Ad Position: Do You Need to Rethink Your Assumptions?

A look at the results of a bid increase test to see if a slightly higher ad position justifies an aggressive bid.

No wacky stories this week – just numbers.

Below I share the result of a bid increase (ad position) test, with the goal being to find out whether that slightly increased average ad position on a core keyword generates enough volume at a reasonable ROI (define) to justify the more aggressive bid.

Did the Auction Change Significantly in 2008?

Bidding to ROI has always been a paid search marketing staple, but in higher ad positions, many of us have resisted throwing more money at keywords. Long-ingrained habits had us avoiding ad position number one because it was often prohibitively expensive, given the apparently similar visibility of the number two or number three ad slots.

But was that stance empirically based? The only way to really learn the click volume and ROI impact of higher ad positions is to occupy those positions by upping bids.

Moreover, it’s easy to misread the true meaning of an average ad position of, say, 2.7, especially in the current auction. Depending on the other contenders in the auction, time of day, geography, competitors’ daily and monthly budgets, and other factors, your ads might still be showing in ad positions five, six, or lower more often than you might think.

It gets even more complex. Quality Score affects not only ad rank today, but “eligibility” in the auction. Bid more, and your resulting Ad Rank powers through any situational weakness, overcoming (rank-wise) competitors who might rank well only under ideal conditions.

Furthermore, we typically fail to appreciate the CTR (define) gap between higher and lower ad positions. Our opinion that average ad position 2.7 is “OK” volume-wise might only be a guess. But 1.9 might be a lot more “OK.” In our real-world test below (average ad position before: 2.7; after, 1.9), CTR increased 22.2 percent between the first and last test period, and clicks increased 34.5 percent. Was 2.7 really “OK” if the business is seeking growth?

Finally, depending on the scenario, conversion rates can be higher or lower out of different ad positions. In certain cases, higher-position ads actually convert better than expected, offsetting some of the increased cost of the clicks.

So, don’t be surprised when you see the Google AdWords bid simulator predicting increases in clicks for bid levels well above your current bid, even though you may already be up there in a lofty average ad position of, say, 2.1. For our test keyphrase, the simulator shows impressions leveling out above a certain bid, but clicks keep ramping up at higher bids, indicating a higher and higher proportion of ads are served in the mega-CTR zone of ad positions one and two.

To find out the real-world impact of bid and ad position increases, simply raise bids by a significant percentage above your current bid, and compare all the key metrics week to week: actual CPC (cost per click), ad position, impressions, CTR (click-through rate), total clicks, conversions, CPA (cost per acquisition), and ROAS (return on ad spend), on a specific keyword or phrase.

Case Study

We ran this test for the (fictitious to protect confidentiality, but data is from the real world) phrase match “jumbo blue gumballs.” There are four test periods: baseline, bid increase number one, bid increase number two, and bid increase number three.

Bid Test: “jumbo blue gumballs”
 BaselineBid Increase No. 1Bid Increase No. 2Bid Increase No. 3
DatesApril 7-13April 14-20April 21-27April 28-May 4
Avg CPC$1.39$1.48$1.86$1.84
Avg. position2.
Conversion rate9.90%10.65%10.51%11.15%
Cost/conversion (1 per click)$14.08$13.85$17.67$16.50


The test results were eye-opening. An unsuccessful, cautiously-managed business certainly couldn’t laugh off a 32.3 percent increase in CPC (define) on a core keyword. Here, we saw that despite increasing bids, after “bid increase No. 3,” the actual average CPC fell slightly instead of rising. Moreover, CTRs, conversions, click volume, and conversion rate all increased, while the increasing CPA (define) leveled off and actually decreased slightly in the last period despite our higher bids.

ROAS (return on ad spend, not shared here) leveled off slightly, but only slightly. The end result in this case – at least for now – is a 51.3 percent increase in conversions based on an aggressive stance toward the auction resulting in a rise in average ad position from 2.7 to 1.9. (The last two periods were both 1.9 “on average,” but volume was still higher in the final period.) The ROAS is fully acceptable and the CPA isn’t running out of control. The revenue increase is close to proportional with the 51.3 percent increase in conversions, and associated profit increased significantly. Not only that, but there is a dramatic increase in new customer acquisition, which may lead to re-orders as customers for “jumbo blue gumballs” have a significant lifetime value.

Clearly, we’re not finished yet. Our last bid increase didn’t even increase the actual average CPC. Our latest increase (today) on the (undisclosed) maximum bid is only 9.5 percent, and I’d expect only a 5 percent or so increase in the actual average CPC. I’ll blog elsewhere about the progress, but in the meantime, the clear takeaways sync up with the assessment in the first part of this column: depending on the keyword, there is more upside potential in click, conversion, sales, and profit than you might think, in the seemingly trivial space above those 3.2, 2.7, and 2.5 average ad positions. In your line of business, on core keywords, you might even want to know how the economics of 1.8 differ from the economics of 1.5.

I don’t blame you if you just drifted off. But if you’re a certain type of business owner, you just zeroed in on the part about “51.3 percent increase in conversions.” Wide awake yet?

Originally posted on ClickZ – Ad Position: Do You Need to Rethink Your Assumptions?