Archive for June, 2006

A double-serving issue can result when you are bidding on the same keywords for the same website, within multiple campaigns. You’ll have keyword overlapping from multiple campaigns…and a keyword overlap strategy is needed. This is a sophisticated PPC strategy.

Here’s a hypothetical scenerio: Imagine a Google account structure that has a My Client Center (MCC) with 2 accounts. And, each account has the same campaign/keywords. Hence, keyword overlap. And to further complicate things, you have the need for the daily budget in each account to be used in full.

You might be wondering why structure an account like this in the first place. Imagine a situation where an organization has separate budgets ( or departments) with the same objectives ( or conversion points/ landing pages) . Your purpose for this type of account structure is the need to separate out the reporting for each account/department, and to use each budget. In this scenerio you’d have two separate initiatives with distinct budget lines. Each department would be bidding on the same keywords.

Google doesn’t allow an advertiser to serve an ad from each account at the same time. That’s double-serving. So, you have a double-serving issue, and a need for a Keyword Overlap Strategy.

This double-serving issue is particularly dangerous with Yahoo & MSN.

Here’s why:
Yahoo will delete the listings. But, you’ll be able to see the reason in the editorial status section of your account. It’s basically survival of the fittest on Yahoo.
MSN will de-dupe your listings by purging one of the ads/keywords without giving you ANY notice. Yikes!

Keyword Overlap Strategy solutions:
Strategy #1 is to just carefully watch the performance of each account/campaign, and optimize bids continuously. This is a manually intensive effort. Only really works on Google. But, there is a danger of low quality score, low impressions, and inactivated keywords. Depending on the size of the account, this might not be very practical.

Strategy #2 is to closely monitor the keyword performance & cost, and shift keywords from one account into the other in order to use each campaign’s budget. Kinda like a tennis match. You’d be pushing the the keyword from one account into the other. Watch the keyword cost, once it gets to a certain level, push it back to the other account. Back and forth until one or both of the campaigns end.

Strategy #3: On MSN , Looksmart, & ASK it’s based on the Display URL. So, if you add the directory structure to the display url, then you can circumvent the issue of keyword overlapping. However, Google’s double-serving policy is based on the root url - so appending in this manner will not help. There is really no way to circumvent their policy - the campaigns will need to compete against each other, and only one will serve on any given search.

Click fraud is a very real problem in PPC advertising for both the advertisers and the search engines. This is evidenced by the recent case of Google vs. Lane’s Gifts & Collectibles. So, what do the search engines do to combat click fraud? Google is diligently monitoring the click fraud problem with their Click Fraud team. Some of the other search engines are automatically discounting a percent of each months clicks in the form of a credit to your account. In case you’re wondering what percent of your PPC budget is getting consummed by Click Fraud, Click Forensics’s has a Click Fraud Index. Currently, the amount of click fraud across all industries is about 14%.

Click Fraud seems impossible to prevent. So how to manage it is the question. There are some third party click fraud solutions like: Click Defense, ClickDetective and WhosClickingWho. But, are they really necessary if you use your own Web Analytics software like Web Trends, Omniture, or NetTracker? If you use your own web analytics, you can see all the same information as these third-party click fraud solutions companies. Basically, you’ll want to monitor your web analytics for recurring IP addresses, Referrers, # of Pages Viewed per visit, Date & Time, and the volume & timing of the incoming traffic.

Most of the major search engines are sensitive to this problem, and will credit your account if you can prove the likelihood of click fraud activity. So, whether you use a third-party click fraud solution, or your own web analytics, the best click fraud solution is to schedule regular web analytic reports & resolve with the Search Engines if and when necessary.

Keyword Stemming in regards to PPC, is taking the stem of a word, and building additional words by adding a prefix or suffix and using pluralization.

Sometimes the words from the output of these keyword stemming tools don’t have any meaning. That’s because the keyword stemming tools are meant to strip a word down to the root and show different variations of the word form.
So, the solution is for the keyword stemming tools to cross-reference the output against a huge dictionary of words (like Merriam Webster Dictionary) to insure that the output is linguistically correct words.

An example of Keyword stemming for the word “search” is:
search
searcher
searches
searched
searching
searchable

Keyword Stemming tools are very useful for expanding keyword lists. Some of these keyword stemming tools will give you the equivilant of long tail keywords. My suggestion is to use keyword matching options to increase your conversion rates. Bear in mind, that most 2nd tier search engines won’t have the keyword inventory. So, keyword stemming is a better strategy on the large engines like Google & Yahoo rather than smaller engines like Ask.

Will one wikipedia link provide more SEO benefit, than a handful of links on a few web directories?
This is a question of quality vs. quantity. Which will drive traffic from more keyword searches, and/or more click-through traffic?

Even if only a handful of good web directories are used, like DMOZ (don’t hold your breath), Yahoo Directory, bCentral, and BOTW. Will one good Wikipedia page still outrank the target site from the collective directory submissions?

In a link building campaign both techniques should be used. But, I ‘m just wondering if given the choice, which is better? It’s really an academic question.

Which do you think provides more SEO benefit? One wikipedia link? Or, a handful of directory submissions?

Here’s how to generate a huge keyword list, and some thoughts on the strategy of keyword bucketing.

Start by generating your keyword list by using keyword research tools like Wordtracker, Yahoo Keyword Tool, Google Keyword Tool to generate a massive list of keywords. Once you have your keyword list, start bucketing your keywords into categories of similar terms. Think of it like segregating the terms into adgroups. This way, you’ll be able to write highly targeted & relevant ads for each keyword bucket.

Besides bucketing the keywords for relevancy, it also behooves you to segment the keywords by each stage in the purchase cycle.
1) Awareness Terms - Consumer is looking for information about products, features, and benefits.
2) Consideration Terms - Consumer has narrowed the criteria in their buying decision.
3) Buying Terms - Consumer is ready to buy.

