Biznology Blog: August 2005

« July 2005 | Main | September 2005 »

August 29, 2005

What are Your Web Site's Top 10 Searches?

Like the weather, everyone complains about Web site search, but no one does anything about it. Or so it seems. Forrester Research famously titled a report on the subject "Must Search Stink?" a few years back, and I am not sure that Web site search is any more aromatic today.

Some things that can improve your Web site's search facility are hard to do, which is why search so often does cast a foul odor. But there is one exceedingly simple thing that every Web site can do to improve their search—track their most popular searches.

Perhaps you already know that concentrating on improving your most frequently-searched keywords is the easiest way to improve your Web site's search results, but many people don't. Time and again, Web site owners know that their search facility is spewing poor results, but don't know where to start. So how do you do it?

Start by figuring out what your most popular searches are. Some of you may know exactly how to do this, but many may not. Most sites do it using one of two methods:

• Use your site's Web metrics system. Most Web metrics packages can monitor input into applications, such as your search application, or they can analyze the URLs of your results pages—most search applications use URLs that contain the search keyword (such as this result for the keyword "websphere" for the ibm.com search facility: www.ibm.com/Search/?q=websphere&v=14&lang=en&cc=us). In the ibm.com URL, what follows the "q=" (query) parameter is the keyword entered by the searcher. Most Web metrics packages can extract that parameter from the search results page views and count how many of each occurred.


• Instrument your search application. On the off-chance that your metrics package does not easily provide the counts that you need, you can add code to your search application to tally up each keyword, perhaps to feed into a report generation program.

Once you have the number of searches for each keyword, you can sort them so that you see which ones are the most popular. What you'll find unfortunately, is that your top queries don't cover a high percentage of searches. Your Top 1000 queries may account for just 10 to 20% of all search volume.

Regardless, that's the place to start. You probably can't start with the Top 1000, or maybe even the Top 100, but surely you can look at your Top 10. Perhaps your ten most popular queries make up just 3% or 5% of your total queries, but it is the right ones to work on first.

What do you do with those popular queries? Start simple. Manually execute each query and eyeball the results. What does the #1 result look like? Is it the page that you would direct that searcher to on your site if you were asked where to go? We'll begin to discuss what actions to take in the next few weeks, but your place to start is with the most popular queries on your site. Find out what they are.

Posted by mikemoran at 8:57 AM | Comments (0) | TrackBack

August 26, 2005

Top Ten Tips for Search Marketing

As I moderated the American Marketing Association's Listserv recently, I was asked what the top ten tips are for search marketing. If you've already bought our book, Search Engine Marketing, Inc., the list may not surprise you, but I suspect that others would not guess at least a few of the ten.

Here are the top ten tips for search marketers:

1. Know your goals. First, define the goal of your Web site and know how much it is worth to you. Co-author Bill Hunt and I call them Web conversions—they could be e-Commerce orders, but they could also be reading a white paper or downloading software or filling out a contact form. Start by making sure that you can measure the value of your Web site so that you know what each incremental success is worth.


2. Decide your keywords. Do your homework to identify the "just right" words for your site. If you pick words that are "too hot" (such as a local physician trying to rank #1 for "doctor"), then you won't get the attention you need. If you pick "too cold" keywords (such as that same doctor expecting people will type "doctor" along with his name) you won't get any searches. Choose keywords that exactly match your site and are used often by searchers (such as "doctor" and the name of the town, or "orthopedist" and the name of the state).


3. Identify missed opportunities. If you know what keywords are just right for you, then you can use keyword research tools to find out how many searchers use those words and how many of them are already coming to your site. The difference is your missed opportunities.


4. Make your case. Use the measurements you took above to value the missed opportunities in search marketing. Estimate a conservative increase in referrals based on your missed opportunities and use your conversion rate to value what each incremental referral is worth. You can then take a shot at your business case by toting up the improvement and subtracting estimated costs to get it done.


5. Convince others. Take that business case around and get your peers to see why search marketing is important and what they each need to do to make it succeed. Do the same with your executives. Show them the value and they will likely invest to obtain that value.


6. Get your pages indexed. Once your plan is approved, you must get your pages into the search index. You must create spider paths and remove traps in its way. If your pages are not indexed, then they can never be found in organic search.


7. Optimize your content. Once indexed, you need your pages to rank well in organic search. Use the keywords you targeted in titles and throughout your copy. You don't have to "chase the algorithm" to get good search rankings. Write your content for people first and then think about search optimization—your searchers must convert when they read the page after all. Artificially high rankings from tortured keyword-stuffed prose won't help you.


8. Attract links. Get attention to your pages from other sites. Go after links to drive traffic all by themselves, and be happy that they help your search rankings too.


