October 17, 2006
Find out what 3,944 marketers revealed about costs, conversions and what is working now for search marketing. Highlights include:
1. How to cope with rising cost per click. 2. New “norms” in search clickthrough and conversion rates. 3. How consumers use search engines for offline purchase decisions. 4. Eyetracking study on Google’s changing “golden triangle.” 5. PR and search explosion.
This is your downloadable version of a presentation first conducted live Oct. 12, 2006, by:
Anne Holland, President
Stefan Tornquist, Research Director
#1. Click this link to download the PowerPoint presentation PDF including 8 new data charts (yes, you may share with colleagues):
#2. Click this link to download the MP3 audio file:
http://www.MarketingSherpa.com/tele/SMBG10-12.mp3 (8MB file)
(Audio download note: This is in MP3, which almost anyone with a PC or Mac can play. If you'd like to listen now, just do a regular mouse click (left side) and it will start playing fairly quickly. If you'd like to download and save to listen later, just do a right-click with your mouse.)
#3. Here is the transcript ofthe teleconference:
Anne Holland: Good afternoon and welcome to MarketingSherpa’s annual search marketing teleconference on our latest research data. Today I am joined by our director of research, Stefan Tornquist. Stefan, how are you doing?
Stefan Tornquist: Very well, thank you.
Anne Holland: And, of course, I’m Anne Holland, President of MarketingSherpa. For now this is live and let’s go for the gusto. Stefan, where did the data come from that we’re presenting today?
Stefan Tornquist: OK, so slide number two, this gives you the full list, but I just wanted to highlight a couple of things. This is the third annual search marketing benchmark study we’ve performed, and for the third year in a row we’re pretty sure that this is the largest study conducted of SEM marketers and agencies. So we’re really excited about that. At 3,944 respondents, that’s a lot of data.
We also followed up with them -- and this is a new feature that we’ve added to our research process -- we followed up with some of the marketers who responded to the survey with in-person phone calls and that was really a great addition because now we’re getting some context around the numbers, especially when someone’s an outlier, we can go and find out, well, why are you spending 100% of your budget on such and such? So, that’s very informative.
Anne Holland: Yeah, why you’d have a weird conversion rate? Yeah.
Stefan Tornquist: Right.
Anne Holland: Was it a typo or is it real?
Stefan Tornquist: We also asked marketers for a special survey about click fraud. It’s certainly been in the news, and we wanted to gauge just what does the average marketer think about click fraud. And we also conducted some partner research as we always do. This year we did phase two of our search marketing eyetracking study, again performed with iTools out in San Francisco. We’ll be looking at some of the results from that. We also performed a search engine loyalty study with the folks at Compete, and then we’ll be looking at some of those results.
And, finally, in cooperation with SEOPR we conducted some SEO PR for one of our own products to see how optimization makes a huge difference with press releases. Of course, we also benefited from the great secondary research that we always get from Research Partners, including a lot of unpublished data form sources like HitWise and Nielsen//NetRatings.
Finally, there’s always the benefit of your interviews and case studies from those of you on the editorial side. That helps us … take a look at the year and when we’re starting our research, know which areas to hit and which ones to avoid.
Anne Holland: Great, so if there is a number in here, be assured that all numbers we’ve got a heck of a lot of data around them, anecdotal, research-based, secondary research, you name it. And if there’s a number that looks awfully odd to us, we do research it three different ways from Sunday to make sure there’s some resemblance to reality before we let you guys have it.
Why don’t we move on. Opportunity number one, search engine optimization yet again. Now, I actually put the word again on this slide on top of Stefan’s chart because sometimes I feel like every year we do this teleconference and every year I’m like, you know, SEO, SEO, SEO and, of course, here you can see from the actual answers of the marketers themselves that when Stefan asked them to rate these one, two, three, four, five, eight different tactics of marketing, which gave the best return on investment at SEO, one. It won for two different types of marketers.
Stefan, can you describe what’s a product marketer versus a lead-gen marketer?
Stefan Tornquist: Sure, well the product marketers are generally looking for a direct sale to occur either on the Web site or offline whereas the lead-gen folks, they’re mostly B-to-B, but it also includes big-ticket consumer items, mortgages and automotive.
Anne Holland: And so what we see here, and, of course, most people know this, but every once in awhile, you know, we always get somebody in the audience who has a question, what is the difference between search engine optimization, SEO, versus, and as you see two notches down, paid search marketing, CPC? The difference, of course, is SEO are the listings that come up naturally, some people call it organic listing -- those are the ones you’re not paying for. The listings that are paid search marketing are the ones where you actually say, “I’m going to run an ad,” and you type in your ad and you pay for them. So there’s a real different there and what we are seeing is that SEO. Now usually it’s a little bit cheaper, but you do get a very good ROI.
What I think is interesting here is that it’s really holding up nicely next to house email marketing and everyone knows your house list, which is your own opt-in list, is usually one of your strongest marketing vehicles ever. So that’s fairly strong, and so there are also a couple of interesting ah-has where direct postal mail is doing better than online ads such as banners, which is also interesting these days, but I assume people are a little more careful about picking the lists for their direct postal mail given how much it costs to send each piece than they are with online ads where you can reach millions of people for much less money.
