Biznology Blog: February 2007

« January 2007 | Main | March 2007 »

February 27, 2007

Will Manufacturers Show Product Reviews?

Last year I talked about the Three R's of Web Marketing, to be real, relevant, and responsive. But the truth is that being real is challenging. Let's take a real-life example—do you expect product manufacturers to post product reviews?

We've all seen customer-posted product ratings and reviews on retail Web sites. Amazon's got 'em. Shopping comparison engines do, too. And customers love to read them.

Some retailers have been reluctant to post ratings and reviews, however. They reason that they want to put their best foot forward—no need to say anything negative about something they sell. After all, why talk someone out of a purchase?

The truth, however, is that reviews talk customers into purchases. In fact, if customers don't get reviews on your site, they'll look elsewhere until they find them.

So, retailers have gradually talked themselves into providing ratings and reviews, because customers can pick between all of their choices and buy something. But manufacturers? Why would they want reviews? They may make only one model of a product—a bad review might cause the customer to buy from a different manufacturer.

It takes real guts for manufacturers to allow reviews to be posted.

Well, take a look at someone with guts: Curt Sasaki, Vice President of .SUN Web Properties at Sun Microsystems. Curt and his team has put product ratings and reviews for all of Sun's products on sun.com—a sign of supreme confidence in what they sell.

Curt told me that what I call being real, Sun calls "openness and transparency," and that it comes down from "the top of the company, Jonathan Schwartz," Sun's CEO. Schwartz is notable as that rare species, a CEO blogger, but providing product ratings and reviews really turns your marketing message over to your customers.

Sun had found that for every bad review they get, they get several good ones. Curt says that the really crazy reviews are "self-correcting"—someone sane comes along to override them soon enough.

Sun's mot stopping, either. They've now started syndicating blog feeds from Technorati alongside the 3700 blogs from Sun employees. Do those blogs from outside Sun always say what Sun wants? Of course not. It's not easy. But it's real. And it's what your customers want from you.

If they trust you are being straight, they will believe more of what you say. And that is the basis of persuading them to buy from you.

Posted by mikemoran at 8:48 PM | Comments (0) | TrackBack

February 26, 2007

Remember Search Marketing is Marketing

I enjoyed my appearance at the ExL Pharma Search Engine Marketing for Pharmaceutical Companies conference today in Princeton, New Jersey. We had a sudden snowstorm yesterday, so you know that the folks who were able to make it really wanted to be there. My pitch, called Remember Search Marketing is Marketing, talked about how you can use search marketing to raise brand awareness and conversions, but my favorite part of the day was listening to Eric Papczun of Performics.

Eric had excellent advice on organic search optimization, but I was most interested in the stats he showed about the pharmaceutical industry:

• 58% of Americans say that online information is their most important source for health information (Pew Research)


• 54 million Americans look for information on pharmaceuticals on the Web (Manhattan Research)


• 80% of Americans (113 million) have researched health information on the Web (Pew Research)

I haven't worked with any pharmaceutical companies on their search marketing programs, so I was interested to see what health searchers do. Not every industry is as large as health care, so you may not have industry-specific statistics for your industry. But if you do, they can go a long way to convincing the boss that search marketing can't be ignored.

Posted by mikemoran at 8:59 PM | Comments (0) | TrackBack

February 25, 2007

What Makes Searchers Click?

We’ve explored what catches searchers' eyes on the search results page. The next step is the click.

What makes searchers click a particular result?

• The page’s title. Searchers are scanning for their keywords in the title of each page in the search result list. But, as we saw earlier, semantic mapping causes searchers to scan for other words that they did not enter. Internet search studies show the title to drive the most clicks—a title which matches the searcher’s concept (both explicit search keywords and those not divulged).


• The page’s snippet. Not far behind in importance is the snippet, the couple of lines of description below the title. Typically the snippet consists of snatches of words from the page’s text with the searcher’s keywords highlighted.


• Depending on your site, other factors may be important, too. Retail sites frequently show pictures and prices in the search results. Auction sites such as eBay show the time remaining before the auction ends. Finding out what causes your searcher to click is important—different kinds of searches require different information on the screen.

But why does a searcher click one result over another? We’ve already seen that, no matter what type of query, searchers tend to click a result that contains the exact query words. In addition, seeing trusted brand names (as well as reviews and comparison information) correlates with Internet searcher clicks. Showing a low price (along with promises of discounts or other offers) enhances clickthrough, especially when the searcher can buy online.

If you're looking for a free search engine for your Web site, one that is easy to try, provides highly relevant results, and that has a configurable user interface you can change to your heart's content, try the IBM OmniFind Yahoo! Edition.

Posted by mikemoran at 10:06 PM | Comments (0) | TrackBack

February 23, 2007

Speedy in Any Language

In the past, we've urged marketers to do it wrong quickly. But how can you make quick content changes when you have to translate into other languages? For global companies, translating their content into other languages can be a time-consuming process that delays the launch of new pages and campaigns. So, exactly how do you speed up your ability to change content when you have to do all that translating at the end?

Well, the first tip is: don’t wait until the end. Get your translators involved early so they don’t object later—German needs a lot more space on the screen than English, for example.

