Eric Enge interviews Kirsten Rasanen about TV Guide's video search engine
Published: April 7, 2008
Kirsten Rasanen joined TV Guide in May 2006 as director of product development for TV Guide Online. She is responsible for product development and management of large-scale projects on the company's flagship site, TVGuide.com. Her focus is on broadband video and new video applications for the Web, including a patent-pending product design. Most recently, Kirsten, along with her team, developed the TV Guide Online Video Guide, the first product to focus exclusively on relevant entertainment video search.
Prior to joining Gemstar-TV Guide, Kirsten spent seven years at Sports Illustrated for Kids, ultimately as executive producer of the award winning Web site, sikids.com. In this role, she was responsible for long term project management, new initiatives, and managing all day-to-day activities. While at sikids.com, Kirsten oversaw several complete site redesigns and piloted the creation of a premium content section.
Previously, Kirsten served at Random House Diversified Interactive Media as a senior producer where she worked on interactive projects, Web site redesigns, and advertising initiatives at 13 different Random House Web properties.
Kirsten has taught Web design and technology at Parsons School of Design in New York City. She holds a Bachelor of Science in Television, Radio and Film Production from the prestigious S.I. Newhouse School of Public Communications at Syracuse University.
Eric Enge: How and when did TV Guide decide to start offering videos?
Kirsten Rasanen: Our video strategy actually began about 2½ years to 3 years ago, before I arrived at TV Guide. The company quickly was realizing that video was becoming very important to the web, and that it was going to continue to grow. Being TV Guide we are all about guidance, but we are also about the visual medium; television, video, it all fits together. So, the initial push for TV Guide was two fold. One was to start gathering video to display on our site as an added service for our users. The content that we got from networks, from movie studios was really just a service for users to say "hey, here is the TV show; here is the preview of the show that might be on this week". This helps the users, and helps promote the contents for the networks or the movie studios if it's a movie trailer. That was our first foray into actually having video on the site.
Eric Enge: When was that?
Kirsten Rasanen: That was in October of 2005. We also at that time began to surface video from our TV Guide Network. At that time it was called TV Guide Channel; now it's TV Guide Network. We surface clips of their content as well to round out our video set. That was the initial foray, and round about the same time we also started reaching out to television networks that had websites and that were displaying video on their websites and working with them to get feeds that indexed their content.
We had deals with MTV Networks, with Discovery Networks, with FOX, and others. They sent us RSS feeds that indexed all of the video content that was currently available on their websites. This was again back in 2005, early 2006; so there wasn't that much content out there yet. This was when the networks were just starting to get into putting content on their site.
The data we indexed, was updated maybe weekly by providers. What we were doing was actually gathering the information so that we could display the relevant video content on the TV Guide object pages, - any television show, any celebrity, or any movie that's been out in the last 50 years, we have a page on TV Guide that's a dedicated homepage for that show. So, what we were doing was working with the networks to let our users know about their videos as related to shows, movies or celebrities.
It was the same principle as what we are doing now with our online video guide. For the most part, videos link out to the partners. So, MTV provided an RSS feed for which we have about 300 videos indexed on our site and the data shows what each of those videos is related to. A user can go to the Real World page on our site, click on a video link and watch the video on MTV.
Eric Enge: Sounds like there is a heavy emphasis on TV and movie type entertainment.
Kirsten Rasanen: Yes. At TV Guide one of the things that we have been working very hard to do is to make sure that we are providing information about content that is going to be the most interesting for our users and for people coming to our site; people in our space, which is the television and entertainment space.
Eric Enge: Right. They are not looking for a bunch of homemade videos.
Kirsten Rasanen: Exactly. Our initial attempt of working with the networks was on a fairly limited basis. We had I think about 17 network partners that we were working with. It was again very television focused right out of the gate.
After we started that, it became clear to us that we needed something more than just a list of videos on a show page or on a movie page. We needed something that was a little bit more comprehensive, that brought all of this information into one search area for users. We realized that there was a really good opportunity for us in video search.
Two years ago, there were a lot of people out there trying to figure out how to address video search. But, nobody had a real clear direction; we took a stab at a product that, instead of indexing all the videos across the web in every possible way, focused on what we saw as the content of primary interest for the majority of our users.
We were actually looking at the short-tail (as opposed to the long tail). We were looking for the content that was going to be the most mainstream and of interest for the largest audience. That's how we started the Online Video Guide. Essentially, we said okay, we know that people are really primarily interested in searching for professionally generated content.
