11. Case Studies

Most of this book so far has been discussion of tools—things that you can use to trap information better.

But you may be at the point where having the tools is all very well, but you need a more concrete sense of when and how you might want to use them. That’s what this chapter is all about. We look at two hypothetical case studies: one for finding external information, such as news, reports, discussions, and so on, and one for finding internal information, such as links and critical comments.

As you might imagine, setting up a thorough set of information traps can take a lot of hours and touch on as many as two or three hundred resources. In one ongoing research job I do, I monitor over 800 separate pages and 1,500 different RSS feeds. The case studies in this chapter touch on only the highlights, or mention resources that might require a certain amount of caution or query revision. If you find your own trap-setting involving far more resources than the ones in this chapter, good for you!

Let’s start with Fred, who needs to learn more about outsourcing to Taiwan.

Fred Learns about Outsourcing to Taiwan

Fred works in a company that makes computer parts. His boss recently asked him to investigate the feasibility of outsourcing some of his company’s manufacturing work to Taiwan. Fred realizes that what he knows about outsourcing to anywhere could fit in an ant’s lunchbox. So while he’s discussing with his boss and others in the company what their needs are and what they hope to accomplish, he sets up several information traps that will keep him up to date and inform him about outsourcing.

Step 1: Generating useful queries

Fred decides to go to a fresh information source—a news search engine—and begin experimenting with query words. He heads over to Google News (news.google.com) and begins generating queries. He discovers that a simple search for outsourcing Taiwan yields about 100 results—very manageable—but the first ten were published in the last couple of days, meaning he could expect five stories a day from this search alone. Fred thinks that may be a little much, and besides, the stories don’t actually fit what he’s looking for. He narrows his search a little using outsourcing Taiwan manufacturing.

Bingo! He’s halved his results, the first ten results were generated over a couple of weeks, and a cursory glance at the page summaries shows that these stories are exactly the sort of thing he’s looking for. So he saves that search. After some more experimenting, he also settles on the phrase "outsource to Taiwan", as well as the queries "DRAM production" Taiwan, and DRAM outsourcing.

Fred has generated two searches that address specifically what his boss has asked him about. But beyond that he has also generated a search that will give him ongoing information about the whole issue of outsourcing to Taiwan in the first place, as well as a couple of searches that remove Taiwan from the equation and look at the whole issue of computer manufacturing and outsourcing in general.

By monitoring these information traps, Fred will be able to both directly answer his boss’s question and perhaps make a few suggestions of his own! (Maybe Fred will discover that Taiwan is not the best place to outsource some of its manufacturing—maybe it’s India, China, someplace else, or nowhere at all.)

With these queries, Fred begins building his traps by starting at a news search engine.

Step 2: Using news search engines

Fred ponders how urgent his need is to get this information. He finally decides that he will have the Google News updates delivered to his cell phone, just to give him constant feedback about Taiwan and outsourcing in general. The rest of the searches he’ll set up as RSS feeds and specify a certain time every morning to go over what’s been delivered to him.

He stays at Google News, reruns the searches, and takes advantage of the option to send the results to his cell phone’s e-mail address. Then he goes to Kebberfegg (researchbuzz.org/tools/kebberfegg.com), chooses the News option, and one by one generates OPML files for his queries. He then imports these files into his newsreader of choice, which happens to be NewsGator, and deletes the Google News feeds since he’s already getting them by e-mail.

Fred now has the cornerstone of his information trapping—fresh information being delivered to him from two or three different angles. He will use the information that he gains from these searches to tweak his queries over time. Perhaps a new company pops up in Taiwan that does what his company is interested in—he may want to monitor them. Perhaps a new advocacy group arises—he’ll want to monitor its activity and news. Fred knows that he’s off to a good start, but he also knows that he may have to change his queries and direction over time.

Fred’s next stop is to see if anybody out there is talking about outsourcing in Taiwan besides his company.

