Making the leap to a much larger keyword set holds more than the promise of increased conversions; it also creates the potential for increased cost. Here's why.

A few months ago when I was enjoying a dazzling and deafening 4th of July fireworks display -- wondering if my hearing would ever return -- I couldn't help but think about our nation's founding fathers; specifically, Thomas Jefferson.

While he might best be known as the principal author of the Declaration of Independence, and as our country's third president, Jefferson was also a big supporter of innovation. In a letter written a mere 200 years ago to colonial entrepreneur and inventor Oliver Evans, the president said that "…ingenuity should receive a liberal encouragement." Jefferson's philosophy on the topic? He felt that innovation would benefit society as a whole. 

Sounds like TJ was onto something as innovation is also good for business, particularly search engine marketing.

Just over six years ago, Google introduced AdWords, and soon thereafter the other major engines launched their own pay-per-click auctions. What followed?  Not surprisingly, a flurry of innovation.

Once it became apparent that manual bid changes over a large number of keywords was humanly impossible, the landscape was ripe for solutions. First came simple rules-based keyword level systems (some of which still exist today). Next came offerings that globally optimized across massive keyword sets. 

Clearly, these were major innovations driven by changed business needs.

The evolution of such major industry-wide innovation happens from time to time as the market grows and technology is developed to fulfill the need. However, between these major events, truly forward-thinking companies continue to introduce a stream of innovative, ancillary developments in an effort to continually make improvements, whether it is to enhance a process, reduce cost or capitalize on a synergy.

Fortunately, there are some recent innovative methods -- both simple and complex -- that marketers may find of interest, particularly if they are scaling to a large keyword set while trying to keep costs to a minimum. After all, making the leap to a much larger keyword set holds more than the promise of increased conversions; it also creates the potential for increased cost in a number of ways.

The below innovations could help keep your campaign a lean-spender.

Landing page tracking
Today it is fairly common for paid campaigns to have large keyword sets, and when keywords proliferate, generally speaking, so do landing pages. Consequently, keeping track of landing pages -- their relevance, changes and removal -- directly impacts cost.

For example, when a landing page is removed, ideally the keywords associated with it are either redirected to another landing page or they are removed. Unfortunately, however, that is not always the case.

Thanks to recent developments, this problem can be remedied by an automated system that scans landing pages to validate that they still exist, and/or provide alerts when they do not. Naturally, though, since some landing pages drive considerably more traffic than others, a page should be checked with a frequency directly proportional to its propensity to drive cost. 

A step beyond landing page validation leads us to measuring landing page relevance. Ideally, a keyword phrase is relevant to the ad text, and the ad text is relevant to the landing page. However, when the relevance chain of connection is broken post click -- that is, when the landing page is not relevant to the creative description -- the overwhelming result is increased cost.

Sure, such a situation is easily remedied with a little human intervention: Once the page changes, the campaign manager is notified. Problem solved. Right? Not necessarily. This hands-on approach may work for some situations, but it's hardly feasible for large-scale endeavors.

Fortunately, recent innovation provides a more complex and automated solution to measure relevance between keyword, creative and landing page. The keyword phrase -- either in its entirety or in constituent parts -- can be detected and measured in the ad title and text, and on the landing page. If thresholds are not met, then an alert is raised to request human review. Even though the use of such a system may be more prone to error than a human reader, it may be the only cost-effective solution available to manage large sets of landing pages.

Inventory tracking
Another recent development that may be of particular interest to online retailers is an inventory-response system. In retail, it is a given that stock will become depleted from time to time. When it does, online PPC spend is wasted when bidding continues on keywords that cannot lead to a sale or back order. When this situation arises, the solution seems obvious: Turn off the keyword. Ah, if it were only that easy. 

The direct approach won't work here, as simply turning off the keyword will have an overall negative effect in most cases. Why? Keywords often contribute to sales across a set of products. The more generic or brand-oriented a keyword is, the greater its ability to drive conversions on a variety of products. Conversely, the narrower the keyword -- a model number, for example -- the more it will primarily drive conversions to a single product. Consequently, simply turning the keyword off could have serious ramifications.

Instead, online retailers can remedy the situation by temporarily negating past sales for the out-of-stock item, and then recalculating the bid plan. This approach will substantially reduce bids on a keyword that directly targets a certain product and only has conversions on that product, and will dampen bids on keywords with a wider variety of conversions.

Overall, history has shown that innovation comes in great and small steps, and search is no different. Major game-changing developments arise from time to time to address changed business needs; however, it's the ancillary developments made in the interim by innovative firms that can really make a difference. 

Charles Haggerty is senior director of operations at iProspect. Read full bio.