Part II:

Transcript

Brett Tabke: This is Brett Tabke Tabke, we are back here in the Silicon Valley and today we have Gord Hotchkiss Hotchkiss of Enquiro. Thanks for being with us Gord Hotchkiss.

Gord Hotchkiss: Oh. Thanks for having me.

Brett Tabke: President, founder of Enquiro. What are your duties these days?

Gord Hotchkiss: You know that is a really good question, and I haven't been asked that for a while, um, generally the more I keep my hands off the client accounts, the better things run, so I you know for me it's been really fascinating kind of looking at where search is going, with things like personalisation and universal, you know I do a fair amount of writing so I've got like a weekly column. I do some writing for search engine land and for some other things, so. That's great because it keeps me connected with a lot of people that are really defining what the search experience is going to look like in the future. It gives me the chance to talk to you know the Marissa Mayers, and on occassion Matt Cutts, and Nick Fox, and Larry Cornett at Yahoo and Justin Osborne at Microsoft, and I really enjoy that because for me it's all about user behavior as you know, um, and so talking to people that are really thinking what does that search user experience look like, um, has really been a rewarding part of my day to day job, and I'm glad that my team at Enquiro let's me indulge that side of things and spend some time exploring that and then of course we bring that back and try to translate that into value add for our customers as well, so.

Brett Tabke: You are also on the board of SEMPO. I was on it for the founding year, I guess. Chairman, what are your duties, you are not the president you are a chairman. What are you duties as chairman on the board?

Gord Hotchkiss: Ok, the official title of the chair is to help direct the strategy of SEMPO, so Jeff Pruit, our president kind of takes care of making sure the ship stays on track. I kind of say, you know what Jeff, maybe we should start steering for there or start steering for there. So, and that is a great team, it has worked really well. Jeff has been awesome, as was Dana before, and Barbara before Dana. The organization has just continued to grow. As you know being on the board in the first time, it is a real challenge in any industry getting an organization off the ground, and certainly kudos to that first year board because they made SEMPO stick around for year two and year three and year four and that is not an insignificant challenge in an industry like the search industry where everyone has pretty strong opinions of what should be and what shouldn't be. And to try to get an organization together that represents and aggregates all those opinions is just a gargantuon task. And it is natural as an organization evolves that some people are going to rally around what the mission of that organization is and some people are going to disagree with it. The organization naturally defines itself as it moves forward, um, and I think that was a lot of the challenges that SEMPO had in year one was just who are we, what do we do, and everyone had a different opinion of that.

Brett Tabke: Oh there were some wars on the board meeting that were just monumental.

Gord Hotchkiss: And you know I have always said that that board in year one did an awesome job because they kept it going.

Brett Tabke: At the end of the day we all came out and said okay we are headed this way, that is our differences, they are behind us, we are going to do this.

Gord Hotchkiss: So, SEMPO is obviously we are still, relative to a lot of organizations we are still a small organization, we are still dealing with limits of resources, both in man power and budget, but I think we are really making a difference in the industry. I think we are emerging as a voice for the industry and we are certainly trying to do that better and more often. Um, you know we are looking at giving people the training and the best practice experience through SEMPO Institute. We are looking at conducting more research, um to help prove the value of search in non-traditional ways. You know beyond the direct response channel that everyone seems to use search for. We have got over 600 members in over 33 countries now, so it truly is a global organization now, so my time at SEMPO has been very rewarding. I'm glad I got on board.

Brett Tabke: You guys have done an outstanding job.

Gord Hotchkiss: Well thank you.

Brett Tabke: The leadership has just been just outstanding in the last couple of years.

Gord Hotchkiss: We have had an awesome team to work with.

Brett Tabke: Well we were talking a little bit about the research SEMPO has been doing, but you are probably the best known for the research you have done. The famous eye tracking study you did, the original eye tracking study done on Google SERP's.

