Search Engine Marketing Finally Getting Respect

To some degree, search engine marketing has been like the Rodney Dangerfield of online advertising -- it's gotten no respect. Or, at least it has gotten no respect in relation to the time and effort analyst firms have put into understanding it compared to banner advertising. The good news is, that's all about to change.

Many of my readers well understand how search engine listings can bring in qualified traffic. You know this because you've experienced it first hand. Given this, you might wonder how the value of search engines couldn't be recognized by others.

The reason is that, until recently, there was no way to quantify the actual value of search engine marketing. Go back a year ago, look at the public financials of any major search engine, and you won't find a break out for paid listings. That's because paid listings simply didn't exist in any major way, shape or form. Instead, it was banner advertising and retailer deals that dominated earnings. In addition, few of the publicly-owned major players were "pure" search companies, so it was hard to discover search-specific revenues.

Of course, a year ago, there was plenty of revenue being generated by search engine marketing. Indeed, there's been plenty earned for several years. However, the bulk of it has been generated by third-party search engine optimization firms, I would wager. Since search engine marketers lack a trade organization -- and since most search engine marketing firms are private companies -- measuring the value of what they generate is a near impossible task. Without those figures, it's hard to get some people to recognize that search engine marketing is and has been a booming business.

A case in point is the recent report from Jupiter Media Metrix affirming that paid search placement is now a successful adverting model. This is the first time I can recall seeing a major research firm declare that search engine listings are something that companies need to consider as part of their advertising budget. If a similar report slipped by me, let me know, and I'll happily apologize. I certainly don't recall one.

We have had stats from surveys showing that people use search engines frequently or that many web marketers understand that search engine optimization is important. Indeed, my Search Engine Index page listed below provides a compilation of such statistics. But a major analyst firm saying to marketers that they need to think about search engine listings? Nada. No respect.

Jupiter Media Metrix is the first firm to do this, and while I'm absolutely pleased to see them come to this conclusion, all I can think is, "Duh, what took you so long?" How could they not realize there was value in them thar search listings a year ago, two years ago, three years ago -- and how much money might advertisers who heeded their calls have saved if they hadn't bought overpriced banners and instead put more thought into search engine positioning.

Let's not single out Jupiter Media Metrix for its long lapse. Forrester released a report back in March on driving traffic to a web site. Search engines weren't mentioned, not as a major way marketers were getting traffic nor as a recommendation they should investigate -- despite the fact that the same report noted search engines as the top resource consumers used to find web sites.

The demand has always been there, and it shouldn't have been that hard to see. For instance, back in 1999, INT Media first suggested that I work with them to do a day-long conference about search engine marketing. I, a veteran of trying to squeeze everything about search engines into a one hour slot at other conferences, was thrilled at the luxury of so much time.

INT Media figured 100 people would be a success. We ended up sold out with over 300 attendees and have been packed at every conference since (and honestly, if you are thinking of coming to the next one in San Francisco this month, you'd better preregister. See the link below).

Now what's caused Jupiter Media Metrix -- and soon everyone else -- to finally see the light? GoTo. You see, while dotcoms continue to crash all around, GoTo is a major player showing growth, due to its listing services.

In addition, GoTo competitor FindWhat has posted a small profit. Meanwhile, LookSmart is running around desperately trying to get analysts to understand that it, too, is in the same marketplace as GoTo (which is it -- so analysts, do finally get it). LookSmart continually points out that its listing services are growing. In fact, last quarter they surpassed more traditional advertising as the company's major revenue stream.

All these companies are public, and they either derive most of their revenues from search listings or break out listings as an easily identified component of revenues. This means that analyst firms and others can finally attached a value to at least a portion of the search engine marketing industry. Moreover, that value is showing sustained growth at a time when other advertising options are declining. This is a big red flag in the face of anyone that there's a product in demand.

Still, the rosy financials are no excuse that the value of search engine listings wasn't seen earlier. Sure, it was easier to measure the spend on banners. And sure, purchasing banner space a year or two ago was much easier than delving into the morass of getting listed with search engines. And sure, getting good search engine placement is still a complicated matter, even with the new programs offered by people like GoTo.

