In the PPC world, numbers are everything. I use numbers and percentages in almost every optimizing decision. Numbers help me to determine which ad is more successful, what to bid on certain keywords, and when to adjust bids, just to name a few ways numbers aid PPC specialists. So when numbers contradict one another, it can really throw me for a loop. I find myself asking the question, “Which stat do I believe?” Below I cite two examples of Google statistical discrepancies.
Rarely Shown Due to Low Quality Score
“Rarely shown due to low Quality score” is a status of a keyword. It essentially means that the specific ad that is triggered by a keyword is not showing on Google SERPs pages, or is not getting any impressions. The reason why the ad is not being displayed to users is usually because the keywords in the ad group are not related. That explanation is simplified, but for this blog it will serve its purpose. Now that I have explained the status, look below and see if you are as confused by these numbers as I am.
I cut out the actual keywords to protect my client’s identity, but each line represents a keyword. Notice that the keyword that triggered the most ads in this ad group has a status that wouldn’t lead one to believe the numbers. If the keyword in question has the status of “Rarely shown due to low quality score” then how exactly does it have more impressions than all of the eligible keywords combined? Either the status is wrong or the number of impressions is wrong. In my daily routine I just have faith that the number isn’t the falsity.
Ad Average Position
Ad Position is a rather important metric to keep an eye on. On your typical Google SERPs page there are about 10 ads. There are usually three ads above the search results, and another seven or eight along the right side. The best PPC specialists aim to get their ads in the first three because anything below the first three rarely produces a click. Rookie PPC specialists tend to aim for the first page, meaning that it is ok with them for an ad to show anywhere in the first ten positions. While it’s rare that an ad produces a click if it’s not in the top three positions, it is unheard of to get a click if it isn’t on the first page so much to the point that Google has a special status for ads that don’t make the first page. See below:
The column on the far right is the ad’s average position. In another discrepancy of numbers we can see here that these keywords are triggering ads to display in the top two positions. However the status would indicate that these ads aren’t even appearing on the first page. I set my bids in this ad group at one dollar as indicated by the column in gray. So I must ask myself, “If my ad is displaying in the first position (the best position), then why is Google trying to get me to raise my bid?” In other words, why bid $3.50, $2.50, or $2.00 when I already have superb position with my $1.00 bid?
Does Google Know About This?
I think this is a ploy of Google to make bidders spend more money on higher bids. After all, this is where Google makes its billions. It’s either that, or a serious error in one of the algorithms. Either way, it sometimes makes the task of optimizing rather ambiguous. Plus, this also hints that any numbers Google gives may not be accurate. I am not saying I don’t trust any of the stats because of these two instances of discrepancy, but it definitely lowers my confidence in the whole Adwords program. I would like to think that Google isn’t trying to make more money, and that they just haven’t fixed their algorithm yet.
Have you seen any discrepancies in the data that Google returns? How do you determine which is correct? Let me know what you think in the comments.
This entry was posted on Tuesday, December 7, 2010 and is filed under Pay-Per-Click.
- DragonSearch Achieves Premier Partner Status with Google
- The Halo Effect of PPC
- Don’t Lose Your AdWords Search Query Data, Start Archiving Now
- The Growth of Programmatic Advertising Platforms and Processes – What Are We Measuring?
- Understanding & Preparing for Seasonal Changes in Search Volume