What s Next for Search Marketing?

What's Next for Search Marketing?

Our crystal ball is as foggy as anyone else's, but we do see a number of trends that will play out over the next few years to change search marketing:

More content. Every year, more and more information is made available on the Web, yet much information remains trapped offline or otherwise inaccessible to search engines. How will this change?
More technology. Like any technology, search technology never stops improving. As software technology marches on, what are the implications for search marketers?
More personalized. Every searcher who enters the same query gets the same results, right? That is something that will probably change in the future, as search results are tailored based on the person and current situation. What will that mean to you, the search marketer?
More competition. The good news is that you are discovering search marketing, but the bad news is that everyone else is, too. How can you stay ahead of the increasing competition in search marketing?

Put on your long-distance glasseslet's examine the latest trends in search marketing, starting with what new content is expected to become available to search engines.

More Technology

Like all the software around us, search technology never stops improving. Even if you have the search marketing game wired today, you cannot be caught unawares as the rules change tomorrow. Let's look at some of the technology changes you should expect to see in the next few years. Although some of you might find the technology to be intrinsically interesting, every search marketer needs to ponder how technology changes might impact search marketing.

Keyword Assistance

Since the advent of text search in the 1960s, one problem has bedeviled searchers and continues to haunt them to this daysearchers often do not know the best words to use in their queries. In time, search engines will use some new techniques to help searchers pick the right words.

As we write this, Google is testing a new feature called Google Suggest (shown in Figure 16-2) that offers suggestions as you type in your queries. The figure shows what you would get if you typed in "dig"several different queries are displayed that all start with those letters, ranging from "digital camera" to "digital camera reviews" and more.

Figure 16-2. Google experiments with keyword assistance. Google Suggest tries to fill in the rest of the query as the searcher types the first few letters.

If Google Suggest remains a beta offering or is withdrawn, it will not have any effect on search marketing. If Google Suggest passes muster, and is made part of the mainstream Google search engine, however, it might cause changes in searcher behavior. Would searchers use fewer misspellings? Would searchers start using longer queries, such as "digital camera reviews," because they did not have to think up the words and type them in? Would the most popular queries get even more popular, with fewer variations in wording?

Other technologies could impact keyword planning, too. You are probably familiar with collaborative filtering technology, even if you do not know it by that name. Amazon and other Web sites use it to suggest items to shoppers (as in "Shoppers who bought books written by Max Readership also have bought books by Minnie Seller"). Suppose search engines started using it to suggest keywords to searchers (such as "Searchers who entered your query also tried these four related queries")? As search engines offer these alternative queries, they could emphasize the alternatives whose results have historically drawn many clicks (indicating the queries produced relevant results). If search engines did start using collaborative filtering to show popular queries, how would that change keyword demand? Would highly popular queries become even more popular?

What if search engines could look for synonyms of words that searchers enter? Search experts have toyed with this function for years, but the synonyms always seem to return more results that are off-topic than searchers are willing put up with. Still, you can bet that some researchers are still working on this.

As you can see, these simple changes in the search user interface could have drastic implications for your keyword planning. You might not know in advance what all the implications for keyword planning will be, but you would at least know that you should be looking for shifts in keyword demand, and perhaps even conversions, if one of the major search engines adds keyword assistance features. You will use your metrics to see how the game has changed and can react more quickly than your competition. Neeraj Agrawal of Battery Ventures (www.battery.com) believes that a new kind of search marketing specialist will emerge that he dubs "Keyword Man," to cope with the increased complexity of keyword planning. Regardless, search marketers must evolve as search technology evolves, in keyword planning and in other areas, as we see next.