To increase speed and to control costs, many Business-to-Business marketers use a tiered globalization approach—each country site, for example, is assigned to a tier that defines how much content is translated. So, the countries that garner your highest sales might be fully translated, the second tier might have only your top products translated, while the lowest tier has only the home page and a few high-level navigation pages in native language (with the rest possibly in English).

When you translate any page, make sure that you re-engage your search marketing experts to use the popular keywords in the new language. Otherwise you risk getting a semantically correct translation, but you aren’t using the words that most people will search for.

Most multilingual sites use on-staff or contracted professionals to do their translation, but a few pioneers have asked volunteers to do it—Google is probably the most famous example. Internet book cataloguer Library Thing offers no money for volunteer translators, but displays an honor roll of those who have translated the most phrases. You need a rabid following to work this way, and you probably can’t have a great deal of content. If you go the volunteer route, you may need to control what types of people you allow to volunteer. For example, a pharmaceutical site might prefer to be translated by medical professionals rather than by hypochondriacs.

Whatever your approach, don't let a little thing like translation slow you down. If you put some thought into it, you can be speedy in any language.

Posted by mikemoran at 11:10 PM | Comments (0) | TrackBack

February 22, 2007

How Can I Check My Navigation?

If you’re like most companies, you spend a lot of time obsessing over your home page, but maybe not enough time on your destination pages. As a proponent of doing things wrong quickly, I don't recommend spending hours on each page trying to make it perfect. So, instead, what’s a quick way to check out the information architecture of those pages?

Use Keith Instone’s Navigation Stress test—Keith is the lead information architect for IBM’s Web team. For those of you who are lazy clickers, I have paraphrased the test below.

To start the test, pick any destination page on your site, and print it in black and white all by itself (with nothing else on the page, not even the URL). Then pretend that this is your landing page for your first visit to the Web site, and answer the following questions:

• What’s this page about?


• What does each set of links represent?


• What site is this?


• What are the major sections of this site?


• Which section is this page within?


• What’s one level “up” from this page?


• How do I get to the top page for this section of the site?


• How do I get to the site’s home page?


• How could I get back here from the home page?

Don’t be upset if your pages usually fail this test. Keith reports that he’s been doing these tests since 1997 and most pages fail. The questions are designed to test the three major questions customers have when they land on your site: where they are, what is here, and where they can go.

When your pages pass the Navigation Stress test, your customers experience a lot less stress.

Posted by mikemoran at 3:02 PM | Comments (1) | TrackBack

February 21, 2007

Is SoloSEO for you?

I've been on a quest recently to find some new organic search marketing tools that do a little bit more than what you're used to. I recently wrote about how Fortune Interactive is innovating with metrics that might help you take on personalized search. Today, I want to look at SoloSEO, a relatively recent entry into the tools market that offers an integrated set of tools pitched to lone practitioners.

I recently spoke with Michael Jensen, co-founder of SoloSEO, who showed me an impressive and easy-to-use set of tools that work together to assist individual search marketers make the most of organic search.

As you might expect, SoloSEO has the usual tools you've come to expect, such as a keyword analyzer, but I was surprised to see that included in the monthly price was access to Wordtracker keyword data even if you are not currently a Wordtracker customer. SoloSEO also offers the ability to scan your site for keywords (the way Google does for its AdWords advertisers).

SoloSEO allows you check your rankings on the major search engines, as you would expect, but, starting today, it also provides reports on how your competitors are doing (as announced in the SoloSEO blog) which include ranking and backlink comparisons and more.

SoloSEO can create a simple Sitemap to help index your site, but it also uses your Sitemap to run reports against your content. You can schedule reports to run automatically, too.

It also offers a handy to-do list that allows you to keep track of your top search marketing tasks—even generating some based on customizable checklists. So, if you always do five tasks every time you create a new campaign, SoloSEO can add those checklist tasks to your to-do list on command.

SoloSEO doesn't have the best tools in each category, but as Michael explained to me, "Free tools just don't get a lot of support and they aren't integrated." I've long been a believer in using free search marketing tools, but I have often chafed at having to constantly switch between them and having to re-key data from one tool to another. And having to manually run the tools to check what I am looking for every time.

I have to admit that free tools are nice, but that I don't check my search marketing as often as I would if it was easy. SoloSEO is just $29 a month for up to five domains, and offers a two-week trial. I've only seen the demo so far, but I am considering signing up for the free trial when I get a few minutes to breathe. For the one-person search marketing operation, it seems like something to check out.

Posted by mikemoran at 3:18 PM | Comments (0) | TrackBack

February 20, 2007

When Search Results Meet the Eye

What are searchers thinking when they stare down a list of search results? Actually, the first thing to keep in mind is that searchers don’t stare at search results. They don’t even read them. They scan them. They quickly skim the page hunting for the word they typed into the search box, along with a few other words they’re thinking but didn’t bother typing.

And they do it quickly. Gord Hotchkiss, CEO of Enquiro says that searchers spend just seven seconds looking at an Internet search results page, and Cornell University says it is less than six. Regardless of the exact amount of time spent, searchers are sizing up the results rapidly.