Users will go to YouTube if somebody sent them a link for a funny video. I want to go and kind of browse around; maybe if I have 10 minutes I'll go watch something funny. But, for the most part when people are looking for content, they have something specific in mind; and they are looking for content that's actually of a descent quality.
That's why we decided to focus on professional content; content that's produced by more mainstream studios or television networks; and now we have obviously branched out to larger web entities whether it's AOL, or Yahoo, MSN, all those kinds of outlets.
Eric Enge: I do see you have a search engine that has searched the web, but it still looks to be focused on the professionally produced content; is that right?
Kirsten Rasanen: Yes, that was a very conscious decision. We said we are not going to go search 10,000 sites; we are going to search sixty or a hundred of the most important sites. The sites that people are really looking to find video on…
In our space we are a TV Guide, and so we know that the people coming to TV Guide are looking for television related content, celebrity content, music videos, and news. These are the areas that research has shown to be the most popular among users, but our own user base also demonstrates it on a daily basis; that that's what they are looking for.
Eric Enge: Right. You also offer full features versus clips.
Kirsten Rasanen: We do. The product has really evolved since we launched. We launched in beta in April of 2007, and at that time we were focusing on just getting the content out there so that if the user types "Lost" into the search engine we will get you clips from Lost. We are making some assumptions about what the user is looking for as any good search engine does, right?
Eric Enge: Right. It fits the audience.
Kirsten Rasanen: Show the user what they want, not what they ask for, right? It was important to us that we surface that relevant content related to television in the best way possible. We had the product out there in beta for about six months, and we did a lot of testing. We had a lot of focus groups, and we did surveys; and then we also looked at research that was out there in the market place.
We are also keeping up as best we can with all of the latest video research - which is plentiful. We found that a lot of people are really interested in finding full episodes online. As more and more networks make full episodes available to users, they are realizing that it's a viable way to watch the show that they missed. So, we made the adjustment to our product to very clearly delineate full episodes versus clips assuming that a user will want an easy way to differentiate between the two.
Eric Enge: Right. So, you have mentioned a couple of things that I want to expand upon. You presented the example of "Lost". Of course the average user who comes to a search engine on TV Guide doesn't want to bother to type out "TV show Lost".
So, you are taking advantage of the vertical specialization of your search engine to really drive higher relevance in response to queries that a truly horizontal search engine would have difficulty dealing.
Kirsten Rasanen: Exactly. We figured that was our space, and we should take advantage of it.
Eric Enge: Do you think there is any chance that you would branch out in a broader way in the future?
Kirsten Rasanen: Yes, I do. I think for us one of things is that we want to make sure we do is make it really easy for users to find what they are looking for. So, I think that that focus actually really does help the user. If you can lead the user by the hand from the beginning of their search through the end result, you can help them a lot.
So, we may offer a broader search product, but I think we'll still always keep it in verticals. We'll let the user know that there are several different verticals that you can search. We'll offer a news and politics vertical, or a health and education vertical, or an entertainment vertical. We are going to give you the jumping off point and let you search for what you want within a pre-determined bucket.
Eric Enge: Right. So, if they typed health and education; can you see yourself picking up videos from the Mayo Clinic, as opposed to a conventional network?
Kirsten Rasanen: Yes. That's exactly what we are thinking. Down the road, we will branch out and do some vertical search, where we'll aggregate many longer-tail sites in specific content areas. I think we are still going to really work hard to stick to the professionally generated content, but professionally generated in an expanded universe.
So, whether it's the Mayo Clinic or health.com or WebMD; we can aggregate videos from all of those sites and put them in one bucket for the user, and just show the user the different buckets when they get to the site. Pick your bucket and then do your search from there. We are definitely planning things like that. I am not sure health and education is going to be our first priority given that we are TV Guide, we'll probably stick a little bit closer to our area of expertise; at least for starters.
Eric Enge: Right. So, user generated content is not in the cards.
Kirsten Rasanen: Not really. What we are doing right now is indexing the most popular user generated content. We realized early on was that there is so much UGC stuff out there that there is no way to index it in a meaningful way for users. It ends up really cluttering your results when you try to index a lot of UGC. However, users are interested in the most popular stuff out there; Viral Videos are popular for about ten minutes, right?