Deep Text Analytics

If you think the technology that search engines use today is magical, you ain't seen nothing yet. Researchers have been working for years on breakthroughs in linguistics and pattern matching that make today's search engines seem positively primitive in comparison. Here are some of the text analytic techniques you can expect to emerge from the labs:

Summarization. Snippets are nice, but some researchers are working on technology that accurately summarizes an entire document in a few sentences. Some implementations string together several sentences from the document, whereas others generate text from the central concepts.
Categorization. Most documents have missing or inaccurate metadata, so computer scientists are developing programs that automatically generate correct metadata. Text analytics are already available to discover the subject (digital cameras) and type (white paper) of a document, as well as its grade level. In addition, some software can categorize the source of information (mass media, educational institution, corporation, and so on), its industry, geographic location, and publication date. All of this information can be used by search engines to limit the results of a search, such as "Find all case studies from the retail industry published after 2002 on the subject of credit card fraud."
Entity extraction. Text analytic software already employs linguistics and pattern detection techniques to identify specific kinds of names within the text: person names, company names, and place names, for example. Some software can identify job titles, company mergers, product announcements, and other concepts. Other analyzers ferret out numeric information, such as company revenue, dates, and people's ages. Search engines could use this intelligence to allow more focused queries, such as "Find documents that mention product announcements from Snap Electronics and its competitors within the last two months."

The combination of these techniques might make true question answering possible, such as "When was George Washington born?" or "Where is Boeing's headquarters located?" or even "Which digital cameras have won awards?" Researchers can already demonstrate prototypes of search engines that answer questions, but this technology has not been generalized for use by the major search engines yet.

If these techniques come to fruition, search marketing might be profoundly changed. Although optimizing for keywords and drawing links would remain important, your content might need to emphasize facts as well, to continue to impress the search engines. Copy that mentions awards won or feature specifications that are meaningless for search today (although very valuable for driving people to convert) might suddenly become important answers for searchers' questions.

Even if this technology stops short of actually answering questions, it might allow searchers to use natural language queries; that is, they type in full sentences to find what they are looking for, such as "Show sales on digital cameras with at least 4.0 megapixel resolution." Searchers have historically been unwilling to do this much typing, because today's search results do not improve with extra words. Some search engine might come along to show better results with these longer queries, providing the incentive for searchers to change their behavior. Think about what effects that would have on your keyword planning.

Multifaceted Search

Our next technology is one of the most interesting of all. You might call multifaceted search the "anti-search" because it is often used without entering any keywords at all. In actuality, however, multifaceted search technology is just a more powerful addition to traditional text searching that enables searchers to restrict the search results by responding to choices offered by the search engine. Multifaceted search enables searchers to continually narrow down their results by choosing another constraint on their search.

Multifaceted search can already be spotted in some shopping search engines, such as shopping.com, as shown in Figure 16-3. Shopping search is a natural application for the multifaceted search approach, because some products, such as digital cameras, are bought based on the facets (features) of the product. Shoppers care about price, brand, resolution, and other facets of a camera. As they choose the facets they want (such as "more than 6.0 megapixel resolution" or "between $450 and $650 in price," and so on), invalid choices disappear, so shoppers never get a "not found."

Figure 16-3. Search without searching. Multifaceted search allows multiple selections from menus without ever serving up a "not found."

This multifaceted magic is made possible by the same XML trusted feeds discussed back in Chapter 10, "Get Your Site Indexed." Each digital camera retailer provides a data feed to shopping.com that includes the values ($649) for each facet (price) for every camera available. The multifaceted search engine does the rest, displaying the precise values available in the search index based on the current trusted feeds. If a particular model of camera sells out, it is immediately removed from the feed so that no shopper is disappointed.

But shopping search is the tip of the multifaceted search iceberg. Multifaceted search could be used in the future for any kind of traditional text search application, if the search engine has enough information (values for facets) about the documents being searched. Remember the deep text analytics we discussed earlier? If search engines can automatically discern metadata, such as document type, subject, industry, and other information, each of these metadata fields can become facets in a multifaceted search. Imagine that specialty content being available through multifaceted search. Patent attorneys could perform their searches while limiting their scope to only patents. Transactional searchers in the Buy phase of the Web Conversion Cycle could eliminate all Use information from their results. As search engines apply categorization technology (the deep text analytics we discussed earlier) to every document on the Web, the metadata is created to enable multifaceted search technology to work on any search.

More Personalized

Beyond technology, a key search trend is toward a more personalized experience for searchers. Although many Web sites, led by Amazon, have personalized their user experience, search engines have been decidedly retro. Different searchers, by and large, get the exact same results when they type the same query into a search engine.