So what? Who said things that are worthwhile are necessarily easy? It would be nice if they were, and hopefully search engine marketing will get easier. However, complicated or not, it cannot be ignored.

If you are marketing a web site, search engines are not an afterthought to your banner ad spend. They are not even something you can consider after you build your web site. Search engines are something you think about from the very beginning, and they remain a priority for as long as you plan to market online. If you fail to understand this, you're going to spend more money than necessary to get traffic and still may fail in your efforts. Meanwhile, those who do understand search engine listings will be enjoying success.

Please -- please, please, please -- don't assume that it's just about buying your way in, either. Yes, you should absolutely investigate the paid participation programs available to you from places like GoTo, Inktomi, LookSmart and even Yahoo. They can certainly make your life easier and generally offer good value for your money. However, if you get your own house in order, you can probably tap into a lot of free "natural" traffic from search engines, as well. Build search engine friendly sites or budget for advice and assistance from third parties who can help you, as well as budget money to spend with the search engines themselves.

Paid Search Engines Picking Up Slack For Depressed Online Ad Market
Jupiter Media Metrix, July 30, 2001

This report points out that advertisers, especially merchants, need to consider paid listings as part of their online advertising mix. It points out that a March 2001 study by Jupiter and NPD found that when users are looking for products online, they are far more likely to type the product name into a search engine (28 percent) rather than go into a search engine's "shopping" channel (5 percent) or click on ads (4 percent, but not shown in the release above).

The report makes a terrible mistake in suggesting that paid listings help users find what they want faster than traditional-editorial style results. Paid listings absolutely have value to searchers and can indeed be exactly what they want. However, you can't prove this from Jupiter's data. To support its claim, Jupiter looked at how long people spent within several major search engines during April 2001. The conclusion was that the less time you spend with a particular search engine, the more likely you are finding exactly what you want and thus leaving quickly.

It then showed that average usage of GoTo was only 56 minutes in April, compared to 106 minutes in the search-specific area of Yahoo and 140 minutes in the search-specific area of MSN. Time spent with Ask Jeeves and Google was even higher, and though this was shown in the accompanying table, it was not called out when making this overall conclusion in the report's press release (and a similar one in the actual report itself):

"Visitors to, a paid search engine, spend an average of just 56 minutes on the site, while the average user of Yahoo's proprietary search engine spend 106 minutesan indication, say Jupiter analysts, that consumers are much more satisfied with their findings from paid search engines."

Using time to measure user satisfaction was a flawed approach. For instance, if many users like Google's results, they would be far more likely to go there. As a result, you would expect their usage time with Google to be higher. As it turns out, Google's average usage is higher -- an average of 269 minutes, nearly five times that spent with GoTo.

According to Jupiter's logic, this indicates that GoTo is far more relevant that Google. That flies in the face of Google widely being acknowledged as a relevancy leader in the search space. In fact, had Jupiter substituted Google's name for Yahoo in the quote above, I doubt many journalists -- who love Google -- would have taken this report seriously.

The only way using time would possibly make sense is if you averaged it by the number of search requests. In other words, let's say I do 50 requests per month at GoTo and spend 50 minutes there in total. That means I spend one minute per search, on average. At Google, I do 100 requests per month and spend 75 minutes there, so each search takes an average of 45 seconds.

Now I can more easily compare satisfaction based on time -- though I absolutely shouldn't do this. If I want to measure satisfaction, I should ask users directly. Someone does this, too -- NPD, the company that Jupiter partnered with on its other report. Based on NPD studies in the past, GoTo has indeed done very well, especially in the category of people who find what they are looking for every time. However, Google has also done well and consistently aced the tests last year.

Unfortunately, the satisfaction of individual search engines is only known if they choose to release that -- and most do not, unless they've been top ranked.

NPD Search and Portal Site Study, July 6, 2000

The NPD study I mentioned above with satisfaction information. This data is now about a year old, but I've been promised new summary data and so hope to update the page later this month.