Step 3: Doing group and conversation searches

Fred begins his search at Yahoo Groups. By doing a general search for outsourcing, he discovers 508 groups are available! He narrows the search a bit to outsourcing manufacturing. He finds a few interesting groups and reviews them. One of the groups looks so relevant to his interests that he subscribes so that he can read it in its entirety, noting that it has relatively low traffic (only a few posts per day). The other two groups are sometimes relevant, but not so much that he wants to read them all; they’re pretty busy. For those he runs a few searches based on his original queries, and saves the results as RSS feeds when that’s an option, and when it isn’t, he notes the URLs so that he can send them to a page monitor.

In other situations, Fred might have found that the groups in which he’s interested were private. In these cases, he would have had to subscribe using his Yahoo account. But he was lucky enough to find that all the groups in which he was interested were publicly available.

From there, Fred heads to Google Groups. (As you may remember from the earlier chapters, Google Groups is a combination of Usenet, a very large set of public discussion lists, and Google’s own discussion groups.) A search of Google Groups’ entire index for Taiwan outsourcing brings mixed results. On one hand, he’s finding some good results. On the other hand, he’s finding too many of them and too much spam. Fred decides to look for individual groups instead. If Fred had found a satisfying and relevant number of search results, he could have chosen to have search results for the entire index delivered to him as e-mail alerts.

One step forward...

Fred remembers that Google Groups offers far fewer lists than Yahoo Groups does, and risks a search on a single word: outsourcing. He finds just 150 results, and after some investigating picks five that he wants to monitor.

All the groups that he wants to monitor are extremely relevant to his interests and have a low number of members (fewer than 50 in each group) so he picks up the RSS feed for each group and adds it to NewsGator.

So far Fred has not deviated very much from the original queries that he picked up from his news searches. That’s because he has been searching mostly for groups, and not for conversations within those groups. Now Fred is moving to search online message boards (Web sites that contain online discussion forums), and may discover that his searches will have to be altered for conversational style.

Fred goes to Yuku (yuku.com), a search engine for online message boards. He cautiously starts his search using the word outsourcing but gets no results. He searches for Taiwan, but gets only a few results that don’t seem related to business. A search for business brings hundreds of results, all far too general.

Fred ponders. How would he pose the issue that he’s investigating in an online bulletin board? After a few moments he comes up with this question:

How do I begin outsourcing in my computer manufacturing business?

...two steps back?

From that question he developed a new query: "begin outsourcing" and tried that. This query, while admirably specific, provided no results. He gave up and moved on to a different search.

Not all resources will provide useful traps to all researchers. Fred tried his original queries and then generated some new ones that seemed more appropriate to his research. While his results were better—a key indicator that his query revision was working—they were still not useful to his search. At this point, he could decide to let things lie and move on to the next resource, or continue to revise his query. In this case, he decided to move on.

Fred has a couple of choices. He can either try to search Google or Yahoo for more forums, using his query words and words that will be more likely to find forums and mailing lists (words like forums, posting, thread, earlier, and later). Depending on how much time he has to set up his traps and how much time he has committed to reviewing them, he could decide to stick with the “big two”—Yahoo Groups and Google Groups—or try to find more forums related to his interests.

At this point, Fred has a flow of news stories related to his interest and a somewhat more limited flow of discussion groups talking about his interest. His next task is to find Web sites dedicated to the topic in which he’s interested and set up page monitors.

Step 4: Monitoring pages

Fred knows that most of the news sources he’s monitoring are general. His next task is to find Web sites that are more oriented toward outsourcing specifically and see if he can integrate any of them into his information traps.

He starts with the Yahoo Directory (dir.yahoo.com). Knowing that Yahoo’s directory is not a list of sites and is more of a subject index, he starts again with a single-word search query—outsourcing. He finds the relevant category and looks at the list of available sites. He finds half-a-dozen that seem to have potential and spends some time reviewing them. In some cases, he decides to monitor the front page. In other cases, he decides to monitor a different page on the site. And in one case, he decides to monitor three different pages on one site—the front page, the news release page, and the page of a journal relevant to outsourcing that is hosted on the site.