To see what searchers are seeing, Enquiro conducted an eye tracking study that recorded eye movements (and clicks) and created graphs that corresponded to where searchers looked on the screen and what they "fixated" on. This heat map from Enquiro shows where on the Google results page searchers looked the most. (Searchers using Web site search probably behave in similar ways.) The heat map shows us intense interest in the upper left part of the search results—searchers scan starting at the #1 result to see their search words highlighted in the titles and work their way down the page.

Now, it’s not exactly earth-shattering to hear that searchers look at the top search results, but if you thought that folks were looking at the ads on the right of the screen or the number of search results—well, they don’t look at those areas as much. Where searchers look is important, because people generally don’t click where they don’t look—although some experts cast doubt on eye tracking studies because human eyesight provides considerable peripheral vision. (Some people actually do click scroll bars and other areas without looking at them.) Regardless, where searchers look gives us great clues as to what has their attention and, thus, could attract their clicks.

Moreover, only about 60 percent of Internet searchers scroll below "the fold" (the part of the page that is off the bottom of the screen when first shown), according to Greg Edwards, the CTO of Eyetools, which performed the study with Enquiro. And when Internet searchers return to a search results page, Greg said, they tend to look further down the search results page for more results, with results above the fold still getting a big edge in clicks.

One reason that searchers spend so little time on the search results page is that they click the first thing that looks like the right answer. They are not looking for the best answer—just the first decent answer. This behavior is consistent with studies that show that people often choose information that is easier to get even when it is of lower quality (a basic principle of foraging behavior also.)

But what are they looking for? The words in their query, yes, but that’s not all. Gord describes a particular searcher behavior called “semantic mapping” whereby searchers associate many possible search terms with a concept but enter only one or two words into the search box. For example, a searcher looking for a digital camera thinks about many words (such as reviews, megapixel, Nikon, Canon, Kodak, easy-to-use, reviews, and testimonials) but ends up typing in “digital camera.” But those other words are not forgotten. The searcher scans for those words in the search results also, not just the words that were typed. In this way, a #2 result that contains more occurrences of the words in a searcher's semantic map could attract more clicks than the #1 result.

Most Internet searchers look only at the first page of search results—the top of the first page at that. But Jarvis Mak, Director at Nielsen/NetRatings, notes that as searchers do more and more searches within a session they are far more likely to look at a second page of search results, perhaps even a third page. These are the distinct minority of searches, but it does show that when searchers are intently searching for something they use more queries and view more results pages.

Within each search result, Internet searchers spend 43 percent of the time viewing the snippet (the multi-line description under the title), followed by 30 percent reviewing the title itself, according to that same Cornell study. Why is that important? Good titles and descriptions are the keys to getting searchers to click.

The more you know about searcher behavior, the more you can do to improve your organic search marketing and your Web site search.

Posted by mikemoran at 4:38 PM | Comments (5) | TrackBack

February 16, 2007

Growing Better Customers

Yesterday, we talked about targeting your best customers by identifying those who score highest in recency, frequency, and monetary (RFM) measures. But what about those customers who do not have high RFM ratings? Most are simply not in your sweet spot—they are not the best matches for what you do. A few were once highly rated in RFM, but are no longer your best customers—at least right now. What should you do?

Your job as a marketer is to try to get them back into that “top customer” segment.

How you do that depends on what's wrong. You can use those same metrics to see how your customers have slipped out of your top segment. By analyzing those numbers, you can decide how to woo them back:

• “Not so Recency.” Customers who’ve had high RFM ratings who have suddenly stopped visiting your Web site (lowering their Recency rating) may now be frequenting the competition. Maybe you should send them an e-mail with an offer.


• “A bit lower Frequency.” Customers whose Frequency rating suddenly drops may no longer need the same amount of your products. Perhaps you may want to offer them a discount for referring a friend. Or expose them to other products you sell.


• “A little less Monetary.” Customers whose order sizes plunge (lowering their Monetary rating), may have found alternatives for some of your products. Perhaps offering them discounts based on raising their order size might bring them back into the top segment.

You’ll find that providing different offers based on RFM ratings will not only increase your conversions, but also raise your average order size and lower your promotional discount costs (because you can selectively offer discounts only in those situations where it raises your overall profits). You may also find that different marketing approaches (e-mail vs. search marketing) may prove more effective for your best customers, or that different areas of your Web sites are best at attracting new high-RFM customers.

Whatever you find, analyzing your RFM ratings and acting accordingly can only raise your sales.

Posted by mikemoran at 8:55 PM | Comments (0) | TrackBack

February 15, 2007

How Do You Target Your Best Customers?

So much otherwise good marketing is misdirected at the wrong customers. "Free shipping" is a great offer for someone almost ready to buy who needs just one more reason to do it now, but it won’t move someone who isn’t ready yet. Similarly, many of your offers will work wonders with your best customers, but may appear to be poor campaigns when unleashed on a larger target market segment—your typical customers. The question is how to tell who your best customers are.

Does it sound like conventional wisdom that you should direct most of your marketing efforts at your best customers? Or that the customers most likely to buy from you should get the most attention from your marketing resources? Yes and yes. But perhaps you don't know how to identify your best customers. If so, then you may want to learn about a technique used by direct marketers that does the trick.