A quick review of the history of search technology reveals that the vast majority of improvements have been based on the contentanalyzing it better, understanding it more deeply, and assessing its quality. But what about applying the same kind of thought to the searcher? After all, what makes a successful search is the best match between the content and searcher, so why have search engines focused on analyzing the content so much and ignored the searcher?

In part, the maniacal focus on content stems from the fact that understanding the content is easier than understanding people. However, despite the inherent difficulty, search engines are already beginning to understand searchers better, and you should expect to see more personalized search results based on several factors:

Geography. Some search engines are already experimenting with local search and geographic targetingthey ask for place names in the query or they guess where searchers live and they try to show results from companies nearby. However, their only means of guessing is to analyze the searcher's IP address, which is not a flawless method. Someday search engines might entice searchers to register and disclose exactly where they live in a more permanent way, so that search engines can use that reliably to enhance results and provide more targeted paid placement opportunities.
Language. AOL, MSN, and Yahoo! already have relationships with many of their searchersthey offer other services that cause people to identify themselves and provide information. If search engines knew what language each searcher prefers, they could restrict results to that languageeven automatically translating results to the searcher's favorite language when necessary. Non-English Web users and non-English Web pages are both growing faster than their English counterparts, so this capability might become important soon.
Interests. Could search results be improved if search engines knew searchers' interests? When searchers enter "jaguar," are they looking for the car, the animal, the football team, or the Apple operating system? If search engines understood the searchers' interests, they might be able to take a better guess. Eventually, one of the search engines will begin to collect information on searchers' interests to flavor the search resultswe see whether searchers disclose this information willingly or if it needs to be collected as a byproduct of their Web usage, as Amazon does.
Demographics. Will search engines begin to collect information about searchers to target paid placement? Would you be interested in buying paid search terms that were targeted to women? To people of a certain age? Or employees in a certain industry? It remains to be seen whether searchers will disclose such information to search engines or whether they would approve of their information being used in this way, but it could radically change the way search marketing works.
The wired home. Web usage dramatically changed with the widespread adoption of "always-on" broadband connections. When surfers can use the Internet by walking up to their computer at any moment, they tend to use it moreand to search more. What will happen when there are Web browsers in the living room, the kitchen, and the bedroomanywhere in the house? Will searching increase even more? Will the types of searches change? Will search engines explicitly support kitchen tasks, such as suggesting recipes based on ingredients on hand or locating the store that has stocked every item on a shopping list? Will voice recognition be required for a kitchen search engine to be truly useful? Regardless of the exact form searching takes within the wired home, it seems likely that devices built in to the searcher's home will have far more information about the searcher, which could then be used to personalize search results.
Handheld devices. Will search results change as searchers use new devices to connect to the Web, such as personal digital assistants (PDAs), cell phones, or other small-screen devices? If you stop to think about it, these small screens with ubiquitous wireless access to the Web are tailor-made for a more search-centered user experience. There is no room to look at long Web pages with lots of links to navigateyou will want to search (maybe by pecking a small keyboard or using a stylus for handwriting recognition, or perhaps with voice recognition). Will Global Positioning Systems (GPS) that know where the searcher is located allow even more targeted geographic pinpointing of paid placement? ("Show me the hardware store nearest to where I am now.") Handheld devices contain e-mail, calendars, to-do lists, and other information about their owner that might be used to personalize search results.
Information alerts. Would the advent of more personalized search results fundamentally change information gathering behavior to a more passive activity? Once search engines know what searchers are interested in, will they begin sending it to them without waiting for them to search for it? If search engines know who wants to buy a particular car, would they send an alert message to the cell phones of each person when that car goes on sale at the local dealer? Would the dealer pay the search engine to do so? Time-saving information alerts might be just the incentive for searchers to disclose their interests and other personal information to the search enginesto get the alerts they want, they will need to reveal something about themselves, which allows the search engines to personalize more deeply in other ways.

We live in a world today where "one size fits all"at least for search results. If one person finds a document using a search engine, another person can enter the same query and find the same document. Personalized search changes the game. Every person might get different search results from the same query, just as they each get different products listed on the Amazon home page. And the same person might get different results based on contextwhere he is, what device he is using now, and what he has been interested in lately.