You get the one clip that's emailed around for two days, and then you never see it or hear of it again. So, we've partnered with a company called Viral Video Chart. What Viral Video Chart does is keep track of the most popular viral clips on the web. And so, we index VVC and we keep the popular content alive in our index for as long as it's on the chart plus five or ten days, and then we expire it. We want to let users find the cool viral stuff that's hot right now, but we don't want to clutter our results permanently by keeping that up front all the time.
Eric Enge: Right. How has this all been received?
Kirsten Rasanen: It's been received very well. We launched in beta last April, and we went out 'live' in October. We have seen terrific growth. We have not done a lot of hardcore promotion; we are still working on the product and promoting it quite a bit from within TVGuide.com, and seeing a lot of traffic coming through there.
We see a fair amount of traffic coming through search. We are optimizing to the best of our ability so that when somebody is looking for Lost videos, we hope to show up the top of the web search results.
Eric Enge: Yes. You have a really good traffic base on TV Guide to begin with; a spot check shows an Alexa Ranking of 1699, which puts you in fairly elite company. So, that's a strong user base that you can take advantage of.
Kirsten Rasanen: Absolutely.
Eric Enge: So, what kind of things do you do to optimize for web search, and what tips can you give to people who read this interview?
Kirsten Rasanen: I will say it's a bit of a science; it's very interesting, and it's funny. Because, when I started at TV Guide a couple of years ago, I didn't know a whole lot about the Search industry. We have a person in our marketing group who is dedicated to the whole concept of search optimization. I worked very closely with him to figure out what we could do with our product to bring it up in search results. So, we have title tags on all of our search results so if you do a search for Lost, the results will give you the Lost related trailers, etc., etc., in the title tag.
We've also done a lot of site mapping. We have over 20,000 shows indexed in the Online Video Guide. Every video that comes in, we classify in many different ways. And, one of the ways we classify a video is by program title or show name, if it's related to a television show. So, we have about 20,000 unique shows indexed, and we have a sitemap that we make available to Google that lists all of those shows individually.
Eric Enge: Now, when you say sitemap, do you mean sitemaps the standard protocol search engines have agreed on or an HTML sitemap, or both?
Kirsten Rasanen: We have both actually. You can actually find our HTML sitemap from within the product. It's not terribly useful, I don't think, for a user, but it is very useful for Google. And, it updates every day, so anytime we get a new show in our index, it is added to that sitemap or any time a show drops out of the index, it's removed from the sitemap. We are working very hard to keep things as optimized as possible. It's a hard science I think. We constantly iterate our approach.
Eric Enge: One thing that you mentioned that was important is that each video has its own page, right?
Kirsten Rasanen: Yes.
Eric Enge: And, that allows you to do on page optimization, so that the title tag of that page; the surrounding text, things like this, right?
Kirsten Rasanen: Every video does have a page and it's got the video title as well as the description of the video. And, we populate the meta-description and the meta-keywords in the page; and we do all the other proven stuff to rank our videos high in search engines.
Eric Enge: Right. So, how important do you think it is to have that separate page then?
Kirsten Rasanen: It's very important. When we put our sitemaps out there, we saw a significant jump in search referrals. I think it's very important. And, the other thing with Search Engine Optimization is that it's not one thing that you do; it's the collection of actions that a site takes that raise its ranking. Every little thing helps.
Eric Enge: Attention to detail is king.
Kirsten Rasanen: Yes. It really is.
Eric Enge: The videos themselves are coming from third parties; so you not involved in putting things in the headers of the videos for example?
Kirsten Rasanen: No, the videos are coming from third parties and they are living on a third party's site. So, if we are indexing videos from ABC, all of that content is actually found on ABC.com. All of the metadata that we are displaying on our site is the metadata that we pick up from ABC.com. Note that we have about 70 different sources now.
Eric Enge: It's intriguing, because you are actually able to rank for someone else's videos and they are still residing over there.
Kirsten Rasanen: Yes. We have different ways that we work with partners. There are some where we are simply acting like any other search engine, and we are actually crawling the websites and collecting the metadata through electronic means. We are collecting that data through our own programmatic technologies.
Then we also work with other sites like CBS. We're part of the CBS Audience Network. CBS sends us a feed of all of their video content. They actually choose what metadata they send us.
Eric Enge: Right. What about the video market overall? What are your thoughts on where we are in the evolution of that market?
