The first row in the second picture shows the overall results for the ad. The second row shows the search engine results, and the third row shows the content network results. Overall, since I launched it about 27 hours ago, my "McCain is 95% Bush" ad scored 10,314 impressions (page views where it appeared), 25 click-thrus, and cost me $15.41.
Almost all of the click-thrus came in the search engine portion of the ad-buy. Despite having less than one-seventh of the impressions, it generated 18 click-thrus (compared to only 7 click-thrus on the content network). Further, 1,214 of the 1,241 impressions on the search engines came from only two search terms: "mccain" (858 impressions,11 click thrus) and "john mccain" (356 impressions, 6 click thrus). The other keywords that I used were not very effective and / or popular, though not completely so.
Clearly, the click-thru rate on the search engine portion of the buy is much higher--18 times higher, in fact. The people who click-thru are the people looking for information on John McCain (aka people who search for him on engines like Google). As such, I considered dropping the content network portion of the ad buy. However, given that it generated so many impressions, I decided to keep it. My belief--or at least my goal--is that the ad is damaging to McCain even if people do not click thru. After all, the ad repeats a frequent Obama and Democratic surrogate talking point: McCain votes with Bush 95% of the time. Just seeing it once again might be a good thing for us.
My results for my Palin ad were completely different:
This ad generated zero impressions overall. None. Zip. Zero. Nada. There could be a bunch of reasons for this, so I made several changes to the ad.
First, I rewrote the ad itself. Maybe Google didn't like the word "Liar." So, I made it a bit more bland:
Second, maybe people don't know how to spell her name. So, I added a few more keywords to try and catch some more impressions.
Third, I tripled the cost-per-click bid to $3. Maybe I am just being outbid.
Fourth, maybe no one is looking for information on Sarah Palin. This seems unlikely to me, but I can't figure out what else I am doing wrong. The McCain ad worked fine, and I actually based it on the same settings I used for the Sarah Palin ad. I can't figure it out. If you have any ideas on what might be happening, please post them in the comments.
If the Palin ad doesn't pick up in performance, I am just going to drop it in favor of two McCain ads. Commenter jeffbinnic suggested this in yesterday's thread:
For each ad campaign, always create at least two versions of your ad and have them run simultaneously to see which ad gets the most clicks. Every few days, drop off the least popular ad and create a new one to run against your most popular one to see if you can get more clicks. (In your campaign settings, under Scheduling and Serving, be sure to click on "Rotate".)
Sounds like a good idea. I'll give that a try soon.
Remember, you can run your own ad campaign like this. If you have already done so, how are your ad campaigns going? I am eager to see if anyone followed suit, and if they have any results yet. This is a process where we can learn collaboratively. |