Traditional direct marketers identify their best customers using the RFM rating. RFM stands for Recency, Frequency, and Monetary. You can use these three metrics to target your very best customers:

• Recency. The more recently that an action has been taken, the sooner it is likely to recur. So, your most recent purchasers are the ones most likely to repurchase soon. Those who visited your Web site yesterday are more likely to visit again today. Those who listened to last week’s podcast are more likely to listen to this week’s.


• Frequency. The more frequently an action occurs, the more likely it is to be repeated. Your most frequent purchasers are most likely to repurchase soon. Those who visited your Web site every day last week are more likely to visit tomorrow. Those who have listened to the last four podcasts are more likely to listen to this week’s.


• Monetary. The higher the value of the action, the more likely it is to be repeated. Customers with the largest orders are more likely to repurchase. Those who have purchased a subscription to your Web site’s premium information are more likely to visit tomorrow. Those who paid for your Webinar are more likely to listen to your free podcasts.

By targeting customers that had bought most recently, most frequently, and with the highest order sizes, catalog marketers can afford to mail catalogs more frequently to that group (and less frequently to less valuable groups). They increased their revenue while decreasing their costs.

Why did this work? Because RFM is a measurable proxy for customer interest. Your most interested customers—the ones who are most valuable to you—have purchased recently (they are current customers), have purchased frequently (they are regular customers) and spent a lot of money (they are profitable customers). These customers are the most interested in what you sell and they are the most responsive to your marketing message and your offers.

RFM is a rating system that you customize for your own business, so you may need to experiment to find the optimum formula. A car dealer calculates purchase Recency in years, while Apple rates iTunes customers in days. Frequency is often calculated over a period of months, but seasonal businesses may want to use one year. Monetary ratings vary widely based on the cost of your product or service. Regardless of how you calculate RFM ratings for your business, the point is to start doing it now so that, over time, you will home in on your very best customers. You don’t have to do it perfectly from the start, but make sure you do start.

RFM has everything to do with how you personalize your Web site. Using this rating, you can decide who your best customers are and show them your best offers.

For example, when customers with high RFM ratings visit your Web site, you know that free shipping is a relevant offer, because they frequently purchase from you. You also know that promoting other products you sell may also be relevant offers, because the more they buy from you, the more likely it is that they’ll buy more yet.

If you haven't been targeting your best customers, imitate those smart catalog marketers and give RFM a try.

Posted by mikemoran at 4:36 PM | Comments (0) | TrackBack

February 14, 2007

How Searching is Like Traveling

I've written in this space before about Web site searcher behavior, but I don't think I've ever talked about how it feels to search. To do so, I think it helps for us to compare the way it feels to search for information to the way it feels to search for a physical destination. I believe that we human beings have many of the same neurons firing when we engage in those two tasks. So, how is searching like traveling? I think it's all about feeling "in control."

Many people, faced with a trip into an unfamiliar city, choose to drive their cars, rather than taking the train, because driving provides more of a feeling of control—they can change lanes or choose a different route to speed them on their way. They may even know logically that the train is faster, but they can't stand the helplessness of waiting at the train station with no control over when or whether they will be picked up. They also may feel more comfortable following street signs than they do navigating an unfamiliar subway system with no physical landmarks.

Now, if the same people find themselves commuting into the city every day, they may eventually choose to do it the fastest way, despite their discomfort, and they eventually will become confident that the train will come.

What does this have to do with search?

Just like travelers, searchers want to feel in control. And just like those once-in-a-while city travelers, searchers don't know where they are going—that's why they are searching in the first place. So Web site searchers start out not feeling in control—maybe that's why they tend to navigate by clicking links before they search.

So how do you put a searcher in control? You might think it would be to offer more options, such as the Advanced Search forms most Web sites offer.

Unfortunately, what makes a searcher feel more out of control than anything is when the search engine says "no results." It's like finding out that you got on the wrong train and that you have to backtrack.

Many advanced search interfaces encourage the entry of too many parameters, leading to a "not found." Rutgers University found that searchers can't predict the effect of changing their queries, which leads to a level of helplessness.

Don't mistake providing lots of control with giving searchers the feeling of control—they must be able to predict with confidence what they are getting and where they are going. That's one of the reasons that multifaceted search is gaining steam over traditional advanced search interfaces.

Posted by mikemoran at 9:44 AM | Comments (0) | TrackBack

February 13, 2007

Are You Ready for Inline Advertising?

You're already buying paid search ads. And contextual ads. You've even dabbled in local search and maybe even demographics. But are you ready for inline ads? A few companies hope you are.

While Google, Yahoo!, and Microsoft duke it out for the bulk of paid search ad dollars, the little guys are innovating. Inline advertising is what it sounds like: contextual ads that appear as inline text links in the middle of a paragraph.

Several companies are pioneering this new form. Vibrant Media has an inline ad offering, as does MIVA, a longtime paid search vendor formed by a merger of eSpotting and FindWhat.

I spoke the other day with Chrysi Philalithes, MIVA's Vice President of Global Marketing, who is excited about MIVA's new Monetization Center, which offers a self-serve ad market for inline and other ads. Check out a live example of MIVA inline advertising or just look at the picture to the left.