Some searchers will be amenable to providing information about themselves to the search engines, just because they want to improve their search results. But others will not want to make any effort at all. Will search engines analyze each searcher's pattern of queries to determine interests? Will search engines provide other services to learn more about each searcher? When search engines perfect the technology to offer more personalized searching, many thorny questions might yet remain as to whether searchers will give up some privacy in return for the time savings of more relevant results.

It will take years to play out, but search marketers must be aware that personalization is coming, and the time to consider the implications is now. If the burden of translating your Web site into multiple languages were lifted, could you profitably operate in countries you avoid today? Perhaps you would benefit from learning more about how offline advertising works, so that you understand the demographics of your customers. It might be helpful to learn how impulse purchases can be influenced, because your retail store or restaurant might someday need to do more than hang out a big signyou might want to digitally alert passers-by or nearby drivers thought to be interested in your establishment.

Regardless of your business, search marketing seems destined to become more fragmented over time, as today's purchase of a paid keyword might become the purchase of a keyword only for searchers fitting a certain demographic model. So, instead of buying the keyword supply chain management, you would buy that keyword only for searchers that work for companies over a certain size in specified industries. You would probably pay a higher price for this more segmented audience, but if your conversion rate were high enough, it would be worth the premium price.

In this brave new personalized world, how would you check your organic search rankings? Rank checking software cannot simulate personalized results, so the concept of ranking #1 becomes meaningless. Perhaps you really want to know for how many individual queries your page was ranked #1. Or maybe what your page's average ranking is across all queries, which only the search engines can reveal. Maybe search engines will start selling you organic search metrics, the way they offer paid placement reporting for search marketers today. We might move to an organic tracking system similar to paid search, where your organic search results are measured by the number of impressions (the number of times they are shown) and your page's clickthrough rate. By thinking about the effects of these changes, you will be more prepared when they arrive.

As you consider all of these changes afoot, you must remember that many others are being worked on, too. Moreover, many of these trends interrelate (keyword assistance could become personalized, perhaps, or maybe voice recognition makes searchers willing to use longer natural language queries), so their effects might be multiplied in ways that are hard to predict. That is all the more reason to stay alert for what is going on around you.

So far, we have discussed only those trends that will be seen by any searcher, but a hidden change is brewing that affects search marketersincreasing competition from other search marketers.

More Competition

You are not the only person joining the search marketing game. At the beginning of this book, we trotted out all the "gee-whiz" numbers to show you how search marketing is growing. So the good news is that you are catching the wave, but the bad news is there are a lot of other surfboards out there to contend with.

The biggest surfers are large corporations, many of which have been oblivious to search marketing, or have experimented and failed. You have seen in this book why search marketing can sometimes be difficult for big companies, but don't expect them to remain on the sidelines for long when there is this much money at stake. Large organizations are now becoming formidable search marketing competitors for three main reasons:

They are addressing their weaknesses. If large companies harness their strengths while correcting the kinds of errors we describe in this book, they will become formidable competitors in search marketing. Enterprise-wide execution will allow large companies to counter the traditional search marketing advantage held by smaller, more nimble companies.
They are leveraging their brand names. As you have learned in this book, large businesses have significant organic search advantages in attracting links and searcher clicks, due to well-known brand names. Larger marketing budgets can be applied to paid search, too. Conglomerates are beginning to see how portfolio management might apply to search marketing, the same way it applies in other marketing disciplines. General Motors, for example, has many different brands of sports utility vehicles ranging from the Chevy Aveo to the Hummer; each brand caters to market segments with different characteristics, sensibilities, and needs. A well-known brand such as Hummer draws many searches, whereas Aveo garners relatively few, but would Hummer searchers who cannot afford that pricey sports utility vehicle be interested in a less-expensive Aveo SUV? GM might be able to leverage its portfolio on the Web in ways that their Hummer and Chevy dealers never could. In addition, could GM have its divisions work together to dominate the listings for "sports utility vehicle"? You can bet they are going to try.
They are defending their brand names. The ease of publishing to the Web has a downside for big brandsit makes it easy for "hate" sites to target a particular brand and get publicity for its complaints. Some search marketing vendors, such as Global Strategies (www.globalstrategies.com) and Converseon (www.converseon.com), are helping large companies with search engine reputation management. They monitor these negative sites and work to develop effective search marketing campaigns among friendly companies to give the negative sites less attention. If the top ten is dominated by your site and your partners, the hate site has less visibility.

