T: This morning I am talking with Rachel Lovinger, a Senior Content Strategist for Avenue A | Razorfish. Rachel will be presenting at next week’s Semantic Technology Conference. In addition, she recently published a paper on Boxes and Arrows. And just this week, she started her blog Meaningful Data, available here http://meaningfuldata.wordpress.com/. Good morning, Rachel.

R: Good morning, Tobi Jo.

T: So, I thought I’d start by asking how does the work of a Content Strategist differ from or overlap with an Information Architect/Interaction designer?

R: Well, this is something I’ve thought about a lot because when I wrote that article for Boxes and Arrows, at first, that’s what I thought I was going to write about. Strictly speaking I think the goal of the Information Architect and the Content Strategist is basically the same - and that is to come up with the best way to present information so it is useful and meaningful to the user. Primarily the way the field is going right I feel like Information Architects and Content Strategists just take a slightly different perspective of it. While the Information Architect looks at the physical presentation and how its organized and structured as the person is interacting with it, the content strategist looks at it from the perspective of the content itself and says “does this content have everything in it that will enable it to be presented meaningfully?”. So, it’s like if you as an Information Architect have devised a really innovative and compelling way to present up the information and the content then my role as the Content Strategist is to make sure all the pieces are in place to support that presentation.

T: I see. So, how did you start doing Content Strategy? What is the first job you had that relates to what you do today on a daily basis?

R: I guess, most directly, it would be Entertainment Weekly’s EW.com <Tobi’s Comment: I met Rachel working on this project :->. I started there as an HTML producer and became a template developer. And at the time we were doing a redesign of the site. And as we brainstormed it became clear that was a camp of two people, of which I was one, who felt like the site should be organized in a very meaningful way based on the subjects of the content and the data that was available about it. And then there was another camp that felt that the content should be organized based on whatever was favored by whoever had a vote. So, the one other person and myself who felt it should be organized in a meaningful, logical way went to work on convincing our boss on our point of view. And we started out just looking at the content ourselves and started having conversations on a better way to organizing the content so that it was making the most use of the entire archive of content they have.

T: When your refer to meaningful organization, what to you qualifies as meaningful organization?

R: Well, I guess the nice thing about Entertainment Weekly is that the content fits cleanly into categories that are very evident to everybody and the subjects are pretty well known to anyone who is interested in entertainment. And even the magazine was grouped by basically what kind of entertainment it was talking about so you have content categories such as “movies”, “books”, “music”, “television”, and so on. So, that was the first thing that was obvious to us. So, if the magazine was the website might as well be also.

T: How does that kind of content organization get decided? Consensus? The Dewey Decimal System? Subject matter experts? Who decides what are meaningful groups?

R: I think this is challenging because different people would tell you different things. I went with the Senior Data Librarian at Time Inc. Interactive to the ASIS&T conference a few years ago, and it was a much more academic conference than I expected. A lot of people there were coming at information science from a very academic perspective. Whereas, she and I were coming at it from more of a practical, we-use-this-in-our-work kind of way. One thing that we noticed is that academic library scientists may want to capture different kinds of data and may organize things from a different perspective than we would in the publishing world.. What we did was very much related to the domain of knowledge of the product we were organizing.

T: Where does this “domain of knowledge” you’re referring to come from?

R: Both the user and the stakeholders have a shared understanding of what the world of their product is about. So for Entertainment Weekly there are movies about every subject you can imagine. So, in theory we could have had a whole section (of the site) where these movies are organized by their story line. So this movie is about travel and that movie is about, you know… um…

T: Rock-n-Roll?

R: Yeah. And steel mills. Right. So, if we wanted to capture that kind of information we could theoretically create a system that covered every subject in the world. But, we didn’t want to do that because the primary focus of that magazine was the domain of the entertainment industry. Whereas, if Warner Brothers Movies was trying to figure out how to categorize their movies, they may want to organize it by what the movies were about. But that wasn’t nessecary for what we were trying to do.

T: Why can’t I just rely on keywords in my meta-tags and let the beautiful perfect Google algorithm work out the rest?