Depending how you set-up your account structure, the above stages of the purchase cycle can be used in a tiered account structure.

Some other keyword segments to bucket:
1) Brand Terms
2) Competitor terms
3) Long Tail Terms - low volume, higher converting keywords. Increases efficiency of campaign.
4) Misspellings - Not heavy search volume. But, you don’t want to miss out. Not very competitive terms, and they could convert well.

And rather than just Broad Matching all your terms, take the time to use Keyword Matching Options.

You’ll need to decide a bidding strategy for each segment for budgeting purposes. For example: Do you want to spend 10% on Buying Terms, 30% on Consideration terms, and 60% on Awareness terms? The subject of bidding strategy is really a post in itself. So, I’ll just mention that the next step is a budget allocation for each segment.

So, what’s the benefit of keyword bucketing? 1) Keyword Bucketing will provide you with the greatest learning on your audience segments. 2) Your targeted ads should see an increase in CTR, and Conversions. 3) By bucketing your keywords into categories, you will increase the efficiency of your search campaign by eliminating many wasteful clicks.

You can SEO your webpages to the max, but if your webpages aren’t indexed it just doesn’t matter. Getting all your pages indexed is not as easy or as automatic as you might think. Many sites don’t have all their webpages indexed by the search engines for one reason or another. I’ll discuss some of these reasons in this post.

To begin with, you’ll need to determine how many webpages you have in your site. And, then, determine how many are indexed by the search engines. There are a couple of ways you can do this. You can ask your webmaster how many pages you have. He/She might know. Or, if you use a content mgmt system, you might be able to get the count from that.

Once you have identified how many webpages you have, then you need to figure out how many are indexed by the search engines. You can do this by using a site: query for a list of all the webpages. And to locate a specific page use the inurl: or the url: query operator. Yahoo has a cool beta Site Explorer tool that you could use also.

The next step is to calculate your “Inclusion ratio”. The formula is: Pages Indexed/Total Pages=Inclusion Ratio
A 100% inclusion ration is rare. Obviously, the higher the percentage, the better.

Here are some ways to get your pages indexed:

Inbound Links from other websites. Inbound links to your missing pages from authority sites, or really any other site that gets crawled regularly. It’s important to get your website deep crawled by the search engine spiders. Remember, your inbound links should be static text links. And if you are buying or trading links, there are a few things to be careful about.

Google Sitemaps: Great way to submit all your pages, and especially helpful for dynamic pages to get crawled. But, just make sure you remove any spider traps. Because, even if you send a spider to your webpage, if there is a spider trap, your page won’t get crawled anyway.

Dynamic URL’s should ideally be changed to static URL’s using a URL rewrite. Google has stated that they can read 3 parameters in the URL, but after that, no guarantees.

Robots.txt or Robots Meta tag: Robots.txt file can “disallow” robots to crawl your entire site. The Robots meta tag will disallow robots to crawl specific pages. If you had these on during development, just make sure you remove them when your site goes live. Otherwise, this is an easy explanation as to why those webpages are not indexed.

Javascript navigation is a spider trap. If you need to mouse over a Drop down menus to see it, then you can figure that the spider can’t do that, and those inner pages won’t get crawled. Text based navigation is the best navigation for SEO.

Pop-ups cannot be crawled. So, if you serve a pop-up with content you want crawled, then it would be best to remove the pop-up and change it to a new webpage.

An on-site Sitemap: This is a good way to get the search engine spiders to crawl all the links.

Requiring cookies: Search Engine spiders can’t accept cookies, so they are blocked from your content if you require cookies.

Requiring Javascript is a spider trap. Using Javascript is different than requiring it. All Javascript should be placed in an external file.

Requiring a Login. If you require anything more than a click, then the spider will not be able to enter and index your webpages.

You can see these problems for yourself in the text-only browser by Lynx, or if you don’t want to download the browser, you can use the Lynx Viewer.

Flash is a spider trap. A whole page of flash won’t get spidered. Google claims that with the BigDaddy update they can read some elements of Flash…but not all of it. So, if you must use Flash, just use some on the page, but not the entire page.

Finally, it is a best practice to Validate your HTML, and fix any errors.

What converts better? A Category page or a product page? It really depends on the ad. If you mention a particular product, then ideally the visitor should land on that product page. But, a lot of advertisers prefer category level pages because they offer a happy medium between the homepage and the product page. For example, if a searcher is looking for 35mm minolta cameras, the category page could be the right choice. If your site doesn’t have a category level page, then maybe you could direct visitors to a site search results page.

Some marketers believe that you need to minimize the amount of clicks or products on the landing page to increase conversion rates. But that might not be the case. Especially, if your brand or site are strong, the visitor might just buy something in addition, or other than what they came for in the first place if you allow them to click around.

Make sure your ad, destination url, and tracking code are pointing to the correct landing page. And don’t forget to make sure the keywords from the search term are on your landing page. Some sites even run a script that highlight the searched keywords so they stand out on the page.

Some Don’ts:
1) Don’t stuff the landing page with information, or with multiple conversion choices. Multiple conversion points are okay, but not multiple conversion choices.
2) Don’t assume the best landing page is the homepage.
3) Don’t assume that the best landing page is the Contact Us form. Especially if you’re in the lead generation business. Give at least a few paragraphs of information to provide the visitor with a reason to fill out the form. Try a hybrid contact page that contains a few paragraphs of content (with the Keywords).
4) Don’t assume that brief copy is the best copy. An extra paragraph with compelling words might be all it takes for the conversion.

After several landing page tests you will not only increase your conversion rates, but you should gain a better understanding of your visitors.