You'll see that the word futbol is underscored twice, indicating that MIVA has dynamically chosen that word as a paid link for one of its ad network participants. For the advertiser, inline ads work the same way that contextual ads do—the advertiser bids on a keyword, submits the ad copy, and pays the rate bid for each click.

The difference is on the delivery side. The first step in delivery is the same—the ad network analyzes the text on the page to decide which purchased keywords ought to trigger ads. But the rest of the experience is different. Unlike contextual ads, which are displayed off to the side as the page is displayed, "inline ads are more unobtrusive", Chrysi says. No more than five inline links are highlighted on a page and the ads are not shown unless the visitor mouses over them.

Time will tell whether inline ads are the next wave in paid search advertising, but with increasing pressure to monetize content, inline ads provide a way to increase the number of ads on the page without removing any content.

Posted by mikemoran at 11:06 AM | Comments (2) | TrackBack

February 11, 2007

Open Season on Search Marketing

In the wake of Jason Calcanis letting everyone know that search marketing is worthless, SiliconValleyWatcher Tom Foremski weighs in with his own views: "SEO, beyond basic principles, is not worth it"—I beg to disagree.

There are several misconceptions here that I need to take one at a time.

First, Tom opens the piece quoting an IDC study by Sue Feldman that "shoots down" the idea that the majority of traffic to Web sites come from the top search engines. I haven't seen the study, but it appears that this study lumped all searches into a pile and says that only 30% of them are happening at Google, Yahoo!, and friends. I have not seen that statistic before, but it doesn't shock me. Undoubtedly many searches are done using Web site search facilities—70% could be the right number for that.

I'm not sure what you do with that number though. Suppose I told you that only 5% of the clicks on the Web were made at Google and company and 95% happened on other Web sites, such as the ones Google leads to. Would that lead you to the conclusion that you shouldn't optimize for search?

The problem with this logic is that there is no distinction being made between how customers find your site and how they find a page after they reach your site—they are two related, but different animals. To say this in the most extreme way, people searching on Yahoo! aren't at your site, and the only way you get them there is to appear in Yahoo!'s results.

After they find you, then they can use your search facility to find where they want to be on your site. And they can bookmark you and return to you without searching again. But being MIA in Yahoo! means that customer never finds you and instead goes on to another site. This is so basic that I am not sure how the SiliconValleyWatcher can watch so closely and not see it.

If they can't find you, they can't buy from you.

The second problem is Tom's belief that because he personally gets only 5% of his visitors from search marketing that it isn't worth it. You might imagine that blogs (especially successful ones like Tom's) get more of their visitors from subscribers than other types of Web sites do. But those 5% of visitors are still important—they discover the site for the first time in many cases, and they become new subscribers, which seems important to me.

Most Web sites get far more than 5% of their traffic from Google and friends. IBM, for example, draws 25% of its traffic from search engines, which is worth millions a year. Before focusing on search marketing, ibm.com drew 1%. You do the math. Someone was getting that business—just not IBM. Now IBM gets 25 times more.

The last problem is that Tom believes that there is a choice that has to be made between optimizing for search engines and for people. What's true is that you have to choose between spam techniques and writing for people. And he is right that you should always choose people.

But there's no need to avoid search marketing techniques to help people. Be careful to use the best words for the subjects you talk about, ensure that your content is crawlable, and be interesting enough to draw links—those search marketing techniques are good for people too. There's no need to choose between search engines and people. Do the things that help both.

Search marketing is the cheapest way to get traffic to your site—my Skinflint's Guide to Search Marketing is living proof. To ignore these simple, free techniques is a huge error.

I haven't seen the study, but knowing Sue Feldman, it would surprise me if her study tells you that we are overestimating the value of search marketing. What it might tell you is that we are grossly underestimating the value of Web site search—it is far more important than most of us believe. When people come to Amazon and type in a query, they expect it to work or else they stop doing that. The same is true for every company's Web site.

If your Web site's search doesn't work the way you'd like, check out IBM's free search engine. It's easy to install and costs nothing. (Did I mention it's free?) My advice is to continue to emphasize search marketing and to increase your focus on search for your own site. It's a false choice to pick between them. One gets you the traffic and the other sells your product to them. Don't choose between breathing and eating—you need them both.

Posted by mikemoran at 11:19 PM | Comments (0) | TrackBack

February 9, 2007

Multichannel Portfolio Management

Most of us know the term portfolio management, maybe as it is applied to investment portfolios—by diversifying your investments among stocks, bonds, bank accounts, and other investments you reduce your risk dramatically but have about the same upside potential. And some of you know how to balance a product portfolio, by making investments in product development based on which ones have the highest potential for profit. But how many of you practice portfolio management for your marketing budget?

Multichannel marketing is no longer a novelty—it is part of most of our day-to-day existences. But as we move to more feedback-based marketing approaches, such as Web search and e-mail, we find that what we should invest in those tactics depends on what they return. The whole Marketing Performance Measurement (MPM) movement is based on that principle.

And it's a principle that I wholeheartedly subscribe to. One of the criticisms that a couple of folks made about our book, Search Engine Marketing, Inc. is that it has too much about measurement and was not purely about search technology. But I think that's wrong. I believe that search marketing is more about marketing than search. That's why measurements are so important.