Kirsten Rasanen: I think it's amazing how far we've come in just two years. Video has taken over the way people use the web. Two years ago, research indicated that about 8% of people watched videos once a month on the web. And, now we are up to 16% of users watching a full television show episode once a week. We are seeing that the audience is just growling by leaps and bounds, and it's not just kids either. It's not just 14 year to 24 year old group; it's everyone.
My grandmother watches video on the web, so I think it's been an amazing technology surge in the last couple of years. I think it's only going to continue. The way people ingest television is changing; nobody is just watching TV on their TV anymore. Linear television is kind of a strange dying concept in some ways.
People are devouring everything; they are watching it online. They are buying it off iTunes, or other services like it. They are consuming content in so many different ways, and I think that's only going to continue. I think that video on web is a first step towards allowing the user to watch and find content that's of interest to them anytime or anywhere.
Eric Enge: On demand. I don't have to be at my TV at 9 O'clock to see a given show. Even with just a DVR, I can tape it and watch it whenever I want.
Kirsten Rasanen: Yes. A lot of people don't even know when their favorite show is on anymore. You know what shows you like and you set your DVR to record it every time it's on. You watch it whenever you get to it, and I think the same thing holds true online. I think that the web gives you the freedom to watch content whenever you want it, and still find new content.
The other great thing about video on the web is that I may not know that a particular show is a good fit for me. I may never have watched it on television, but you can discover new things more easily online. I think that's happening more and more.
Eric Enge: Right. The discovery process is easier when you can interact and provide indications of the kinds of things you want to see. I think with video that users are going through is a discovery process. Yes, they may want to go see the episode they missed, but they are much more likely to be fishing around just to find something to watch.
Kirsten Rasanen: Yes, exactly. But people still want to watch something good. They want to watch something that's of certain level of quality. I think that one thing YouTube has shown us is that people will watch really lousy quality video online; both in terms of the quality of the video stream itself and the quality of the content, people watch anything when its impromptu and short-form.
But, when they want to find something specific - like a TV show, music video or movie trailer, presumably they would rather watch something that's of a higher quality; that's of a certain standard. It doesn't have to be quite television quality, but it's got to be close enough. I think that video on the web really expands possibilities for people.
I think it gives you an opportunity to find content that you might never have watched before; that you didn't know was even there, but that you might love, you never know.
Eric Enge: Do you think the model is going to move towards commercial free content and a pay per view model?
Kirsten Rasanen: I don't. I think that by and large people expect the web to be free. People would rather sit through a 30 second commercial and watch the show for free then pay two bucks for it. I'll sit through the 30 second commercial, I'll watch the show. If I decide I don't like it - okay, I've only lost 30 seconds of my time. If I pay $2 and then decide I don't like it, then I've lost $2 and my time.
So, I don't think that it's going to move more towards the pay per view model. I think they are going to find new ways to advertise that are less offensive to users, and that are more effective for advertisers.
Eric Enge: Yes. What's happened with commercials in general is that the burden they have is to be as entertaining as the show itself.
Kirsten Rasanen: Right. The interesting thing about the web though is that the opportunities for advertising around video online are so much greater than the opportunities you have advertising on a linear television show. When you are online, you can create ad units that are interactive, while the show is on as opposed to interrupting the user experience and annoying the user as most commercials do.
I think that interactive advertising or advertising around the video content, but not directly within the video content; I think that's another way that content owners are going to realize that they can monetize more effectively then a straight pre-roll ad.
Eric Enge: How about a bit of integrated web search related to what they are watching, so they can search on related things while continuing to watch.
Kirsten Rasanen: Yes, exactly or product placement. This is the thing people have been talking about for years. As I am watching an episode of Friends online letting me 'click here to buy Jennifer Aniston's sweater', right? I mean that's the example that's been used for years and years.
Eric Enge: Right. It would be one price to get one which is the exact same sweater and another price to get the one she is actually wearing right now.
Kirsten Rasanen: Exactly, or a third price to get the knock off. That's one of the nice things about the web is that there are many more interesting opportunities. The technology is there to really blow it out whereas linear television; you are kind of constrained by the box If you don't capture the viewer immediately, then you've lost your opportunity whereas I think on the web; you have a lot of different ways that you can capture the user's attention and engage them.
Eric Enge: Right. Well, one of the things that happen with TV right is that people see the ad and it's kind of sticks or doesn't stick. And, if there is something you are trying to get people to do, it's very, very hard to tell what your results are. In a web environment you have a much tighter feedback loop and much better sense of how successful you were in your advertising goal.