Stefan Tornquist: Well, I think the other thing that’s interesting about this is that because it’s a question about ROI and for many companies SEO and house e-mail are very small expenditure and in some cases people will say, “Oh, well SEO is free.” And we’ll talk a bit more about that, but since it’s a question about ROI, this chart emphasizes the power of paid search to me because paid search unlike these other two tactics is not inexpensive, well, for very few. And so even with rising click prices, paid search performs very well and what makes that even more impressive is that as marketers who haven’t been at the search game all that long, certainly not as long as with email, and measurement of the full impact of search is really still in its infancy. We know from studies like that performed by comScore that the overall impact of search, most of that happens out of site of our basic measurement which covers sort immediate actions and latent conversions of people who have been cookied and come back to the Web site.
So most of that is happening out there in the world in places that we can’t measure. Even though I know most of the conventional wisdom on searches is that it’s kind of plateauing, it’s not going to keep growing the way that it has, and I would generally agree with that, I think we’re going to see strong growth for quite a while because as more information goes in and we get more view into that 80% of search that’s not going measured, we’re going to be able to assign dollars to that 80%.
Anne Holland: That’s pretty cool. Well, we’ve got some more slides about that further on. Why don’t we move on to the next one?
This is, of course, the slide that everybody wants to know, which is how are clicks converting? Now a conversion could be whatever the marketer in question decided it was. For example, for an e-commerce marketer a conversion is somebody clicks you and then they bought something. For a lead generation marketer, it would be somebody clicked through and then perhaps they registered, maybe they registered for a webinar or a white paper, or they opted into an email list. For a brand marketer, they might, or a regular retail store, they might have clicked through and then gone off and seen where they could find stuff in the real world or read a whole bunch about a particular product, done enough activity to indicate some serious interest.
So conversion can be whatever the goal of you campaign was. Now, Stefan, can you explain what these bars are that we’re seeing on either side?
Stefan Tornquist: Sure. This is the most general chart that we’ve got about conversion. In the book we look at it through a number of different lenses, you know what is the goal, the type of marketer, price points, that sort of thing. But here we’re looking at all different, I think we look at eight different types of conversion goal in the book, this is all of them boiled down to one comparison.
Anne Holland: Right, so your conversion may not match this precisely at all. This is a big giant aggregate number.
Stefan Tornquist: And like any average it’s dragged up and down by the outliers, and so that’s why we’ve started to really break these averages down so we can look a little bit more granularly at these things. So in the middle we’ve got all site visitors. Now that includes people from search, paid and natural, but it also includes folks that are driven by email, direct navigation, etc. And the three overall conversion rates don’t vary that much, although what you see is that when you start breaking down that average, there is a big difference. The top quarter percentage, that means that a quarter of our respondents to this question, if you average their conversion rates, you’re up at 10.13% for paid search and up at 5.11% for natural.
If you’re wondering why there’s that big discrepancy for paid search, well, this is my theory. My theory is that a lot of the folks that click on paid search ads have just conducted what you might call a navigational search, you know, I just put Nike into Google and one of the first things I see is a top listed ad for Nike and I go there. Chances are I had a product in mind and I think that’s what raises our conversion rate. Whereas, the natural search results, even though they typically get the majority of the attention and the majority of clicks, people have a much wider degree of interest when they perform a search.
Anne Holland: OK, so maybe I’m looking to educate myself as opposed to, yeah, I’m shopping, I know this is an offer?
Stefan Tornquist: Right. If I put in Bombay.com into the search bar, which by the way is very common –
Anne Holland: A great brand and a place.
Stefan Tornquist: Right, but then it is very common to use the search bar as navigation. You know? It’s one of the things that complicates some of the data on click fraud is that a lot of people instead of putting in, you know, Nike.com into their favorites or –
Anne Holland: Or browser.
Stefan Tornquist: – or browser directly, they just type it into the search bar because they know that if they misspell it a little bit or they get the dot-com versus dot-net wrong, they’re going to see the result that they want thanks to the search bar.
Anne Holland: Or if you’re my mother, you think you actually have to do that to get to a Web site.
Stefan Tornquist: Well, that’s probably true of a lot of folks as well, but navigational searches make up a big chunk of the most popular searches that we see are brand and site specific.
Anne Holland: Alright, but what we are seeing here is that conversion rates, natural versus paid, I mean there’s not a huge difference there, so that organic traffic is converting pretty darn well just as the paid traffic is converting. So it’s not like there’s this complete out of whack, you know, gee, one’s much worse than the other.
Stefan Tornquist: Absolutely not, and one thing that this doesn’t show is the conversion for specific terms that got folks there versus via natural search. You know, that’s one of the best tactics for figuring out what keywords you want to optimize your site for and to buy in auction are those terms that convert highly, and natural search can potentially bring in more people than the paid side.
