive or supportive. Not the best publicity for any company to deal with.

You can imagine the howls of frustration when some customers, who had ordered identically configured systems, received their computers weeks ahead of those caught in the data-entry error.

Lessons learned:
The ‘Butterfly Effect’ is real. As you’ve seen, small errors can balloon out of control. Smash the barriers between departments, particularly when there is a crisis. In this
example, imagine how much better the crisis might have been handled if there had been open communication among manufacturing, shipping and customer support. Let
professionals in each area do what they are supposed to do.

In this case a marketing manager should not have been allowed to take on the burden of communicating during a customer service crisis.

Butterflies, Assets and … You?

By Phil Verghis on March 17, 2008

(I wrote this opinion piece for IBM, for whom I’m an independent thought leader. See more on the ‘Verghis View’ at IBM’s resource center.

Butterflies are not only beautiful, but they can teach us something about managing assets in a complex world. Here’s what I mean.

For years, we have heard that we need to be aligned to the business. If all that required was documenting the mapping between business services and IT components, this would have been successfully accomplished years ago.

With the increased pressure on productivity, there’s barely time to keep up with everything that has to be done in your own job, much less other departments or the organization itself. Who has time? It’s like brushing our teeth after every meal. We know we should, but how many of us actually do it?

Interestingly, a 1972 talk titled Predictability: Does the Flap of a Butterfly’s Wings in Brazil Set Off a Tornado in Texas? holds a clue about why alignment is so important to the field of asset management. Meteorologist Edward Lorenz’s groundbreaking (and then-radical) paper concluded that the slightest difference in initial conditions – even smaller than what we could measure – made it impossible to predict past or future outcomes. That idea went against conventional thinking of physics.

OK, but how does this affect you? Well, the very idea of an asset, and who manages it, has dramatically increased in importance over the last few years. After all, managing your assets, and knowing what they are used for, and why, can materially affect your company’s efficiency, operations, even its stock price. For example, a tiny screw that’s missing or damaged because it wasn’t maintained properly can cause a multi-million dollar backlog in a your supply chain. Multiple groups would love visibility into the maintenance of that screw, for different reasons.

That’s why smart managers around the world are taking the time to peek over the silos that separate them, and discovering how their work impacts other parts of the business and vice versa. Having a comprehensive view of your assets enables different groups to look at the same asset from very different perspectives. This provides a chance to pool everyone’s information and make informed choices.

A problem with that screw is like that tiny butterfly’s wings having an impact on the weather thousands of miles away. Understanding your organization’s assets helps you understand the complex ripple effects that can happen downstream – and turn them into a competitive advantage.

That’s why butterflies, in addition to being beautiful, have lots to teach us about managing assets in a complex world.

The Verghis Group | (800) 494 9142 | (617) 395 6613 | |
Copyright © 2010 The Verghis Group. All Rights Reserved.