Kirsten Rasanen: Exactly. And then, you can adjust your advertisement based on that behavior, right? I mean you actually have real user metrics to evaluate the success or failure of your advertisement. You can make adjustments to that ad content, based on that behavior to really maximize effectiveness. More and more advertisers are going to start realizing that potential here is great.
Eric Enge: Yes, you can do real AB tests. The first user gets one version, and the second user gets a different version. So, how dramatic is the impact of all this going to be the box in your living room?
Kirsten Rasanen: I don't think the box in your living room is ever going to go away. I think that what's going to end up happening is that the box in your living room and your computer, your PC are going to become maybe not one device, but certainly very closely tied together so that your PC and your set top work together to provide you outlet for entertainment wherever you are and whenever you want it.
They are going to give you that fully integrated experience so that eventually users aren't even going to really know or care where the content is coming from. Is it on television, is it online; I just don't care. I just want to watch it right now.
I think that that's where it's going to go. I don't think the TV is ever going to go away. I think by and large it's still the way people prefer to be entertained.
Eric Enge: Rather than sitting at a computer screen.
Kirsten Rasanen: Yes. The computer has a smaller screen; but it provides you a little more interactivity which I think is very useful in a lot of contexts. But, I do still think that there is a desire to be entertained, to sit back and watch and not have to be active.
You want to really just sit and watch TV; I just think that the delivery mechanism of the content is eventually not going to matter. It's all going to be one, and you may watch it on your PC; you may watch it on your TV. It depends on what you want and when you want it.
Eric Enge: And, what's convenient to do at that particular point in time. So, the TV will have to become similarly interactive although you don't want to have to be interactive. You just want it available when you want to do it.
Kirsten Rasanen: Exactly. You want a choice to be passive or not. I think that's just a question of a little bit of time and a little bit of technology.
Eric Enge: It sounds to me that we have really just scratched the surface. It's probably going to unfold over many years.
Kirsten Rasanen: I think so, but think about how far video has come online just in the last two years. Two years ago we were all looking at 320/240 videos and now it's pretty much all there. Now the user wants HD Video on ABC.com. You don't even have to go to a specialty site to do it. It's readily available, and that technology has evolved in less than two years. It's probably going to be another two years before we see the integrated set top PC experience. But, I think we are getting there. I don't think its 10 years away.
Eric Enge: A lot of things happening in a short period of time.
Kirsten Rasanen: It's really amazing. The evolution of technology in the last 10 or 15 years has been astonishing. It feels like a new industrial age is upon us. It feels to me like we are in the middle of this great technology revolution; it's pretty incredible.
I am realizing that anything is possible, and I am constantly surprised that what new technology is coming around the corner every day. I think that I am starting to realize that I can't even imagine the things that are coming; that it's going to be pretty amazing.
Eric Enge: It's going to be fun.
Kirsten Rasanen: It is fun, isn't it? I think it's really exciting. Actually I love being in this world; I think that video has so much potential and it's all about entertainment, and information, and finding new things. And, I think that video is at the convergence of television and technology; and it's all really exciting.
Eric Enge: Yes. Thanks Kirsten!
Kirsten Rasanen: Terrific. This has been fun. I love talking about this stuff as you can tell.
Have comments or want to discuss? You can comment on the Kirsten Rasanen interview here.
Other Recent Interviews
- Did-It's Kevin Lee - March 31, 2008
- Wordtracker's Ken McGaffin - March 24, 2008
- Site Tuner's Tim Ash - March 10, 2008
- Icisive's Matt McGowan - March 4, 2008
- Avinash Kaushik - February 18, 2008
- Danny Sullivan - February 11, 2007
- Google's Adam Lasnik - February 4, 2008
- comScore's James Lamberti - January 28, 2008
- Incisive's Kevin Ryan - January 28, 2008
- Eurekster's Grant Ryan - January 14, 2008
- Eurekster's Steven Marder - January 7, 2008
- Google's Sep Kamvar - December 17, 2007
- Microsoft's Grad Conn - December 10, 2007
- Seth Godin - November 26, 2007
- Microsoft's Ramez Naam - November 12, 2007
- Google's Matt Cutts - October 8, 2007
About the Author
Eric Enge is the Founder and President of Stone Temple Consulting (STC). STC offers Internet marketing optimization services, including SEO, Social Media and PPC optimization, and its web site can be found at: http://www.stonetemple.com.