Google AdWords Select Clicks/Day Inaccurate?

Clicks/Day seems to vary depending on number of keywords plugged in...

[Winooski]

I've been testing Google AdWords Select's "Select Maximum Cost-per click" tool, in particular, testing the Clicks/Day data for keyword optimization purposes. What I seem to be finding is that the Clicks/Day data often (but not always) changes depending on whether the keyword in question is part of a larger set I've input versus just a solo keyword. Am I doing something wrong, or is this a significant inconsistency? (Note: In the "Target countries & languages" page, I specify "English" and "United States" only.)

Here's what I mean. Here's a screenshot of a larger group of keywords I plugged in [natosongs.com], using a maximum cost-per-click of $15 to "amplify" the Clicks/Day to their maximum. (And no, these aren't keywords I ever target, and the ad's completely fake, too.) In particular, check out the Clicks/Day for the three marked terms, "auto" (370.0), "car" (1900.0), and "free auto" (3.2).

OK, now here's a screenshot of those three keywords once again [natosongs.com] using the same ad, but this time I've just plugged in the keywords alone without any others. (I've made one composite image of the three relevant sections so you could see 'em all at once.) This time, the Clicks/Day are 380.0 for "auto", 2000.0 for "car", and 3.4 for "free auto".

For any keyword, I would expect the click/day estimate to stay the same regardless of whether that keyword happened to be part of a larger set you were querying with or just on its own. Yes, futzing with the Max CPC would affect the click/day estimate, but in this case I maximize the possible amount to keep that variable steady.

...So what am I doing wrong?

[Winooski]

...So I assume that everybody's currently really busy just putting the Google AdWords Select "Select Maximum Cost-per click" tool through its paces. Hence, no replies yet.

Or is this not an important issue for keyword research?

[DrCool]

I don't think the small variance you are talking about will make that much of a difference. The numbers are estimates anyway. If you use them as a guideline the numbers can be very effective but I wouldn't use them as anything concrete to base results on.

[Winooski]

I'd love to shrug the variances off (and yes, they do seem small), but I keep coming back to the same problem: there shouldn't be any variances whatsoever. It's like taking temperature for 20 cities at the same time -- you wouldn't say, "Well, if we inquire about 20 cities' temperature, then New York City is currently 60 degrees, but if we inquire about New York City by itself, it's currently 59 degrees.."

I don't think this is a minor issue because real money hinges on this data. It's not just Google AdWords Select budgeting, too: you can use the Google tool to help decide which keywords to target for overall search engine marketing independent of whether you want to do the AdWords Select program.

Bottom line: Should we recommend targeting keywords based on popularity data that varies for no apparent reason?

[WebGuerrilla]

O.K.,

I've finally had some time to go back and play with the AdWords tool some more. Another inconsistency I've found is the differences in impression numbers reported by the regular AdWords tool and the AdWords Select tool.

If you take a your list of terms that you are running on Adwords Select and dump them into the original AdWords tool, you will see that on average, the orginal AdWords tool will report an estimated number of daily impressions that tends to run about 3 times higher than the number of impressions AdWords select shows for your ad.

When you consider that Google always boasts about their average clickthru rate being much higher that that of banners, it makes you stop and think. I've got one ad that reports to have an average CT rate of 2.2% and an average position of 1.1.

Based on the data AWS is giving me, the ad is being shown about 1000 times per day, but the original AW tool estimates that the daily impressions for the term is 3800.

So are they only showing my ad a third of the time, or are they only counting every third impression in order to maintain an inflated CTR?

[sparrow]

I thought this statement out of Google's Adword Select Help Screen might help

How does Google keep my ad below the daily budget limit?

"Google tries to show your ad evenly over time so that you reach your daily budget limit by the end of each day. That way, you do not collect all your clicks in the morning each day and then have your ad disappear in the afternoon and evening.

Once you select your keywords, you will see a recommended daily budget limit. If you set your daily budget limit to this value, your ad should appear every time someone searches for your keywords. This maximizes the number of clicks you receive. If your daily budget limit is lower than the recommended amount, Google will only show your ad some of the time in order to keep your clicks below your limit. However, your ads will be delivered evenly throughout the day."

[WebGuerrilla]

None of the phrases I checked had daily budget amounts set lower than what was reccomended. So based on what they say, My ads are being shown all the time.