9. Optimize paid search. Use the measurements you saw above to choose the right amount to bid on cost per click ads. Remember that highest overall profit, or lifetime customer value, is far more important to optimize than highest click rate, lowest cost, or highest profit margin.


10. Make search operational. Every day, you must watch your numbers. Check that your pages are still indexed. Ensure that no new keywords are emerging that you are missing. Test your content for keyword richness and proper tagging. Make sure that all of your inbound links are still working. Carefully analyze profitability of paid search. Do it every day and share the results with executives and peers. Celebrate what is working and fix what is not.

A lot of hard work lies between these tips and search success, but these are the basic ideas that you must execute.

Posted by mikemoran at 11:26 PM | Comments (0) | TrackBack

August 20, 2005

Can Search Marketing Raise Brand Awareness?

Another topic that came up during our week of moderating the American Marketing Association Listserv is search marketing's role in branding. Savvy marketers have long known that monitoring keyword popularity can help you measure brand awareness, but can you actually increase brand awareness with search? I say you can, pointing people to a presentation Bill Hunt and I made at Search Engine Strategies that shows you how.

Posted by mikemoran at 4:50 AM | Comments (0) | TrackBack

August 19, 2005

Search Success on the Cheap

As we hosted the American Marketing Association Listserv this past week, we heard from many search marketers about what they are struggling with and concerned about. One item that came up was the differences between search marketing for big companies vs. small companies. We've discussed what makes search marketing difficult for big sites, but we haven't really talked through what small sites can do.

Small sites have many advantages over large sites. They are less likely to have over-architected their site, introducing expensive, spider-trapping technology the way large companies sometimes do. (It is certainly possible to implement high-end Web technology to be spider friendly, but it takes more knowledge than with boring old Web servers.) They are also more easily able to get their entire Web team to work together (maybe because it is only one person). Small sites also make decisions faster and are able to change their site faster than larger sites.

But small sites are often encumbered by a small budget. Whereas big sites sometimes have the resources required to succeed at search marketing, small sites can be starved for cash. A little needs to go a long way.

What can search marketers at small companies do? Enter the Skinflint's Guide to Search Marketing, the cheapest way for search marketers to break into the game. There is even a free way to do shopping search!

Don't let people tell you that you need deep pockets to succeed at search marketing. Consult the Skinflint's Guide today.

Posted by mikemoran at 10:32 PM | Comments (0) | TrackBack

August 16, 2005

Search Engine Marketing, Inc. Book Excerpt Available

Our publisher has made a free excerpt available from our new book, so if you want to read Chapter 1 of Search Engine Marketing, Inc. without spending a dime, here you go. (Warning: It is a 30-page PDF, so it can take a while to download, but it's worth waiting for, we hope.)

Posted by mikemoran at 12:03 AM | Comments (0) | TrackBack

August 15, 2005

Catch Us on the American Marketing Association Listserv

My co-author, Bill Hunt, and I are co-moderating a listserv this week for the American Marketing Association. If you're a member, subscribe to the E-Commerce listserv. If you're not a member, you can join. We'll be posting some questions and answering others on the topic of search engine marketing. We'd love to have you join us there.

Posted by mikemoran at 8:26 PM | Comments (0) | TrackBack

August 13, 2005

Why is Search Marketing So Hard for Large Sites?

It is actually harder for large companies to succeed at search marketing than it is for smaller ones. The very specialization that makes large organizations work hampers search marketing efforts—too many groups need training and watching for anyone to manage it all. Unless you know how to do it. Check out the August Biznology Newsletter to see what you need to know. You might be surprised.

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

August 11, 2005

In-House Search Marketing Talks

Yesterday, I had the opportunity to present as part of three panels that included my co-author, Bill Hunt, at the Search Engine Strategies conference in San Jose. Here is some coverage from WebProNews. The full presentations are listed here for download for both Big Site/Big Brand SEM and Working Together. (The third session was strictly questions and answers.)

Bill and I got plenty of positive feedback and a number of folks saying they will go buy the book, so that is always gratifying. So many medium to large companies (and even a number of small companies) are telling us that they are struggling with issues that we cover in the book—getting out to speak with folks is a great reminder of why we wrote it in the first place.

Posted by mikemoran at 11:59 AM | Comments (0) | TrackBack

August 2, 2005

Do You Know the Two Kinds of IT Projects?

Part of the uncertainty of the role of the IT group stems from two complementary, but fundamentally different uses for IT, which I will call Necessary IT and Differentiating IT. Investments that must be made to keep the business running, to lower risk, or to raise productivity and lower costs are Necessary IT. Differentiating IT, in contrast, drives competitive advantage, new markets, and improved revenue. Does your company know the difference?