So it may not have the same top-end average, but in terms of ROI it’s fantastic.
Anne Holland: Let’s move on to the next chart, which talks about spend and this, I got to tell you, SEO versus PPC and spend, this is the chart that surprised me. We’re, you know, our goal here is always to say what happened? What’s working, what the data is right now as opposed to making any forward-looking remarks because, you know, the future could be anything and what matters is what’s going on right now. However, last year, between 2004 and 2005 what you saw was what was expected, a huge jump in paid search, but there was also a huge jump in SEO, which is the smaller bars. So the bars on the left is paid and then the shorter darker bars to the right of each paid bar are the search engine optimization bar, the little tiny bar there, and when I saw last year, gee, it went from $330 million to $740 million between 2004 and 2005, I thought, whoa! Look at that arc, it’s more than doubling and I was so excited, I was like, ooh, people are getting search engine optimization religion and they’re seeing the light and it’s great and incredibly efficient, you know, wonderful traffic we all need to do it. And then I thought, boy, by this time in 2006 it’s going to crack $1 billion , you know, I absolutely was like it’s gonna crack a billion, and, boy, was I wrong.
Look at that, it basically just held dead on steady. I mean 740,000 to 790,000, not a big change. And it looks, Stefan, from other data that we have, including surveying 104 search engine optimization firms about their clients lists as well as surveying clients and non-clients about their likelihood of switching to a firm or not, taking in in-house or hiring out of house and how they’re budgeting, it looks like this is really a realistic number, that suddenly just the growth halted. Do you have any feeling for what went on here?
Stefan Tornquist: Well, I don’t see this in quite the same grim terms that you do. For one thing, we’re talking about a $50 million increase in what is considered by many to be a virtually free tactic. You know, here it is at just about $.8 billion paid experienced dramatic growth, I don’t think anyone would have expected SEO to accompany it, and you know we’re looking at SEO as a percentage of paid spend to continue to participate in paid search marketing last year you were looking at roughly a 22% increase just to stay where you were. Now SEO doesn’t place that same kind of demand on marketers, so I –
Anne Holland: Right, prices aren’t going up every year necessarily.
Stefan Tornquist: And this is the truth in advertising about the data around SEO is that a lot of marketers for one thing don’t have SEO as part of their budget. SEO falls as part of the time to someone’s who is on the Web team or on the IS/IT team, so in the eyes of a lot of marketers and we see this in their responses, they’ll put down the cost of SEO as zero and then they’ll also give us lots of data about their SEO so we know that they’re doing it. So in part this is a lack of communication internally about SEO and what its practical costs are.
Anne Holland: The nice thing is that the marketers are measuring, and thank goodness they are, the traffic they’re getting from optimized listings and from natural search. It might make sense that especially if you are budgeting right now for your next year, if you’re on a calendar year or fiscal year, to go to whatever department it is and make sure they’re aware of how much is coming is as a result of their expenditure into SEO. If some other department is spending on it or they’ve got a staffer who’s helping out with it, they should know that this is affecting the company’s bottom line. It would be important to see a lot of inter-departmental communication on this front.
Stefan Tornquist: It’s definitely true that company structure and politics are a big issue here. In a lot of companies, marketing has only tangential control over the Web site especially since SEO’s really at the meat of how a site is structured and how pages are written, which isn’t really marketing’s gig, at least not in the way that we’ve come to think about it and certainly not the way the technologists think about it.
It’s interesting, in a recent survey by iProspect, I think 65%, 64% of companies that use outsourced SEO, they found at least one obstacle to implementing those recommended changes when it came time to actually getting it done internally.
Anne Holland: Office politics?
Stefan Tornquist: Well, yes, office politics and just site construction, you know, these things are not as straightforward as they seem, and the biggest reason that those people who saw those challenges, the biggest reason they cited was a lack of human resources. So you can imagine if that’s how it is when you’re paying a company to provide you with information, and there’s obviously a real initiative around it, it makes you curious about how it is when the program is entirely in-house when you’re not spending those outside dollars and you can just as easily, if you’re the IS/IT folks, say, “We don’t have time for this right now.”
Anne Holland: Right, the marketers kind of holding their cap in hand saying, “Please, please, please,” and you have to get to the end of the line because they’ve got a long project list.
Stefan Tornquist: Yeah, finally the other why is SEO not getting the amount of attention is deserves and the amount of money it deserves, I think it’s sort of analogous to email. It’s not one of the big-ticket items. Emphasis is on page search, emphasis is on other expensive tactics –
Anne Holland: Even though with SEO, the ROI is fabulous.
Stefan Tornquist: Well, I think, you know, one of the things that happens for a lot of companies, they’ll go and take a look at those key, those main keyword phrases where they expect to show up, and they’ll see themselves doing OK in those areas, and they don’t necessarily check to see whether they’re optimized against the secondary keywords or the keywords that are converting really well in paid search.
