Search Optimization Update: Google's New Adword Program Tested | Update: Google's New Adword Program Tested |
| Wednesday, 20 June 2001 | ||||||||||||||||||||||||||
#040501-PSE - Google's New Adword Program - We test 6891 impressions to determine the true cost per click In the April Edition of the Journal, we posed a (doubly) provocative question: How did a belly dancing diva help our research team test an effective new advertising medium, achieve clickthrough ratios at 300 times the industry standard, and generate (eventually) qualified site traffic at $.22 cost per click? More pointedly, we focused on 6 questions...
You can view the full experiment here. Google's Adword program proved to be a viable option for the careful marketer... but the operative word in this sentence is "careful". Google can sometimes help you achieve traffic at far less than the prevailing bid per click at GoTo -- Google can also "eat your lunch". With Google, you pay by the impression, not by the click. The challenge is to create an ad so seductive it captures significant attention. What can you say in 2 short (35 character) lines that will compel a searcher in "autoscanmode" to STOP and click? We recently learned what NOT to say... How did a truly compelling ad completely bomb, failing to generate a single click?Here is an ad that (we like to think) was carefully smithed in our "copy forge". But here is an ad that was less than disappointing. Yahoo Store "How To" Book The ad was anchored to these 7 search terms: Yahoo Store The ad received 249 impressions, but did not yield a single click from Google... (though it would probably perform well at GoTo *1). Here is our scientific, scholarly, learned, conclusion: NOBODY was interested. We infer this weighty conclusion, in part, from studying the results of the individual terms:
It would seem that the people who searched "Yahoo Store" and "ecommerce software", were looking for something other than a book. In fact, the most relevant term, "Yahoo books", garnered zero attention. What was the cost of this education, besides tedious labor, heart wrenching disappointment, and the necessary therapy? Just $1.67(*2) We adapt fast. We reasoned that our market wasn't looking for a Yahoo book they were looking for a shopping cart - which infers they were looking for information on the right shopping cart. We changed our ad and our terms. Our new ad read as follows: Yahoo Store Shopping Cart Our new terms included: shopping carts We also kept: Yahoo Store The results? Our click through ratio jumped to 4.06%. Was this adjustment profitable?Well our current conversion ratio for the offer page remains at 7% (nearly 3 times the industry standard). (*3) This means that for every 100 people who visit the Yahoo Store page, 7 actually place an order. We are currently capturing $2.19 for every single visit. If you navigate the slippery math, Google cost us $.36 per click for the (highest yielding term) "Yahoo Store". So we are capturing $1.83 in gross profits per visit. While this is a solid return, our most effective marketing channel is proving to be yet another source: online discussion groups. Traffic from these sites has helped us to achieve (remarkable) conversion ratios as high as 42%. Case in point... Yesterday the offer page received 24 visits from one such group; this attention promptly resulted in 10 sales. In the Lab Edition for the Google Report, we discuss in more detail how to achieve a strong yield. You can visit that report here. In the meantime, there is one more question, completely off the subject, we should ponder. Will online publisher's, themselves, pay for content?Dear Reader... There was a reason the last issue of the Journal had it's Section 2 material protected with a ridiculously vulnerable "scramble" code... We were trying to answer a thorny question: Would publishers pay $7.95 to get the material they need (on how to sell content) or would they "cheat" the system? My... did we glean some fascinating data. We'll be sharing it, soon, in an upcoming Mid-Month Update. We're smiling.
#070501-MPM - We Test Marketplace Manager: a service that utilizes Artificial Intelligence (AI) to simultaneously post your products in Yahoo, eBay, Amazon, and 700 other markets. Which is more effective: spending money to drive traffic to your product, or spending money to put your product in the midst of someone else's traffic? This research may truly challenge your current marketing strategy. #070501-AWA - We study 600 awards sites to test and select the top performers. Can you attract traffic by winning recognition? What does it take to win? How do you prepare your site? What is the easiest way to submit? OTHER EXPERIMENTS
Email us with your comments and let us know which experiments you would find most helpful.
(*1) The ad is so specific it can keep you from wasting money with a pay per click service. (*2) Google charges $15 per impression for the top position. (*3) This number is not discounted by the fact that the Yahoo Store page is also visited by marketing experts who are only interested in studying our approach, and who do not have any intention of purchasing. |
||||||||||||||||||||||||||