Gord Hotchkiss: Yeah, we are continuing down that road. We got an interesting variation on that. You know we did the first one in 2005, I think, or late 2004. We did a follow up last year on comparing Google, Yahoo, Microsoft. One of the questions that was coming out was, obviously I'm sure you've seen it, I've seen more innovation on the actual interface, the look of the interface in the search space in the last 6 months than I've seen in the last 10 years. You know, Ask unveiling 3D search. Google unveiling both universal, blended results and personalization. Yahoo has been making changes on their interface. Microsoft is playing around with some stuff. I really think that although most of the innovation in the search space has been a little bit hidden in the background, tweaking algorithms and improving infrastructure, I think the next big moves will all be on the interface side. The UI side, and of course for us watching what that user behavior looks like. The question is well how is that impacting users? Obviously the engines are doing their own internal testing, but that tends to stay internal. So we were asked, we said, well have you guys ever done eye tracking on personalisation and we weren't exactly sure how to do that because personalisation introduces some interesting challenges and methodology, it's based on past history which is tough to do in a one off panel, uh but we wanted to get something out there, so what we did is we kind of took an interesting approach, we said okay universal is fairly easy. We can do certain searches and see universal results. We can eye track that now. So we did that. Then we said well personalization as it is currently being implemented by Google, we think we are at the very earliest stages of that and I don't think its truly impacting the user experience that much, but what if we look 12 months down the road, and say they have got the algorithms working, so the results they do show that are personalized are relevant, they have got strong information sent, and what if we say okay there is three slots on the page that are ear marked for personal results, if they exist, what if we set up a scenario, we get a panel that it was kind of a friends and family panel because we were imposing on them, we set up an initial session where we give them a scenario to just go out and do anything they want and we track all their click stream so we have a personalized, a personalization data set to look at. And then we call them back for a second session and say okay pick up where you left off, but then we start giving them personalized results. So that is what we did. We did eye tracking on that to see how strong a draw those personalized results were, and then we went one step further still, and we said well I wonder what a search results page could look like in 2010. So I did a series of eight interviews. I interviewed Marissa Mayer at Google. I interviewed Danny Sullivan, and Chris Sherman. I interviewed Greg Sterling. I interviewed Larry Cornett at Yahoo. Justin Osmer at Microsoft and Jacob Neilson. Um, and we kind of aggregated some of the themes, and our in-house designer put a mock up together of what a Google results page could look like in 2010, heavily personalized, um and we did eye tracking on that, so we kind of looked forward 3 years in the future, took some bench marks, and said I wonder what the user experience is going to look like in those things. I don't know if I am sad or happy, but the golden triangle does not look like it is going to be as standard in the future as it is now. The introduction of graphics on the page, introduces some really different scanning activities. We are used to interacting with a page that is predominantly text based. So it makes sense to scan that in a linear manner. It makes sense to start at the top and kind of just work your way down, and that produces the golden triangle. Well, when you start putting different graphic elements like an image on their, in a universal search result, suddenly the top to bottom scanning doesn't make as much sense. There is an image there and the image could be appealing. It offers a different flavor of information sent, so it tends to attract the eye. So what we are seeing a lot is that if you have an image in the third or fourth spot you fixate there first and then start your scanning there. So rather than the F shape scan, creates the golden triangle, where you scan down and scan across when something catches your attention, we actually saw more of an E shape, where the middle of the E is adjacent to where the image was. You look at the image, you scan across, then you make your choice whether to go up or down and then scan the top or the bottom. So rather than kind of a linear scan, we saw actually what we call page chunking, where you kind of slice the page off based on where the image is, so it is everything above the image, everything below the image, everything beside the image, and then you kind of pick and scan those chunks as units rather than a top to bottom scanning. Now in a lot of ways that is not that significant of a thing, but when you think about a fundamental change at how users look at search results, it suddenly changes a lot of it because the power of the top sponsored ad is it's right at the beginning of the scan path. It's right where you tend to fixate and start down. Well if that's suddenly not at the beginning of everyones scan path, if it's different depending on what else is appearing below it, that introduces some very interesting implications for marketers.

Brett Tabke: Well, just the fact that you got the image there and it is pushing the content down the page, you know you are looking at a break now at the fold of ten twenty four seven sixty eight, at about position three. If you got weather or graphics or whatever up top it can push it down all the way to one being the top result, organic result.

Gord Hotchkiss: Right, and it is not only pushing the content down, it is changing the way you scan that content so you combine those. Then you add something like personalization. When we did the test, we did it for the same query and we showed one group basically what is coming out of Google right now, and then we did the mock up with three personalized results that happened to be in organic spots three, four, and five. So, they appeared, but not in top spot. The personalized results out performed the standard Google results by a factor of 2x in everything in gaze time, in number of fixation, and 3x plus lift in click throughs. So personalization does work, and does work well when it is implemented really well. So if you have universal chunking your page, and you've got personalized results appearing below that, now you are even drawing more attention from the top of the page down. So it's kind of a new ball game. You know you combine that with the fact that everyone I talked to at the three major engines, Marissa, Larry, and Justin all said there is a lot of change coming in the search interface over the next two to three years, um, and you know the linear presentation almost everyone said, "that paradigm is dead". We are done with the ten blue links, the top sponsored, the side sponsored. We just don't want to change it so much that it is jarring to the user, but we know that the future is richer, more functional, interface, we are going to be trying some new things. So not only are we changing the linear results we are really changing the look to be probably in a lot of cases more like what Ask has done with 3D search, but ultimately the user is going to decide what they respond to and what they don't.

Brett Tabke: Now, you just put out this eye tracking study, correct?

Gord Hotchkiss: Yeah, actually we did a special sneak preview for a few clients, out right now, so we are probably a week or two away from making it publicly available.

Brett Tabke: Ok. ok. Speaking of time, we are out of it. I appreciate you taking the time to talk to us.

Gord Hotchkiss: Thanks Brett Tabke.

[edited by: Vanessa_Zamora at 8:56 pm (utc) on Jan. 20, 2009]