Anne Holland: A CEO is going in and looking at his company name, he’s like, “Oh, we’re here,” and he may not be saying, “Well our best converting paid search terms are this and this and this. Are we there, too?”
Stefan Tornquist: They look on their own page, their competitor’s page, and maybe one or two key product feature pages, and consider it a job well done and that’s just part of the story.
Anne Holland: Why don’t we move on. Opportunity number two, rising costs per click and, boy, it is continuing to rise. It’s a little bit slower of a rise than it was last year, but it is certainly rising. Again, Stefan you have this broken out in the main guide by all different kinds of marketers and campaigns, but so this is a very general chart.
Stefan Tornquist: This is the most general of the charts. Internally, we’ve sliced it by types of goods and campaign goals and we also break down the averages just like we did on the conversion slides. So you’re not just looking at one number, you’re looking at how different segments of the market experience that, how that average breaks down.
Anne Holland: But we’re basically, you know, average cost per click is arcing up toward two bucks. Either way it looks like Google and sooner or later Yahoo and MSN may follow.
Stefan Tornquist: Right, and what we see when we do break down those averages is that actually not that many people are paying two bucks, but some people are paying a whole lot more and, well, when it comes to Google people are paying a lot.
Anne Holland: Let’s move on to the next chart, how are your marketers coping with these rising click costs? This is actually data that was collected by SEMPO, which is the Search Engine Marketing Professionals Organization, and they do a nice annual survey. What’s interesting was that this data follows a lot of our anecdotal data that we get from case studies and talking to marketers one to one. What we discovered was if you’d asked this about a year and a half ago what you would have had marketers telling you was “Oh, I’m reducing my average cost per click by just broadening my campaigns like crazy.” Everyone a year and a half ago went from like 500 click terms to 2,000, 20,000, 200,000. I’m just going to expand like crazy, take advantage of the long tail, get these five-cent clicks and it’ll offset my $15 clicks.
That to some degree has changed partially because, of course, Google and Yahoo! and the rest of the search engines made it so easy to expand your campaign, they saw the good in that. They were like, “Wah-hoo, let’s sell advertising like crazy on every term.” So they made it so darn easy that it no longer became a competitive angle to actually do that so much. Your competitors were doing it, too. It was awfully easy for everyone to expand their campaigns and then what you saw were marketers saying, “OK, well I’m going to stabilize the number of key words I’m betting on, maybe just pick the absolute best ones or at least stabilize my growth.” Instead I’m either going to, as you can see here, improve my big management, which is again being more careful about which terms you’re really spending the money on, but where I’m really seeing the growth here was up at the very top where it says, “Improving the site’s conversion efficiency.”
What that means really is saying, “Have I fixed my landing page? Where are these clicks going to? Do I actually know they’re working? Have I A-B tested it? Have I multivariate tested it? Have I made sure the darn thing is doing the best job it can?” I think this shows you that search marketing, a lot of investment we may not see in a chart or two before we were seeing all that money going toward SEO and PPC actually search marketing is also driving a heck of a lot of money in the budget toward site improvements and toward sometimes hiring agencies or just using different little ASP firms that can fling up a landing page for you if you have to get in line at the Web development at your company and maybe there’s no time to improve your landing page using alternate sources for landing pages while you’re waiting. And we’ve seen some real growth there in everything. You know, we’re seeing, a few years ago nobody was multivariate testing, now everybody seems to be, everyone I can think of is doing at least A-B testing. If you’re going to test it’s going to be on that traffic page that you paid for. I think there’s some really interesting stuff going on there.
Stefan, any thoughts on this chart?
Stefan Tornquist: What jumped out at me at first was the degree to which people are not looking to shift budget to any search. Now that can mean a lot of different things. In the business to business world people are definitely very interested in vertical search, but in the B-to-C world, that’s really how much do you spend on experiment with the second tier?
Anne Holland: And the second tier being –
Stefan Tornquist: Well, in our studies we’ve used a 10% cutoff. So it’s the search engine –
Anne Holland: So, 10% of total search usership?
Stefan Tornquist: 10% of search market share meaning did anyone use the search engine in the course of the month? And so for us the top tier is, of course, Google, Yahoo, MSN, and then Ask and AOL also fit into that category. For us, everything else is in the second tier, but you’ll find different definitions and so what’s the opportunity there? We dug into that quite a bit this year, and the number one response is these sites don’t have the kind of traffic or the quality of traffic that makes it worth dealing with yet another whole set of learning tools and all of the complications that come.
Anne Holland: It’s just a pain conducting a campaign.
Stefan Tornquist: That’s exactly right, and if you’re not seeing the kind of traffic you want and you don’t see the kind of conversion on that traffic, you’re not going to continue.
Anne Holland: It may be good clicks, but it’s maybe only 500 of them, so why bother?
Stefan Tornquist: That may be the case. Now we did try to focus on some of those marketers who had had positive experiences because you know this is an opportunity here. Remember, the cost per click is a fraction of the price of the main, of the big guys. So if there’s an opportunity here, we really want to hear about it, and we did get some positive response typically from very sophisticated search marketers, some of whom had almost exhausted the opportunities in the top tier for their –
Anne Holland: They get that ceiling; they expanded as far as they could.