R: Well for one thing, the Google algorithm wasn’t available at the time when we started doing this. But, for another thing what keywords don’t give you - and let’s assume that the keywords are somehow controlled and they’re consistent and nobody ever spells anything wrong (laughing: that’s a big assumption, because that doesn’t happen) assuming that did happen - what keywords don’t give is the distinction between something like Harry Potter the “book” and Harry Potter the “movie”? Or how do you indicate that these seven Harry Potter books are part of the same “series”? With keywords, you could indicate the name or type of the product but you couldn’t really indicate what kind of product it was or create relationships with other kinds of products that are in it’s family.

T: So how did you tell the computer what a “movie” is? I mean, how did you explain to the system what the relationship was between this Harry Potter “movie” and the “book” with the same name? And how did you name that relationship?

R: We worked with our IT department to create a proprietary system for categorization that had a hierarchy built into it so that you could group things under a heading of “movie” and all of those things would be movies. And you could group other things under the heading of “books” and the system would know all of those things were books. And then we created the ability to draw relationships between those things. So now that we have defined books and movies, we can connect them with the following relationship: this movie is an adaptation of the book.

T: How is that different from a spreadsheet or other hierarchical or nested information like site maps where you would have connected information?

R: Well for Entertainment Weekly that was just one of many many relationships we wanted to capture and this is the reason why it was so important for us to approach the content in this way. This is because entertainment products inherently have many clearly defined relationships for each entertainment product. So, when you have a movie there are all the people that were involved in making it, there’s the release date, grade, related products and we wanted to be able to capture all those relationships in a central location because if they had seven different articles about that movie they had to manually enter all that information in each article. And we thought, wouldn’t it be great if we had all these relationships in one central location and they were available to every article about that movie.

T: Yep. So, if I was an editor entering an article about Harry Potter, and tagged it with “Harry Potter” that tag would reference the central location where you’ve defined all the actors, directors, producers, release date, and so on for me?

R: Yes, when using this kind of content organization system everything that you’ve tagged with Harry Potter inherits those relationships.

T: Well, that seems like it simplifies a lot of the pitfalls that arise from standard tagging and the whole world of user generated tagging that’s gaining awareness from the popularity of Flickr and Del.icio.us. What’s your take on that?

R: Well, I think there are definitely difficulties with user generated tagging. I use Flickr and I love it. And I think it’s wonderful when I’ve tagged photos or when I want to find photos that other people have tagged. Usually you can find pictures of anything you can imagine. But, some people don’t tag their photos and other people add forty tags to a photo. Some of those tags may be less relevant that others. There are also problems when you have many different terms for the same thing. For instance if I wanted to say my picture was related to New York, do I say “Manhattan” or “New York” or “NYC” or all of them because someone may be looking for any one of those things. That being said, I think user generated tagging provides a lot of information that might not otherwise be there.

T: So, if I’m on a site that relies solely on user generated tags, when I search for NYC I probably won’t see photos tagged only with New York unless it was also tagged with that. The system doesn’t know they’re synonyms; but, a structured data system could have as many synonyms as you wanted available. But, your job is to put in place the organization system so that when I type in NYC the system also searches for the synonym New York. Is that right?

R: Right. So I think the place where user tagging is going to be successful is somewhere like Flickr or Digg where the user is tagging to keep track of his or her own photos or stories. This seems to work best if there is a critical mass of people who tagged a similar thing similarly, and those things can rise to the surface. Where I think user generated tags may not be as successful is somewhere like Amazon where there is less benefit to the user to tag something. I think for Amazon the content that’s on there - with the small exception of people who are very avid reviewers – your use of that particular piece of content usually ends when you buy it. Because of that limited relationship with the content, I think people are less motivated to want to tag it.

T: So, do you have a book list of basics for people interested in content strategy? Blogs? Conferences?

R: Well, the ASIS&T conference is a good place to start. But, I’ll be updating my blog on a regular basis with information on content strategy.

T: Great! Well, thank you for taking the time to talk with me today. I’m looking forward to following up with you in the future to talk about how all this is evolving.

R: You’re welcome, Tobi Jo.

This entry was posted on Thursday, May 17th, 2007 at 2:02 pm and is filed under User Experience, Content Strategy, Basic. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.