But big businesses are not the only source of new competition. You might already be doing battle in global markets, and you see that competitors seem to be getting more sophisticated about search marketing, lessening your advantage. Or, worse, you do business in just a few local markets but you are starting to face competition from other regions or countries that could never do business in your territory before the Internet, but now they can. Not since the advent of the fax machine has such an important way of crossing geographic boundaries come along.

Before the Web, companies seeking to enter foreign markets used exporters, licensees, joint ventures, or wholly owned subsidiaries to create a local presence in each market. Although these techniques still have their place, the Web allows businesses to sell directly to a customer, no matter what country each one is in. The rise of global search engines help a business from across the world seem just as "local" as one a block away. It is likely that your business will be competing with new entrants around the world, if it is not doing so already.

But companies from around the world might not be the only new competitors you face. You might meet new combatants right in your own backyardlocal businesses. If search engines can personalize results by location, as discussed earlier in this chapter, millions of local businesses whose ad budgets are spent on Yellow Pages advertising might now be able to profitably engage in search marketing. If you work for a large company, such as Home Depot, you might have had search marketing to yourself, but will someday face increased competition from local hardware stores.

What happens as more and more marketers realize how well search marketing works? As changes in search marketing make it profitable for more and more businesses? The simple answers are that it makes it tougher to rank at the top of organic search results and it also makes paid placement more expensive. So what do you do about that?

For organic marketing, your best approach is to redouble your efforts to make search marketing an ingrained part of your extended team's job. No more remedial programs to fix problems! Instead, prevent those problems from the start. Every new page must be designed with search marketing in mind, and no changes should be permitted without considering the search marketing impact. In that way, you will keep your costs as low as possible while reaping the highest possible gains. However, you must always focus on your message. If you truly create the content that is the best answer to a searcher's question, and you carefully optimize your site, you will continue to get high rankings. If you allow others to create better matches for your targeted searchers, however, increased competition will crowd you out of the top spots.

The growing competition in paid search is a much more difficult trend to cope with. Economics teaches that the marketplace in paid placement will force bids to the highest affordable levels for the players in your industry. This relentless efficiency will drive prices up as long as more competitors enter the market. To cope, your business must ruthlessly drive out costs so that your Web site is the most efficient conversion machine around and your product costs are as low as possible. That way, you have the maximum amount of budget available to raise your bids whenever necessary so you continue to drive conversions.

If the conversion rate of your Web site is lower than your competition's, or your product costs are higher, it leaves you with less money to put into search marketing. When the bidding gets too high, you will be forced to drop out. Consider the situation you will be in as bids mount. If you find yourself asking, "How can my competitor make any money bidding that rate?" then there are two possibilities. One is that your competitor is losing money, so you have only a short-term problemeventually your competitor will lower the bid or go bankrupt. The other, more serious, possibility is that your competitor is actually making money with a bid that would lose money for you. In that case, you must address all inefficiencies in your Web site, in your manufacturing, and in every cost you have, so you can remain competitive in your bidding.

One way to cope with rising bid costs is to shift money from one budget to another. If search marketing is truly your most efficient marketing spending, you might want to reduce other marketing outlays (offline advertising, brochures, and so on) and move that money into search marketing. That way, you can keep up with rising bid prices without spending any more money overall.

Another way to improve your efficiency is with passionate attention to deep metrics. The kinds of metrics that you need to do by hand should become automated in the future. Adopt software tools that allow your campaigns to go on autopilotthe tool constantly monitors your bidding, your conversions, and your budget. The tool can make adjustments to take advantage of inefficiencies in bidding, not unlike stock market arbitrageurs looking for stock price anomalies. If you constantly automate your campaign management to be more adaptive without human intervention, you will drive your conversions to their maximums while keeping a lid on costs.

Before leaving the topic of competition, we should point out that the changes leading to more competition offer you business opportunities, too. Just as your business might face competitors from new places, your business can seek customers in new markets that were not cost-effective in the past. If you can become efficient enough, you can become a feared competitor in any market using search marketing and the Web.

Enough heady thoughts on where the search marketing industry is headed! What about you? As you get started in your search marketing career, what's next for you?