Stefan Tornquist: Right, and –
Anne Holland: And they were forced to grow, they had to go elsewhere.
Stefan Tornquist: Right, and that means, of course, that they have budget to do it, but there is also a group of careful testers who have found good fits between certain products, certain categories and specific second-tier products. So if you’re in a really margin tight business where the difference between paying five cents per click and $3 per click would really make a difference, then it is worth experimenting.
Anne Holland: That’s pretty cool. I also have heard from some ecommerce sites and some lead-generation sites that they are on the lookout for affiliate marketers who specialize in second-tier search, and we did interview one of those affiliate marketers earlier this year, and he said it was actually a lot of fun for him because there were so few people who specialize in this that it’s an open field for someone who wants to be an expert, you can go in there and do very well, but he also said, he’s like, this is his whole business. It is such a complex thing in and of itself and he said his biggest challenge was writing software to manage all the campaigns on the different second-tier sites because … it’s just a pain to manage the campaigns, it’s an administrative thing.
The other thing was that certainly not all, but some of the second-tier sites are not what one would call -- I don’t want to say dishonest, how about they’re less than scrupulous, shall we say -- when representing what their traffic is, where it came from, what sort of demographic they’re reaching. I mean if you go to ad:tech, and a lot of us are in a few weeks, you’re going to see 50 booths from different people who all claim to be search engine networks. “Oh, we’ve got, you know, a gazillion, hundred million searches per month.” And you’re going to be like, “I’ve never heard of any of these people, they’re all second-tier guys.” How do you know which ones are good or not? And he said, “You really have to cautiously test them.” It’s a lot of work, testing, determining what they’re good at and then moving on, which is, again, why some of the bigger lead-generation marketers and the larger ecommerce sites are actively recruiting those what I’m calling tier B, or second-tier affiliates, people who specialize in nothing but that and say, “OK, look, you represent my brand or you can even do my trademark search on those tiers, just don’t take the ones I’m good at,” which is kind of a neat thing to do.
Why don’t we move on to the next chart. This chart, loyalty data by search engine, this one’s a little bit confusing. Now, Stefan, when you explained this chart to me it, to the initial eye it looks like Ask – which used to be AskJeeves; now it’s Ask -- has 22% of the traffic. That’s not actually the case. What is this chart measuring?
Stefan Tornquist: What this chart looks at is if you’re a -- let’s take Ask as an example -- if you’re an Ask user meaning if you’ve used the Web site for a search in the last month, how much of your searching do you do with Ask?
Anne Holland: So if you do 10 more searches in the following 30 days, what percent of those 10 searches will be on Ask as opposed to a different search engine?
Stefan Tornquist: If you did 10 searches in the course of the month and you used Ask at all, you probably used Ask for two searches. Does that make it clear enough? I’ve had the same issue with this data, but it’s really interesting because we also took this data from Compete and we looked at a specific kind of search, and that was ecommerce, and the way we did that was by creating a group of 500 brand and purchase related terms and we compared that with the loyalty data and what we found is that in that specific kind of search loyalty is even higher. So if you see Google here at about 75%, meaning that if you’re a Google user you use them about three in four times. If you’re a Google user, chances are you use it more than 80% of the time to buy things.
Anne Holland: In your ecommerce searches that month, you would’ve used Google, but in different kinds of searches that same month, you might have gone to another search engine?
Stefan Tornquist: That’s true. Different types of search garner different types of loyalty for their sites –
Anne Holland: OK, so where is Ask winning in terms of types of searches? Is there a particular kind of search people tend to do there as opposed to somewhere else?
Stefan Tornquist: We do have some preliminary data, not from this study, but from another that the places where MSN and Ask are really doing well right now are in local search -- and I’ve got to emphasize this is preliminary data. Both of these sites have made a tremendous investment in their overall search product and specifically in their local search product, and users do seem to be responding and marketers are seeing some good returns. Again, this is a very preliminary stage for this data.
Anne Holland: Now, we do have a neat chart that I’ll actually be releasing in my blog, which is also from the Benchmark Guide and I’ll be releasing it in a week or two, and that is as you may know if you watch TV, Ask has been advertising like crazy. It’s been going on these advertising jags about every six months and just ratcheting up all over TV. What we’re seeing here is that you can see a little blip, but loyalty doesn’t seem to be changing that much. I mean here they are advertising like crazy every six months and spending millions, I’m assuming, and we’re not seeing loyalty change or shift. Do we see traffic shift?
Stefan Tornquist: Well, we do, but it’s temporary. What was interesting about the Ask data was that you can just look at the chart and you can see that there’s a direct correlation between ad spend and the number of people that come and try the search engine. OK, advertising works, we knew that. But the question is do they stick around after giving it a shot? And while it does seem that they get included, like once you’ve used Ask once, it’s then included in the cornucopia of search engines you may use in the future, it doesn’t necessarily become your favorite search engine. The fact that the traffic tends to drop when the ad campaigns do when they drop their ad spend suggests that basically people are pretty loyal to Google and Yahoo! and it takes an awful lot to move them.