Much has been written, by me and by people smarter than me, about how you can make changes and use the feedback to improve your approach for the next time. I like to call that Do It Wrong Quickly, but you can call it whatever you want as long as you do it.

But multichannel marketers have a larger problem, one that goes beyond optimizing within a single channel by tweaking your Web design or changing your ad copy. Shouldn't we use this feedback to allocate resources among competing channels? Why must we agree at the beginning of the year that Paid search gets 15% of the budget when we see as the year goes on that it is driving far more business than magazine ads? If you use a portfolio approach—constantly balancing your resources across channels—you'll optimize your total budget rather than sub-optimizing each channel individually.

If you use a multichannel marketing approach, you must constantly decide how to allocate the budget among competing marketing tactics. Imran Khan, Chief Marketing Officer for Internet lender E-LOAN, uses a portfolio approach to determine which tactics get “credit” for each new customer.

E-LOAN spends 70% of its marketing budget between TV and search marketing, and uses rules to determine which gets the credit for a new customer. Imran explains: “When people are searching for our brand name when they have never been to the site before, we credit an awareness channel like TV. When people search for home equity loan, we credit search.”

The tactics that get more credit become larger parts of the marketing mix over time. “To do a good job of budget allocation, we need to do a good job of credit allocation,” Imran says.

Do you measure which channels get credit for your sales? If you do, then you already have the tool you need to optimize your marketing channel portfolio. If you don't, it seems like a good place to start working on before your competitors get there.

Posted by mikemoran at 9:34 AM | Comments (0) | TrackBack

February 8, 2007

More on Personalized Search

My recent column in Revenue Magazine has prompted some interesting responses from around the blogosphere, but I think my favorite is Mike Marshall's in WebProNews. Mike gives a real example of the personalization that is creeping into results today and speculates on what becomes of hoary search engine optimization techniques when every searcher can get a different result for the same query. I think the result will be a back to basics approach.

In my column, I talked about how the search engines might become the ultimate source of rank-checking data, because only the search engines will know for what percentage of queries your site ranked #1. Or for what demographic groups. So far, however, the search engines don't provide (or sell) this information.

So what's a search marketer to do?

You may need to get back to basics. Even without personalization muddying the waters, search rankings are a blunt instrument at best, showing you where your page ranks at this moment. What you really want to know is not whether you are #3 right now, but how close you are to being #2, or how far ahead of #4 you are. You want to know if your page has a strong position or one that could fall away easily.

To find out that answer, you need analytics tools that weigh the factors that search engines use to produce their ranked lists, such as the strength of your inbound links, the use of keywords on the page, and other factors. Tools that measure pages based on these factors abound, but they frequently work in isolation, making it hard for the search marketer to put together the whole story.

Worse, different search engines weigh the factors differently—and different queries use different factors even within the same search engine. A query with few results is more likely to emphasize on-page factors, such as where the keywords are placed, while a hyper-competitive query such as "digital cameras" demands link strength with little regard for what's on the page.

How do you figure out the strength of your pages?

I've spoken recently to someone who claims to have the answer. Mike Marshall, the author of that interesting piece, also heads up Fortune Interactive, which offers tools that analyze the results for each query in your campaign and provides insight into how strongly each search result is entrenched in its position. It literally shows you who is strongly staking out a position and who can be easily had.

To me, this is the future of search marketing—to go beyond simple rank checking to measure the intrinsic strengths and weaknesses of your pages against your competitors for the queries you care about, keyword by keyword and page by page. I'm still not sure if this is the ultimate solution for coping with personalized search, but it is an improvement. Anything that gets deeper than simple rank checking provides deeper insight.

I asked Mike whether using the top results in his analysis was problematic when personalized results could bring in new pages that were not being analyzed. Mike told me that his tool aids search marketers in personalized search situations because "search engines usually move around the results" but don't produce completely new lists of pages when personalizing. So, #12 for one person might become #4 for another, but #300 doesn't move to #1. His theory is that analyzing the top pages of generic search results is still the most useful thing to do for personalized search.

I think there's no doubt that this approach beats traditional rank checking—expect to see more innovation in search result analysis, both from Fortune Interactive and from others. Search marketers need to use deeper analytics techniques to learn how they are doing, if search engines don't provide the information they need. Just be careful not to fall into the trap of "chasing the algorithm"—your pages need to reflect what your company and offerings are about, not just what makes them popular with search engines. And your pages must first and foremost appeal to people, because getting a #1 ranking with no sales does you no good.

So, continue to use the same approach you've always used to write for people first and search engines second. But it's OK to use these tools to assess your strength so that you can make sensible changes that might improve your search visibility, as long as you use your judgement about which changes really provide value to the searcher. Not everything that gets you a higher ranking (personalized or not) will be more relevant to the searcher, which is what really improves your sales in the end.

As search changes, the tools must change with it. I like the idea of going beyond simple rank checking and expect to see more innovation as search marketing comes of age.

Posted by mikemoran at 8:39 AM | Comments (0) | TrackBack

February 6, 2007

Do You Know How Searchers Behave?