From there he moves to the Open Directory (dmoz.org) and tries the same thing, finding a few more useful pages. He doesn’t feel, however, that he’s getting full coverage of the outsourcing industry, especially as it relates to Taiwan. So he goes to Google and tries a specific search based on the search that failed when searching forums: "begin outsourcing" advice. He gets 371 results and is able to narrow that down to 10 pages to monitor what he did not find through the directory. He’s now more comfortable that he’s getting complete coverage.

Many information trappers might stop here, believing they’re well on their way to becoming informed. In Fred’s case, however, there’s one more category of resource that he wants to cover.

Step 5: Monitoring stocks

In his research into outsourcing, Fred sees some publicly owned companies whose names are coming up over and over again. These companies have made decisions about outsourcing to Asia—some of them are including it in their strategies, some have pledged never to do it, and so on. Because Fred is interested in outsourcing as a possible direction for his company, he’s interested in the fortunes of these companies.

Fred wants to monitor a specific type of data that’s found on the Internet. In this case, it’s stock performance. Perhaps in your case it will be something else. Fred probably has done various kinds of stock monitoring before. Perhaps his company even subscribes to an information service. If it does not, Fred could go to Yahoo Finance, for example (finance.yahoo.com) and either set alerts for certain companies or subscribe to their RSS feeds. In any case, such a specific type of information doesn’t have to be part of the information trapping process, but it’s relevant to Fred and his strategy.

Fred’s future strategy

At the conclusion of his work, Fred has 75 or so feeds in his reader, several e-mail alerts set up, and a hundred or so pages in a page monitor. He has made his boss aware of his traps and will be spending an hour or so in the morning reviewing them before proceeding with the rest of the day’s work. Fred has a base of constantly updating data to sift through as he builds his report about outsourcing in Taiwan. How long he uses this information depends on when his report is due, whether or not his boss wants him to treat his report as an organic, ever-changing document, the directions his research takes him, and so on.

In the case of Fred, most of the information he’s looking for is external—it’s not directly related to his company. In Ethel’s case, however, there’s a need for tracking more internal information.

Ethel Tracks a Blog

Ethel has a blog that she wants to use to make herself better known in the community. She understands that she’ll need to create good content, make sure she’s publishing an RSS feed, and submit her site to search engines. Beyond that she wants to see who is linking to her site and what is being said about her site.

Unlike Fred, who is monitoring information for a report that will one day end (ostensibly), Ethel is setting up her information traps to be monitored on a more-or-less perpetual basis. And unlike Fred, Ethel will be looking at resources like blogs, so she will probably have more luck with discussion groups, and may not need to monitor news at all. However, like Fred, her first step will be to figure out the backlinks and keywords for which she should be monitoring.

Step 1: Determining backlinks and keywords

Ethel’s blog is a consumer advocacy site called “ICalledandAsked.com.” That’s both the name and URL of the site. The name is unique enough that she’ll be able to use it in her trapping. If she had gone with a different name—"I Don't Know"—she might have had to confine herself to monitoring only for her blog’s URL. If her name is uncommon enough, she might be able to monitor for that as well. "Ethel Mae Potter" is probably unique enough.

Ethel has three main things she can monitor: the name of her site, the URL of her site, and her own name. Is there anything else she should add? It depends on what’s on her site. If she has a tool that other sites might link to—a dictionary, a featured article, a database of information—she could track that URL as well. If she doesn’t, it’s probably best to monitor just the URL of the home page.

Now that Ethel knows what she’s looking for, she has to figure out where to look for it.

Step 2: Choosing resources to search

Fred, seeking external information, had a good sense of where that information was located. He was looking for credible information, data, and experiences. It was easy for him to determine where to look.

Ethel, on the other hand, is seeking internally focused information—links to her blog, mentions of its name, and mentions of her name. Where she looks is based on information about herself. Is she famous? Is she controversial? Is she known from some other arena?

It’s entirely possible that for the purposes of monitoring information, Ethel is neutral—neither known nor unknown, controversial or not. That’s the case we’re going to look at here.