MSN is an interesting case because they’re going to be able in the next 12 months to really focus a lot of different kinds of resources because they’ve got this wide suite of products that are all emphasizing the Windows Live experience. It will be really fascinating to see whether they’re able, not just to move that search share, but to keep it and keep it growing.
Anne Holland: It’ll be very interesting. I mean can Google’s back be broken? What we’re seeing here is it’s just getting stronger and stronger.
Stefan Tornquist: It’s not happening this year.
Anne Holland: Let’s move on, Google’s triangle, opportunity number three. Here we are at the eyetracking results which for me are always the most fun part of any study partly because they’re colorful, I’ll admit, I come from the creative end of marketing and I’m attracted to things like this. This is our second annual study into eyetracking, seeing how search users use search engines, and there were several different findings from the study this year and this is just one of them.
Stefan Tornquist: Right, we looked at all of the major, the top tier -- we did everyone in the top tier except AOL. We are very curious how changes on Ask might have made a difference, etc., but one of the things that we’ve found was in the last year you’ve seen the advent of what’s called the One Box or Google’s site search responses and if you can get in here you’re really doing well because user attention streaks to this location. It’s sort of the perfect confluence of natural results, or what are perceived as natural results, combined with some kind of editorial element. So people seem to feel like, “Wow, if it’s up there, it means that it’s kind of got Google’s stamp of approval. It’s really relevant.”
Anne Holland: And it’s a fun little checklist. People like to have their eye travel down a little list like that.
Stefan Tornquist: Right, it looks a little bit different than the content around it, but people’s attention really goes there.
Anne Holland: OK, so this is, this appears if Google has a One Box on a results page, it doesn’t always, but it would appear as part of the very first natural listing underneath the paid ads on that page. And that’s where it appears. A lot of people don’t know its name is called the One Box.
Now, Stefan when you’ve talked to search agencies and search engine optimization firms about how do you optimize your site, what do you do tactically to make sure that if you’re in the number one listing you’re going to get that One Box, all that One Box goodness, all those lovely juicy links? Are there any answers?
Stefan Tornquist: Well there’s a little bit out there, especially if you look to the bottom right-hand side where –
Anne Holland: The History Channel.
Stefan Tornquist: - we see the History Channel and you see those page selections, if you’re routinely showing up in the top ranking of Google, then you can play around with your site construction and you should be able to show these individual site pages assuming that all of those pages are getting sufficient traffic for Google to consider them relevant. They haven’t released a whole lot of information about how that algorithm works and we haven’t heard a whole lot from the search marketers themselves. Now when it comes to local, there seems to be an association with how your links are set up for Google local and mapping.
Anne Holland: Right, you have to submit yourself to be mapped by Google Local, which I believe you can either send them a fee or it’s actually like a mail-in form, insane but true. And so if you do have local organizations, in other words, you’ve got, retail outlets or offices in different areas, you absolutely, absolutely want to get those into Google Local, and I’m not saying that because people go to Google Local -- it’s a very tiny portion of overall Google traffic. I’m saying that, and you know, sure, what the heck, get that traffic, but I’m saying is that because your chances of showing up in the One Box are so much higher and if that puppy’s feeding the One Box, you want to be there.
Stefan Tornquist: One interesting thing that we’ve seen over the last four or five months is that these product search results as we see in the middle of the screen, these are harder and harder to find and a couple of the search pros I talked to say this has something to do with them taking down Froogle -- Google seems to be restructuring the way that product searches are conducted and presented. I would keep my eyes open and not really worry too much about optimizing for the product search results of today because it looks like something new is right around -- well, it’s coming down the pike. Whether it’s right around the corner or not is –
Anne Holland: Well, gee, holiday season 2006.
Stefan Tornquist: Yes.
Anne Holland: Now of course the top one, news results, that obviously helps if you can get some PR out there, always a happy, happy thing, and we’ll talk a little bit more about search engines and PR in a little bit. One other thing to bear in mind if you do, just as the History Channel does, in the bottom right corner of the screen, have yourself listed with little site sections there, be wary of the site revamp. Site revamps are dangerous for all manners of reasons, nevertheless we all go through them sooner or later, but if you’ve got the number one listing on something and Google’s One Boxed you and you do a significant site revamp that could really affect your One Box. I was actually to one of the top 20 ecommerce sites in the America this week privately, I was talking to their search marketing team and they were like, oh, God, you know, they thought about the One Box after the revamp, the week after the revamp somebody went oh, the One Box and now they’re a little concerned about what’s going to go on there. They’re a big enough site, I think they’ll be find long term, but it definitely is going to cause a blip, so I think that’s worth doing.