GoTo Branding Study
GoTo, January 2001

In this study of 2,663 people, GoTo had research company NPD test how users interacted with three types of advertising placed in a search environment: listings, banner ads and small, square "tile" ads that ran along the side of the listings. The study found that listings were far more likely to be read and clicked upon than banners or tiles. Listings were also far more likely to produce purchases.

The real interesting twist was that brand awareness was also raised more by listings than banners or tiles. That's not a bad pitch for GoTo to take to large advertisers who are who are still dubious about paid listings and instead looking to raise brand awareness. It also raises the issue of whether GoTo will have to consider cost per impression pricing, in the future. After all, if you are trying to raise brand awareness, you want to appear often and perhaps maybe don't even need to attract clicks. In that case, GoTo's cost per click pricing may not make sense.

Don't discount this study just because it was conducted by GoTo. From the briefing I had, it looks fairly well conducted. I am also planning to take a longer look at it in the future -- just need to find the time!

Driving Customers, Not Just Site Traffic
Forrester, March 28, 2001,1317,12053,FF.html

This report surveyed 50 online marketers about how they drove traffic to their web sites. Banners were the top choice, used by 96 percent of them. They were asked about search engines, even though these don't show up on the chart, Forrester reported to me. This is because only 4 percent said they found search engines the most effective way to reach target customers.

Such low respect for search engine listings by marketers flies in the face of the fact that a different report mentioned in this article showed that search engines were the top way used by consumers to find new web sites.

Forrester also told me that in another report, "Online Advertising Eclipsed," they asked what percentage of total Internet marketing budget will be spent on search engine optimization. For 2000 it was 7.6 percent and for 2005, it actually drops to 6.9 percent. Want to wager that this percentage actually rises, instead?

Popping Up Everywhere and Under Everything
ClickZ, July 31, 2001

Despite some complaints about paid listings, when positioned and labeled appropriately, I think even the worst critics would agree that they can be useful to consumers. In contrast, the growing popularity of pop-up or pop-under ads seem widely considered to be an irritant, as this article explores.

Banners Can Brand, Honestly They Can, Part II
CyberAtlas, July 30, 2001,,5941_804761,00.html

Large skyscraper-style ads are effective in building brand awareness, in comparison to banner advertising. And that's good -- because perhaps it will finally help drive home the fact that advertising is not necessarily about clickthrough. Many good and struggling web sites have the same audiences that advertisers will spend millions to reach in more traditional venues. But on the web, success has been traditionally measured by clickthrough, which is poor for visual ads.

Advertisers spend millions on billboard space, but no one clicks on that. Why should placing visual ads on the web be seen as less valuable? The message is still being delivered, clickthrough or not. In contrast, measuring the effectiveness of search listings by clickthrough makes a lot of sense, because these ads correspond precisely with a person's desire to receive more information on a topic.

The argument that visual ads should be purchased to build brand awareness makes a lot of sense. But also notice how none of the studies mentioned, and done in conjunction with the Interactive Advertising Bureau, examined the effectiveness as paid listing-style ads among the choices tested. Once again, search placement gets overlooked by players in the online advertising world.

Search Engine Index

If it's an interesting search engine-related stat, I try to post it here.

One Net Stock That Isn't Poised to GoTo Zero, July 26, 2001

GoTo saw a 20 percent rise in revenues for the quarter ending on June 30. Revenues hit $62.5 million, nearly triple the amount from last year. Its loss is down to about $3 million, compared to $20 million from a year ago. The company expects to make $1 million in profit next quarter and $19 million in 2002.

Other interesting stats: average click price is now 19 cents, up slightly from last quarter but down from the 23 cents average paid a year ago. Perhaps those advertisers are getting smarter about watching bids or making more use of bid management tools.

All the money coming in isn't without a price. GoTo paid 59 percent of revenues to affiliates and partners such as AOL, though this is down from 67 percent in the first quarter of this year and expected to drop to 55 percent next year. (Affiliates make up 95 percent of GoTo's clicks, by the way). The article provides more details and analysis about these stats and GoTo's fiscal future. Announces Record Second Quarter Results
FindWhat Press Release, July 23, 2001

FindWhat had revenue of $4.4 million this past quarter, up 73 percent over the previous quarter's $2.6 million. In addition, the company even made a small profit of $133,849, compared to a loss of $2.9 million the same time last year. The average clickthrough rate was 13 cents last quarter, compared to 10 cents from the previous quarter and 7 cents for the same time last year. The percentage shared with distribution partners was not listed.

