Google Adwords

Hi Acornia,

Just curious, how long have you been testing AdWords before making this post?

I'm not sure what point you were trying to make in your post, but I can say with some confidence that you must know very little about Google AdWords and even less about statistical analysis.

You refer to "keywords" and not "search terms". You may be focused on the wrong reporting dimension. In AdWords "keywords" are used to target "search terms", it is the search term metrics that tell the true story, not the targeting mechanism (keywords). It was not evident in your post that you understood the difference between a keyword and a search term.

Different search terms will have totally different audiences, often with very different search intentions. The data you referenced seems to reinforce that point. You weren't comparing apples to apples, more like apples to oranges. looks like a classic failure to control the variables of your experiment.

You make a vague reference to Google "given access to companies in the background who are running algorithms". Duh... Of course they have. That is the whole point of the AdWords platform. It doesn't have any purpose outside of that function. It is a self-serve, real-time, auction based advertising platform that companies use to advertise their products and services, using any algorithm they choose as part of their marketing strategy.

Congratulations, you have discovered AdWords and it's intended purpose, just like millions of others before you. Welcome to the party.

Now, if we can just teach you how to perform statistically valid data analysis you just might be able to succeed in the intensely competitive AdWords advertising platform. In your example you included much too little data to be considered statistically valid. The results have a high probability of being completely random with such a small data sample. To draw any conclusions from that small of a data sample shows a total disregard for valid methods of analysis.

I realize that this forum has a lot of members that are trying to learn how to become a marketer, though most fail to even grasp the basic definition of marketing. There's a difference between marketing and selling. The masses on this forum seem to be focused on online selling, and often resist the very idea of marketing. Those two separate disciplines have completely different goals.

While selling and marketing work together to deliver results, the each have a different central objective. The primary objective of selling is to get a sale. The central objectives of marketing are to improve the scale and efficiency of acquiring customers.

In your post you made reference to impressions, clicks and ad positions as they relate to keywords (not search terms). You also mention that the ad was written specifically to include the term "free", and no mention of the word "cheap" in the ads targeting term "cheap". The results seem to reflect the differences in the message match aspect of your ads,and nothing more complex than that.

Your reference to algorithms not be able to determine what humans can. Well, that seems absurd to me. Algorithms are nothing but rules that are created by humans. If a human can understand how to discern something then they can create an algorithm based on that understanding. This is a central concept in everything Google has done.

AdWords isn't fake, it's as real as it gets.

I think you need to spend a little more time learning about how the AdWords platform works, and how to perform valid statistical analysis before jumping to any conclusions about what is actually happening in your ad campaigns. You seem to have a few misconceptions.

Feel free to start asking questions and learning from the more experienced members here. Many are willing to help and have invested the time to learn the platform. We can demystify the platform and help bring some clarity to your understanding of how it works, provided you can invest the time and keep an open mind to learn.

HTH,

Don Burk

Ok thats rare, i need to re-read this.

"So who's doing this, our judgement is that Google has given access to companies in the background who are running algorithms, we would not expect it to be Google directly as that would be plain idiotic."

That seems like a wild shot in the dark, do you have any evidence to back this up ?

I am a bit confused by your whole opening post. There are various reasons why free might be placed lower than cheap. Some might be that there is more competition for free, and another might be that Google feels that cheap is what you are offering due to reading keywords on the page, where free might not be mentioned, or even mentioned less (less relevance).

I am not sure if I am understanding your posts correctly, but seems to me you are jumping to wild conclusions, based more on assumption than actual information.

In my experience Google DOES at times make idiot mistakes, I have run into them myself several times, I guarantee you they do. Not everyone at google is as sharp as everyone else, and people from time to time are less sharp then they are at other times. Sometimes people break things unknowingly and other people don't know exactly what they have done, and sometimes algorithms overlap in ways they shouldn't, Google's algorithms are so large now I have heard it said several times that no one person at Google knows everything that is contained in them.

Acornia, I too believe you are jumping to wild conclusions. Worse is that you just don't have enough information and seem to base your conclusions on just two clicks. As Don said, not statistically valid.

You are basing your judgement on 2 clicks and just 39 impressions. Also, that is average position, not actual position. Those 2 clicks could have come when the ad was in the #1 position. Sometimes competitors only run ads at certain times of the day or they have a small budget and when that runs out, the ads stop.

The ad could have had zero clicks all day at much lower positions, then had 10 impressions at #1 or #2, and received 2 clicks. So although the average position was lower throughout the day (or whatever time period this test was done it), the clicks could have come at higher positions.

Without more data, who knows?