Why don’t we move on to opportunity number four. We’re on page 11 of our presentation: search to promote offline. This was really surprising data to me, Stefan. Now this is data from a Yahoo! study so, obviously, I mean Yahoo! is in the business of selling search advertising so one would assume that they’re going to end up with a study showing that search advertising works. I mean I’m taking the results with a big grain of salt. How true are they?
Stefan Tornquist: To be fair, they surveyed over 4,300 people; it’s an online panel –
Anne Holland: OK, that’s a lot of people.
Stefan Tornquist: They certainly did it the right way.
Anne Holland: OK. So what are these results? Can you talk to us about them?
Stefan Tornquist: To me, these just stress the fact that search is becoming part of our daily lives. That’s not tremendously surprising. I think we’re going to see people who are currently defined as heavy search users, those folks who are conducting 10-plus searches a day and typically buy more online. I think those people are probably the model for everybody just down the pike a little bit. What we see here is that the number of people for whom that old chestnut of, like, “well, you know, I need to see and touch things,” that’s less than 40% of the folks that responded, and a significantly larger percentage say, “I use the Internet for research, but then I end up going offline.” Of course that’s a really difficult thing to measure. As I mentioned earlier, the comScore study suggested that 80% of the impact of search never got measured because things are happening out there in the real world where it’s really tricky to measure things. When you talk about some of the tactics that you can use to measure the offline impact, none of them are really perfect. Some of them are couponing, Web specials, a few sites have started to associate future Web actions like looking for a … let’s say you do a search for an MP3 player. You don’t buy on the Web site, but then later you turn up as a customer because you’re going and looking at the online FAQ or you’re coming to sign up for the warranty or look at the product manual, or you can survey people in the locations. None of these are really bonafide 100% accurate methods, and they’re all fairly labor intensive.
Anne Holland: One thing this also shows is that a lot of people, certainly including myself, wind up using search engines the way you want to use your print telephone directory. I might have in the past gone to my print telephone directory to look up the name of the local veterinary clinic and then call over there and say, “Gee are you guys open until 6?” Now, I’m just going to Google it, and I’m going to go OK, where are they? What was that street address for them? OK, let me go find them. Or if I’m traveling and I need to get a doctor, I will probably Google wherever I am on the road as opposed to, I don’t know where I’d call, and maybe pick out a local doctor or a local walk-in clinic. I actually just did that last week.
So people, if they’re traveling, if they’re vacationing, if they’re a new mover -- and, of course, we all know new moving households are high on the consumer expenditure list -- are probably Googling local services. Google may be the new Welcome Wagon or, you know, obviously Yahoo!, MSN, the rest of them. The search engines are to some degree being used as local contact, but it’s nothing measured. I mean how often did a walk-in clinic that you’ve ever been to ask you where you heard about them. They’re too busy taking down your insurance ID. I think you’re right, this stuff isn’t being measured and there’s huge opportunity and they need to start marketing on it.
Why don’t we move on to opportunity number five. Now this is actually one of the most sexy new opportunities last year, and as we see PR and search, it has exploded in the past 12 months. Stefan, can you talk about this? What is the difference between this and 2005?
Stefan Tornquist: If you look at the bottom of the chart, you see the dramatic difference that people are conscious of the ability of press releases to go far beyond their traditional function. Obviously, the number one way of optimizing the release: putting a link in it. You know, it’s amazing when you go out there and find links online or find releases online that don’t actually contain a link.
Anne Holland: I’m dumbfounded. I read a lot of press releases, I mean professionally. I’ll be reading it and there’ll be no hot link, or it’ll be like way at the bottom.
Stefan Tornquist: In terms of some of these other tools, these are all coming along very slowly with the exception of writing releases emphasizing key words in the headline and the body text. It’s just like Web sites themselves, it’s about a lot more than metatags, and we think it’s great that the good wire services -- Market Wire, Business Wire-- they’re offering optimization now.
Anne Holland: That’s right. Last year it really was all about PR Web. This year you’re seeing Market Wire, Business Wire, PR Newswire, they’re all coming online offering some form of optimization.
Stefan Tornquist: Yes, and it’s important to note exactly what they do versus what you need to do in-house. They’re probably not going to rewrite your press release, although I know some of them do offer that feature. So you really need to be aware that just checking off the optimization box, that doesn’t necessarily cover all of your bases.
One of the things that I really wanted to emphasize for folks who are experimenting with this that’s fairly straightforward is producing multiple versions of releases emphasizing different keywords. We’ll look at some results in a moment from our own study that really show you why you ought to.
Anne Holland: So if you’ve got one piece of news, you should do more than one press release about that same piece of news?
Stefan Tornquist: Absolutely and emphasizing different keyword terms, and you can look at it in a couple of different ways. One, are there different constituencies who are going to be looking for different things? You know, professional journalists on the one hand, maybe bloggers on another, customers perhaps most importantly. And then even within those customers, what are the terms that people might be using? Are they looking for –
Anne Holland: Right, kids versus their parents.
Stefan Tornquist: Yes, airline deals versus cheap tickets or what have you. For the relative cost of distributing these press releases, if done well the ROI can be tremendous.
Anne Holland: Millions in fact.