LookSmart Reports Second Quarter 2001 Results
LookSmart Press Release, July 26, 2001

It sounds bad at LookSmart. Revenue is down for the second quarter to $18.9 million, compared to revenues of $27.3 million for the same time last year. Its net loss was $13.6 million, just above a $12.9 million loss for the same quarter. So much for the bad news. On a cash basis, the net loss was only $3.6 million, reduced from $11.1 million in the previous quarter and $8.4 million for the same time last year.

As for revenue, a year ago most of that came from advertising and syndication. Now listings are LookSmart's primary source, and that revenue is on the rise. It was $8.1 million last quarter, up 26 percent from $6.4 million the previous one and $1.6 million for the same time last year. Distribution payments ate 40 percent of that revenue last quarter, but that's less than what GoTo shared for the same time period and down from a peak of 76 percent in the fourth quarter of 2000.

Excite@Home Shakes The Piggy Bank
Forbes, July 27, 2001

Once a search engine, then a portal, now primarily a broadband company that also runs a portal, Excite@Home's not finding the financial picture anywhere as bright as GoTo. The company had a $65 million loss last quarter. Revenues generated by the Excite-branded media properties were $28.6 million, a drop of 62 percent from the same time last year. The company is now looking to get rid of the Excite properties. Meanwhile, though Excite@Home has $183 million in cash, one analyst thinks that will drop to $20 million by the end of the year.

LookSmart Ranked #1 in Network Reach by Jupiter Media Metrix
LookSmart Press Release, July 16, 2001

I've written before that GoTo really isn't a search engine but instead an ad distribution network. LookSmart, with some validity, makes the same claim. Now Jupiter Media Metrix not only backs up that claim but shows LookSmart has having greater reach than GoTo, 76.7 percent to GoTo's 70 percent.

The only problem is, it's hard to trust these numbers -- not just for LookSmart but also for GoTo and perhaps some of the other distribution networks named. For instance, how is the reach derived? Is Jupiter Media Metrix looking at all the sites where LookSmart or GoTo results appear, then looking at all the users it surveys and determining how many of them go to each place where LookSmart or GoTo listings are available?

That's what I imagine, though Jupiter Media Metrix has yet to respond on that question and LookSmart flat out says that providing any clarification would violate the confidentiality of its partners.

OK, so let's assume that I'm right. The thing is, we're talking potential reach, not actual reach. Not everyone who goes to AltaVista will see a LookSmart listing -- in fact, they are harder to find there now than just three months ago. Similarly, not everyone who goes to will see GoTo links, only those who specifically perform searches -- and maybe not always then.

No doubt, both LookSmart and GoTo have very broad reach. That's what makes them valuable advertising resources. But how much is actual reach is uncertain. One good study would be for someone to measure which links produce the most visits, GoTo's or LookSmart's. That would really compare them better as advertising venues.

Search Engine Placement Tips

Key tips to making your site search engine friendly.

Tapping Into Natural Traffic

Fast, simple things that you can and should do as part of your site building process, to ensure you are friendly to search engines. This is only available to Search Engine Watch members, as described at

Jupiter Online Advertising Forum

Part of the reason behind the many reports Jupiter has been issuing on online advertising is to drum up interest in its upcoming conference on advertising. I was very hearted to see that search engine listings are being touched upon for what I believe is the first time, through an address by GoTo.

Search Engine Strategies

Hard to believe, but the Search Engine Strategies conference that I produce now begins its third year of existence, as it comes to San Francisco on August 16 and 17. What originally started out as a one day, single track event is now spread over two days and multiple tracks. If you are involved with search engine marketing -- or thinking you should learn about it -- you'll find plenty of advice here from a wide range of experts and search engine representatives.