Hack42.Measure Paid Search Engine Marketing
Hack 42. Measure Paid Search Engine Marketing
Paid search marketing is one of the most popular marketing acquisition strategies in recent years, primarily because its effects are so easily measured. Taking the time to understand the "what" and "how" of search marketing can dramatically improve your return on investment for search keywords. Paid search engine marketing can be measured and evaluated by simply modifying your click-through URLs to identify incoming paid traffic from popular pay-per-click engines like Google and Overture. Using your web measurement application, you can then segment those visitors for further analysis and profitswhich, considering estimates that keyword costs will rise 30 percent between now and 2009, are worthy indeed. 3.7.1. How to Identify Paid Search TrafficDifferent levels of tagging will yield small, medium, and large amounts of data. Whichever method you choose determines the amount of intelligence you can gather about your paid search marketing. 3.7.1.1 Low data gathering strategy.The overall big picture view of paid search is simply the identification of your traffic as originating from paid search. This can be done with a single query name/value pair on the end of all incoming paid search traffic. If there isn't already a query name/value pair, just add ?source=paid to your landing page URL:
This segments your data into two distinct groupspaid versus organic. The amount of information that you gather from this approach is somewhat limited, is critical to differentiating paid from organic search [Hack #43]. If you can determine that people coming from organic search results are generating conversions, you can purchase that term for additional clicks and, hopefully, conversions. 3.7.1.2 Medium data gathering strategy.The next level of granularity comes with identifying which paid search engine generated the click. In this case, you will be appending information to the click-through URL that will include the name of the search engine from which you are buying the keyword. For example: http://www.mysite.com/landingpage.htm?source=Google 3.7.1.3 Large data gathering strategy.The recommended method for measuring paid search engine marketing is to measure at the per-keyword level. Whilesomewhat more time consuming to set-up, this level of granularity is most actionable and will help you best decide where your money is spent well and where it is wasted. Each term, for each engine, would have its own unique query name/value pair. You can do this one of two wayseither by embedding all the information you want to track in the URL or by building a lookup table that associates the information after-the-fact. If you're going to embed information, you'll use URLs that look like the following in your paid search interface (when you set up the keyword campaigns):
Alternatively, the value can then be related back to your web measurement program with all of the related information in it, possibly through data integration [Hack #32]. An example of this is:http://www.mysite.com/landingpage.htm?campaignID=112, which, when combined with data from an external spreadsheet (Table 3-5) would tell you thatcampaignID 112is related to Overture, the term "web analytics," and costs $1.05 per click.
Obviously, despite the additional work, the "large data gathering" strategy should be used to ensure that you'll be able to accurately determine what is working and what is not. Once you've set up your strategy for identifying the clicks, the next step is to start collecting data. 3.7.2. Which Data Should You Be CollectingThe metrics that you'll want to consider when determining how your paid search marketing efforts are paying off include many of the standard marketing measurements. 3.7.2.1 Response rate (click-through rate).The response rate for each of your keywords is simply the number of generated visits divided by the number of impressions served at the search engines. You'll need to get the denominator from your paid search application, and again, don't be surprised if the number of clicks they report differs from the number of responses measured on your end [Hack #51]. 3.7.2.2 Cost-per-click (CPC).Measure your cost-per-click simply by dividing the total cost of all of the clicks for that term in the time period by the number of requests in that time period. This should roughly match the per-click cost as reported by the advertising system that you are measuring. If not, then examine your system for possible points of loss [Hack #51]. 3.7.2.3 Conversion rate by search engine and keyword.Measuring conversion rates on a per-engine and per-keyword basis is critical to understanding your success in paid search marketing. A keyword either drives success or it doesn'tit's that simple! You can compare the conversion rate across your paid search campaigns to see if similar terms across all of the search engines convert at the same rate. This is particularly useful when looking to see if a similar term on a new paid search campaign is going to be as profitable as it was in the past. To calculate this metric, divide the total number of clicks by the number of conversion events [Hack #39] for a particular time period. 3.7.2.4 Revenue (or loss) by search engine and keyword.Measuring the profit or loss per click is the ideal metric that all marketers should strive to measure. Especially if you're able to calculate profits based on margin, having this information will allow you to make truly actionable decisions about increasing or decreasing your bids for specific keywords and phrases. To calculate, simply subtract the total cost of your search campaigns from the total revenue driven directly by paid search, again being careful to differentiate paid from organic search [Hack #43]. 3.7.2.5 Lifetime value of visitors from paid search marketing.Considering the amount that many companies spend on search marketing efforts, it is worthwhile to keep track of the lifetime value [Hack #84] of visitors who originate from paid search efforts. If you discover that paid search visitors are very profitable over the long run, you can feel that much better about your per-click costs. 3.7.3. What Do You Do with the Data?So, in your web measurement tool, you know which search engine a visitor came from, which keyword he clicked on, how much you paid for that term, if that term converted, and what the value of that conversion was. Let's run through an example with each type of data collection for a term that may be purchased to help sell this book"web analytics," purchased on Overture. In the first small measurement example, you would knowwhich clicks came from paid search and could track those from the referring URL to the conversion. You can then add that term to your Google AdWords account or other paid search engines. 3.7.3.1 Low data gathering strategy.If you're using the low data gathering strategy, you can measure the total paid search engine profits and determine your organic search term [Hack #43] profits:
3.7.3.2 Medium data gathering strategy.In the medium measurement example, you would know which paidsearch engine generated every click. This is particularly useful for search engines that display results from multiple paid search publishers (like Dogpile.com or other meta-search engines). Also, it allows you to evaluate which search engine you should be spending your time and money on:
3.7.3.3 Large data gathering strategy.In the large measurement example, you would know for every click which paid search engine generated the click, how much you paid for the click, what the total cost is for all similar clicks on that engine, which clicks resulted in conversions, and what the value of each conversion is. This data can then be compared to the total profit/loss with the original click as the source to determine what the value per click is for that term (Table 3-6). (Note that keyword #1 is losing two cents per click ($0.02) while keyword #2 is driving a profit of $0.75 per click.)
Remember, for any individual search engine, the sum of revenues minus costs for each keyword you're buying is the profit (or loss) for that engine. Analysis like this arms you with the knowledge that keyword #1 isn't profitable and should be examined to see if there is anything that can be done to make it profitable. Keyword #2 is a profitable term and should be purchased on all paid search engines for that price or lower to evaluate the term over a broader distribution network. Dylan Lewis and Eric T. Peterson |
EAN: 2147483647
Pages: 157