Stefan Tornquist: And there are case studies to prove it.
Anne Holland: There certainly are.
Stefan Tornquist: Should we go to the next slide?
Anne Holland: I actually want to say one thing, which is from personal experience, but I’ve also talked to a lot of marketers about it, one of the things that some of the wire services are now offering is pay a little extra or check this extra box in your checkout process to get the release distributed and we will stick you into RSS feeds and, of course, RSS is such a sexy term these days. What we’ve actually found is that, while we believe very strongly based on data that search engine optimization of your press release is a very smart thing including the hotlink, including all of the metatags, including everything Stefan listed in this chart, the RSS feed may not be as useful, so at the very least don’t pay extra for it. What I am seeing is that the RSS feed tends just to be feeding those, well, I guess you’d call them splogs, they’re the junk blogs that are just built to sell Google ad words and they’re not actually useful sites. They just sit there and suck up RSS press release feeds and spit them back out. It’s not a site anyone goes to. I have not seen a lot of value from RSS feed press releases, I’m not saying I don’t see value from RSS in other ways.
Why don’t we move on? Yes, this next slide, this shows you actual real-life results from two different press releases that we did about the same time.
Stefan Tornquist: These are in many respects the same release and, therefore, ultimately they’re promoting the same product and some of the same findings, but as you can see from the highlighted words the top example, this is the number one ranking on Google, on Google News at that moment for podcasts, plural, and the lower one is for lead generation we were the number one ranking in Yahoo! Very effective. One interesting note is that it’s important that we used “podcasts” and not “podcast” because even though a lot more people are searching for podcast, there’s an awful lot more content for it as well. So the happy accident was this discovery that something as simple as a plural, which we didn’t expect Google to care about. We expected it to be sort of broad match, and lo and behold because it was podcasts and not podcast even though we were only looking at however results for that moment, was 2,300 or something, that’s much smaller than we were looking at podcast, our ranking’s number one and that makes a much bigger difference.
Anne Holland: And the nice thing is because we had such a high ranking there, you can see directly under that in the little Google News One Box that five, seven, 12 other actual news sources had also picked up on this release and written their own stories about it. So we got a lot of play in the press, which we hadn’t planned on; we hadn’t realized that we’d actually get reporters calling us. It was kind of fun. We were just looking for sales.
Stefan Tornquist: What was really interesting also in looking at the results with hindsight was the role that blogs played. Bloggers were very often the first ones to pick up the news and that led to some of the traditional press picking it up, which led to more bloggers, and the blogger-driven traffic converted better than the traditional sources.
Anne Holland: It’s fascinating. Now this all points out a key thing. I’ve talked to a lot of marketers about this, and they say that much like search engine optimization the confluence of press releases and search marketing is causing a political battle in their organizations. What I’m hearing is, of course, press releases would be the PR department or the marketing communications department or perhaps even outside PR firm are in charge of press releases and you don’t want to go behind their back and that’s their job, that’s their territory. Here you are in the marketing department and you may be in the direct response marketing department or the Web marketing department, you’re trying to get, you know you want to wield the press release as a marketing tool. You’re going out there and saying, “I’m going to do this because I want to get clickthroughs from actual end consumers that are surfing Google and Yahoo! News and MSN News.” Whereas the PR department is wielding the press release as a media relations tool. They’re thinking, “Hey, I’m going to get USA Today to write a story about us.” It’s such a different goal and yet sometimes it’s the same press release or sometimes it’s just you’re not allowed to do a press release as a marketer because that’s the PR department, period, and they’re got their little ivory PR wall up around it.
This is definitely going to be something where there needs to be a lot more communication and some bridges built between PR and marketing, and it needs to be recognized that PR needs to train marketing in what they expect out of press release and I’m sure they’ve got formal brand rules around that and they want some communication around that and they need and want to communicate, have some control over the scheduling, but marketing also needs to train PR in what’s a keyword? What’s a landing page? How do you do a URL? How do you create a landing page? How do you track these things?
These are two such different skill sets, and both departments have to get together to cooperate. In the case, most famously, of course, from Southwest Airlines, when both departments cooperated, they ended up making millions in tickets sales from a series of three press releases. So it is possible. Not only that, but it’s probable if you know what you’re doing. It is absolutely, even though it’s been exploding, still an unexplored opportunity for a lot of companies. But it’s going to require internal work, and I hope that some of the data we’ve presented today can help you as you’re budgeting, as you’re going in and you’re having those committee meetings and you’re trying to bring these ideas to the table and saying, “Look, let’s move our organization toward this for the future.”
That’s our goal, and if there’s some data point or some fact or something that you’re like, you know, I’m dying to pitch this to my CEO … or pitch this if you’re an agency, I’m dying to pitch it to clients, but I need the data to be able to prove it can work or to be able to present this idea, let us know. We may have the data in the report; we can let you know if it’s in the report; we can let you know if we know somebody else who has the data. That’s our goal is to be a service to you as a practical service for your marketing data needs.
[End of audio]