Too often, all IT projects are lumped together and people try to apply "best practices" across both kinds of projects indiscriminately. The result is that success in either tends to be sporadic, driving down the overall value of IT to the organization, and raising risks and lowering credibility in the minds of senior management.

One reason the line between these two types of IT is blurred is that every successful differentiating technology eventually becomes a necessity. For example, in the 1980s, Citibank emphasized ubiquitous, simple Automated Teller Machines to drive higher customer satisfaction and lower costs. This initially differentiated Citibank against its competition, but as customers began to demand this service from all banks, the same technology became a necessity. Federal Express package tracking is a similar story—what started as a FedEx exclusive quickly became the cost of doing business for UPS, Airborne, and everyone else. This pattern seems to recur fairly frequently across many different technologies and industries. The first mover (or first few movers) employ a differentiating technology, while the mainstream and laggard companies later scramble to employ the very same technology as a competitive necessity.

Another reason for the blurry line is that the Chief Information Officer (CIO) has typically managed both. This can be very challenging, because Necessary IT focuses on reliability and lowering costs while the differentiation plays are about competitive differentiation and, eventually, incremental revenue. Few parts of any business are simultaneously focused on operational and strategic goals to this degree. Every business unit has plans for the future, but they usually involve doing their main mission better. IT has two main missions with two different sets of goals and timeframes for success.

The problem with blurring these lines is that the best practices for each kind of investment are often different. One key difference is that new technologies that can emerge as differentiators typically emerge as integrated solutions. Citibank worked with technology vendors to create custom hardware and software for its ATMs, because that was the only way to solve the problem. FedEx implemented highly customized tracking systems, even handing out PCs to their largest customers so they could access a private network. It would not have been possible to solve these problems when they did except by employing highly integrated and customized technology. These intgrated systems were tested carefully with users and modified until they succeeded. There were no best practices for either of these applications because they did not exist. Tight integration coupled with relentless user testing revealed the best practices required for the business breakthrough.

Necessity spending is quite different. Necessities must be implemented in a standard way. Best practices already exist and standards exist for how to implement the technology—tech vendors may even have arrayed themselves into various niches that each plug into the standard implementation in a well-defined way. As ATM technology matured, for example, vendors emerged that used PC hardware with standard operating systems and industry data networks to drive down costs with the same user satisfaction that originally demanded integrated systems. Package tracking was made ubiquitous through Web interfaces that obviated the need for private software and networks.

Because the two kinds of IT are so different, they require different measurements. Necessary IT can often be justified on the basis of cost savings while Differentiating IT usually has a revenue justification. Differentiating IT typically requires more innovation in your business justification as well. How did FedEx justify package tracking? Maybe they thought they'd reduce phone call inquiries (reducing costs) but they actually created a feature that caused customers of other delivery services to switch.

Many companies are using the Chief Technology Officer (CTO) to handle the differentiation mission of IT. The CIO handles the necessity spending while the CTO takes the rest. To do this job well, the CTO needs to have a much closer relationship with senior management, especially as it involves corporate strategy, than other IT managers. The CTO must deeply understand corporate strategy to help technology meet that strategy, but the CTO is also responsible for influencing corporate strategy based on technology trends and capabilities.

When you examine success stories such as Citibank and FedEx using differentiating technology as a competitive weapon, the more optimistic among us want to run out and start several big Differentiating IT projects. But the truth about these projects is sobering. While IT undoubtedly has the power to differentiate, as these successes show, IT is a two-edged sword. For every successful differentiation project, there are probably hundreds that fail. Each one that fails puts a financial drag on the corporation that reduces or even eliminates the benefits of the successful Differentiating IT projects.

How do you use metrics to limit failure and accentuate success? Most successful CTOs follow similar paths:

• Select projects carefully. Study the risk and business value up front to pick the winners.
• Implement small projects quickly.
The bigger the project, the bigger the investment, and hence the more money at risk. Keeping projects short and cheap allows you to test the riskiest factors first.
• Monitor constantly. Here's where your metrics are critical. What are you testing for? Customers switching from the competition? More revenue from existing customers? Know what you want to learn and design your metrics to learn it fast.
• Act on your results. This might be the trickiest part of the project. You must be willing to quickly double down your bet by expanding what works. Or ruthlessly kill what does not. Too often, successful pilots are left slowly running along and take years to become game changers. Even more frequently, someone's pet idea continues being tested long after it is clear that it isn't working.

Does your business distinguish between Necessary IT and Differentiating IT? If not, you may wake up to find out that you are doing only one. Or that you are doing both but managing both the same way.

Posted by mikemoran at 7:19 PM | Comments (0) | TrackBack