Step 3: Monitoring blogs

Because Ethel is not particularly famous or infamous, she decides to start her monitoring with blogs. Most blog readers read other blogs, so blog search engines are a good place to track mentions.

Ethel starts with Feedster (feedster.com), and begins the process of tracking her name, her URL, and her site’s name. She decides to go with all RSS feed alerts in her blog searches—she’s not interested in getting any notifications via e-mail. After she’s set up these feeds, she wonders if she should be monitoring blogs for mention of her site’s theme—consumer advocacy. After all, if she finds relevant blogs, she could ask them for links or begin a dialogue with them. She does a quick search for "consumer advocacy" and finds a small number of results scattered over a couple of weeks. Pleased, she adds these to her RSS feeds.

Continuing her search, Ethel visits some other blog search engines, such as Technorati (technorati.com) and Google blogs (blogsearch.google.com/), and generates several more feeds to add to her feed reader. With those added, she’s confident that she’ll be notified quickly whenever someone mentions her or links to her blog. She’ll also have several chances to build her profile in the community by reading and responding to consumer advocacy posts on other blogs. If she goes a step further and registers with Technorati, she’ll be able to get even more information about who’s linking to her blog.

Of course, blogs are not the only place where people discuss and link to things. There’s group discussion, as well.

Step 4: Monitoring group discussions

Ethel, like Fred, proceeds to Google Groups and Yahoo Groups to set some traps. Yahoo Groups is a tough one, because she can’t search the entire database at one time, but has to pick groups that she thinks would mention her blog or her topic of interest. (Ethel can do a Yahoo Groups search for communities containing topics relevant to her blog, but she can’t do searches for messages containing topics relevant to her blog.)

She finds a couple of active groups for consumer advocacy and decides to subscribe to them, both to monitor for mentions of her blog and to find bits of news that might make worthy blog items.

Google Groups is much easier for her. She sets up monitors for her name and the name of her blog, but decides against monitoring for the phrase "consumer advocacy" after she finds neither groups nor good search results that represent conversations in which she wants to get involved.

Ethel spends some time looking for online forums as well (using the techniques used by Fred earlier in the chapter) and does find some boards that catch her interest. Since Ethel is monitoring issues that are more consumer-level, it makes sense that she might get more use from online discussion boards. Her problem is making sure that she doesn’t get overwhelmed with the number of results that she gets.

Step 5: Monitoring search engines

Ethel has to use search engines to set up some traps, but not for searching for consumer advocacy. Instead, she’ll need to use the search engines to monitor for links to her site. She goes to Yahoo, Google, and MSN and searches for links to her site as well as the name of her site. She then sets these searches up as RSS feeds when that’s an option, and alerts when it isn’t.

Consumer advocacy is too broad a topic to include in a search engine. But Ethel could monitor for her name, as well as her blog’s name. This may have to be adjusted over time if another famous Ethel Mae Potter appears on the scene.

Ethel could, if she wanted, access Yahoo and the Open Directory and find consumer advocacy sites that she could cull for news and posts. However, she wants to concentrate on blog content to build that community first, so she skips those resources. Her last big decision is whether or not she should monitor news search engines.

Step 6: Deciding whether to monitor news

In Fred’s case, monitoring news was the cornerstone of his strategy. He needed fresh data and statistics about outsourcing. Since Ethel’s research is more personal and site-focused, monitoring news is less of an issue for her.

There are three scenarios, however, in which she would want to monitor news:

• She is actively courting mentions from the media—by putting out press releases, having herself listed in expert directories, and so on. In this case, Ethel would want to monitor news search engines for her name and the name of her blog.

• She is trying to build journalist awareness in a more low-key way—for instance, by monitoring news search engines for her topic of interest and then contacting the journalist with comments, additional resources, or pointers to her blog.

• She wants to use the news search engines to find bits of interest for her blog. Using news search engines to find stories to add to the blog is better than using general search engines—the stories are fresher, flow is lower, and generally they’re easier to target.