We all know that the searchers on our Web sites frequently find search frustrating—it’s simply a more difficult task than most things they do with computers. Because we are all searchers, we can recognize our customers’ thought process when searching, which helps us to design our sites around that knowledge. We all recognize the search pattern: enter a few keywords, see the search results, choose a result to click, and find your answer. There’s just one problem with this simple scenario. It hardly ever happens.

The keyword-result-click scenario is the ideal one. It is the one we keep in our heads when we design our sites. But it is not the way searchers actually behave most of the time.

We need to break out of the tunnel vision that skews our understanding. We need to view the world as our customers do, which is quite difficult. After all, we are experts in the information on our Web sites. We know exactly what words to search for. We know where everything is. It’s terribly hard to forget what you know and look at your site with fresh eyes. But that’s what we need to do.

We must also forget all the analogies to looking up information in a library. All the behavior that people applied to card catalogs and the Reader’s Guide to Periodical literature has no application here—when people search, they revert to much more basic human instincts.

Searchers are foragers. They are berry pickers. They are basically using the behaviors hard-wired into all of us to hunt for food. Research first conducted at Xerox PARC (and validated in numerous studies) show these behaviors time after time.

Searchers do not follow the straight line path of the ideal searching scenario—instead they follow the "scent" they are after. A searcher has a few words in mind when looking for something. The searcher types a few of these words into the search box and keeps a few others in mind that are not typed. The searcher scans for these words in the search results and in all the pages viewed in a hunt for information. The searcher follows the "scent" of these words to lead toward the goal, the answer—the mind food.

And like any good forager, searchers are always balancing the amount of effort to capture the food with the value of the food itself. Searchers that are both hungry for information, and believe the information will be very good, work much harder to get that information. So, if your site looks like it has the information needed, and it looks like that information will be easy to find, searchers will try harder to find it.

But not that hard. Because foragers are constantly making new decisions every minute. They don’t plan three clicks ahead. At any moment, a berry picker could decide that the berries left on this bush are too high to be comfortably reached, so they move on to the next bush.

Imagine a searcher at Home Depot's Web site searching for "fan"—when a group of ceiling fans are shown, the searcher realizes that the keywords should have been "bathroom fan." Then, a series of bathroom fans are shown, but they are the wrong color. Maybe the searcher notices that they come in multiple colors—perhaps the searcher adds the color to the growing list of keywords and searches again.

Or maybe the searcher abandons the Home Depot site.

Searchers can abandon the site at any moment to shop at a competitor. They can traipse back to Google to search more broadly. This winding trail is the typical behavior of someone seeking information. This isn't Home Depot's fault in this case; they may have a perfectly good search facility. But this is the true state of searcher behavior.

Just 34 percent of site searchers find what they are looking for, with 47 percent of those that failed trying only one search query before abandoning the attempt. That tells us they can’t detect the scent of what they seek, so they are moving down to the next berry bush.

Throughout the information finding experience, searchers are looking for the confidence that they are making progress toward their food. A study by usability expert Jakob Nielsen found that 40 percent of e-Commerce site searches yielded useful results (that the searcher explored further), 37 percent caused searchers to search again, with the remaining 23 percent of searchers abandoning searching to try a different finding strategy that they decided was better at the moment. That’s what information foraging is all about.

You can't change searcher behavior, but you can change your search engine. If your Web site search isn't all you want it to be, you can try a simple, free search engine to see if it works any better for you.

Posted by mikemoran at 10:19 PM | Comments (3) | TrackBack

February 5, 2007

Now Leaving for Panama

Many great blogs cover the day-to-day news of search, so I don't spend much time duplicating them. But once in a while, a story is so important that I can't leave the coverage to others. Yahoo!'s long-awaited delivery of its Panama project is one such story. At 3 pm US Pacific Time today, Yahoo! will convert its top-bidder paid search auction system (that it bought from Overture) to its new hybrid auction system. The world will be watching—including search marketers.

Search marketers care about Panama because they must learn a new tool and, for some, a new way of working. Any search marketer that exclusively uses Yahoo! for paid search is accustomed to the traditional top-bidder auction, where the highest bid for a keyword grabs the top spot. Top-bidder auctions are easy to understand, but hard to manage, with bidding tactics such as bid jamming—requiring search marketers to use bid management software to respond to what competitors do.

Google pioneered the hybrid auction, where bid amounts alone are just one factor in which ad is shown first. Google uses clickthrough rate and page quality in its paid search rankings, but the concept of a hybrid auction allows any criteria to be used, the same way complex organic search ranking algorithms do. Last year, MSN Search revamped its paid search auction to adopt the hybrid approach, dumping Yahoo!'s Overture system in the process.

Search marketers usually find hybrid auctions easier to manage. You still need to use bid management software, but because clickthrough rate is not as volatile as bids can be, the rankings are less volatile and marketers can think more strategically. Search marketers will be poised to see how Yahoo!'s new auction makes their lives easier and fills their coffers higher.

But search marketers have another reason to watch. MSN Search has seen no real gain in popularity since it moved to its new hybrid paid search auction. Yahoo! Search's popularity has been flat in recent months, also, while Google has gradually increased its share of searches.

Can Yahoo! turn around this trend around, making the search market more competitive? Recent criticism of Yahoo! has raised the pressure on the company for Panama to be a big win—so much so that Yahoo! has advised analysts not to expect any revenue boost from Panama until later this year.

I've had conversations with search marketing experts that believe that Yahoo! is so far from catching Google that it is more likely to fall behind Microsoft. I don't agree with this viewpoint, but it shows how critical Panama is to Yahoo!'s prospects. Search marketers should be rooting for Yahoo! and Microsoft to give Google a better game because they want as much innovation and price pressure as competition can bring.

So cross your fingers today and let's see what Yahoo! can do.

Posted by mikemoran at 10:34 AM | Comments (0) | TrackBack

February 3, 2007

The Right Turn Theory

In this month's newsletter, we asked if you were creating a Do It Wrong Quickly Web site—but you probably already have a Web site. The question is, "Are you improving it wrong quickly?" Or are you constantly planning large, costly redesigns as the only way forward? If you've been in the Web business any amount of time, you've probably been part of a top-to-bottom Web site redesign.

But what if you think about Web improvement the same way you consider home improvement? How often is it the right thing to knock down the entire house and build a new one in its place? Only if you really hate your house, you have a lot of money, and you can stand a huge disruption in your life. Most improvements should be small, like adding a new piece of furniture or repainting a room, but sometimes it makes sense to rip out a bathroom and put in a new one. Once in a while it makes sense to do something major, such as expand the house so your mom can move in with you. But it’s rare that you should tear your house down and start over.

Your Web site is no different from your house. Most Web sites have relatively happy customers, but like most houses, can always be improved. Unfortunately, our approach is usually to tear the site down and build a new one where it stands—which we call a “site redesign” or a “major re-launch.” We don’t repaint a few rooms, or even put in a new bathroom. We hire a new set of designers and information architects and do something dramatic to justify the cost.

But is that what your customers want?

Some companies have learned the hard way that their customers prefer more subtle evolution in their user experience. For example, eBay changed its yellow background in favor of the more typical white—and received so many complaints they felt compelled to change it back.

Now, did those customers really have some intrinsic problem with white backgrounds? Doubtful. White is the most common background of Web pages today. What eBay heard from its customers is that they like the eBay site and they don’t want anyone mucking around with it.

So, eBay did something very clever—gradually removing the yellow one shade at a time until eventually they had that white background they wanted. This evolutionary approach went unnoticed by customers.

Could eBay apply what it learned to other kinds of changes? Yes. When eBay discovered improvements they could make to the forms on their site based on user testing, they had a dilemma. If customers complained about something as innocuous as a color change, what would they say about modifying the forms that were used millions of times each month?

The folks at eBay decided to leave it up to their customers. They introduced a new form link on their page that showed each customer the improved version and allowed customers to adopt the new form for themselves. When eBay saw how many people chose the new ones, they developed the confidence to replace the old forms site-wide.

Why does this work? Think of it as the “right turn theory.” In countries where drivers keep to the right side of the road, studies show that far more accidents occur when drivers make left turns in front of oncoming traffic than when they make right turns. Big changes in your Web site are left turns. Yes, you get where you are going faster, but left turns are riskier. Instead, make three right turns. You get to the same place, but you’ve taken in a bit more scenery along the way and it’s much safer. So remember, two wrongs may not make a right, but three rights make a left.

Use a gradual approach when making changes. You can offer new experiences as “beta”—giving your customers the choice of what they want. Usability expert Jared Spool says, “The best teams not only design the changes, but design the process for introducing the change...to overcome the users’ natural resistance to change.”

Google is one of the biggest practitioners of “beta” experiences—it introduced its Gmail and many other ideas just that way. Google goes even further, by testing tiny changes with small groups. If you notice that your Google search screen looks different from the person next to you, it’s because Google is testing a new feature on a random subset of searchers.

You, too, can tweak your experience a little at a time, rather than gutting it and replacing it every few years. What happens when you don’t keep tweaking? One marketing director laments:

“So we’re in the situation now where we built our Web site in 2001 and 2002 and it was a great Web site at that point and it provided a really nice jumping off point for individual brands to layer on campaigns and support for their specific business objectives. Well, nobody invested in keeping that foundation solid and current—and online the landscape changes so quickly. We are now in 2007 and we need to spend a ton of money simply to rebuild the foundation. That’s a big lesson we are trying to impart to people. We are not going to do a Web design project and walk away from it. We need to turn this project into a program.”

Learn from this. Every time you make a change to your Web site, you need to measure its results and make it better. When most people think the project’s over, you need to know it’s just starting.

Posted by mikemoran at 6:29 PM | Comments (0) | TrackBack

February 1, 2007

Syndication on WebProNews

I'm pleased to announce that blog entries from Biznology are now being syndicated on WebProNews. If you don't usually check out WebProNews, there are several excellent columnists there, including Nathan Weinberg and Robert Scoble.

Posted by mikemoran at 11:34 PM | Comments (1) | TrackBack

A Do It Wrong Quickly Web Site

If you need to create a Web site, isn't that a complex undertaking requiring lots of research and painstaking details? It doesn't have to be. See how to apply an experimentation approach to your new site in February's Biznology newsletter, called A Do It Wrong Quickly Web Site.

Posted by mikemoran at 11:33 PM | Comments (0) | TrackBack