Ethel does have plans to start issuing press releases, so she monitors for her name and her blog’s name. In this one case, she generates e-mail alerts and has them sent to her phone. She wants to be able to react quickly when a journalist mentions her in a story or her press release gets republished. All her other monitoring is done strictly by RSS feed.

Ethel faces the future

Ethel’s needs are somewhat more basic than Fred’s, but she will also have to be ready to tweak and change her traps in the future. She may become well known for her commentary on one particular topic, in which case she’ll need to monitor that topic. She may develop a positive (or adversarial) relationship with a retail company over its consumer-relations policies, and decide she wants to monitor its press releases. It’s even possible that a particular iteration of her name becomes prevalent (Ethel Mae P., for example) and she’ll have to adjust her traps to accommodate that. As she realizes her goal of increasing her profile in the community and the blogosphere, her traps will have to be adjusted to reflect that.

At the moment, Ethel is committed to review her traps three times a week, with ongoing updates to the traps as necessary. Her timeline is open-ended; she has no plans to end the monitoring at all.

Different Strokes

As you can see from these two case studies, different needs are going to use different resources. A resource that’s essential to one trapper may be all but useless to another. That’s the first thing to remember. The second thing to remember is that your initial set of traps is not static. To keep up with the changes in your topic, be it internal or external, you will have to change them over time.

In a perfect world, you would be able to monitor all traps indefinitely. You would never have to back off, or do other things, or drop your topic, or anything. In that world you could let everything putter forever. But it isn’t a perfect world, so in addition to knowing what to look for and how to look for it, you also need to know how to shut things off.

Shutting Off the Flow

The project is done. Or another project is coming along and you’re drowning in work. Or some emergency has occurred and you need to stop monitoring your traps right now. In the cases where you have to put your trapper hat aside for the moment, there are several different ways you can back out of information monitoring, depending on what you’re monitoring.

Shutting down RSS

Shutting down RSS, if you’re using a feed reader, is no big deal. Just stop reading the feed. Preparing it so that you have a way back in is a little tougher. A lot of news is going to go by while you’re preoccupied. Consider using Blogdigger (blogdigger.com) to save old posts from RSS feeds, but remember it’s not a panacea—if you leave an RSS feed unread for more than 30 days or so, you’re going to leave some things unfound and unread no matter what happens.

Shutting down e-mail

You have two options for shutting down e-mail, and they both involve filters.

The first way is to filter all your mail straight into an archive. That way you’ll have it when you need it. However, some old alerts, especially for news searches, will get too old to be found in 30 days or so. Most news search engines don’t keep archives for more than about 30 days, and most online news outlets also have a limit to how long they keep their stories online.

The second way is to filter your alerts straight into the garbage. If you’re sure that you won’t be able to read your alerts for a long time, or you’re not quite finished with the topic (so you don’t want to unsubscribe from the alerts completely), filter into the garbage.

You might be thinking, “Ah, that’s not much different from just deleting them from my inbox!” It is. You’ll find that the unused, unneeded alerts will drown out the current work you’re trying to do. Go ahead and take the time to remove the alerts from your main e-mail. It will make it a lot easier to pay attention to your current tasks.

Shutting down page monitors

If you are using WebSite-Watcher or another client-side page monitor, shutting down your page monitors is as easy as never activating that software again. This simplicity comes with a tremendous disadvantage, however. You can’t keep track of page changes as they occur over time.

With Web-based monitors that send you results by e-mail, it’s different. You can filter those just as you filter e-mail. And you can keep track of several changes that appear over time. But like regular e-mail alerts, be sure to take the time to filter the unneeded alerts where they need to go—either into the garbage or into the archives.

Branching Out from These Basic Patterns

The purpose of this chapter has been to give you a bird’s eye view of how people in two different situations might use the tools that we’ve discussed so far in this book. While you don’t have to follow these steps exactly (and, indeed, lots of references to smaller resources have been left out) this should give you a basic pattern on which to base your own trapping strategy. Build on it from there.