Hard Choices for PPC Advertisers: Part I

by Andrew Goodman

The basics of paid search advertising are fairly widely understood by many of today's guerrilla marketers. Savvy search marketers in some industries can count themselves lucky that their best tricks haven't yet been adopted by competitors, but the harsh reality is that it's only a matter of time before so-called best practices spread.

Online resources like ClickZ and Search Engine Watch Forums have helped get the word out on nuances like how to adjust bids to correspond with a "success metric" like cost per order (CPO). Indeed, seminars and tutorials offered by Google and Overture themselves can help advertisers come up to speed on some of the basics. My own report on Google AdWords tactics has been out for nearly three years now and has been updated several times. As I sometimes tell clients, when everyone else has become aware of best practices, it may soon be time to switch gears and adopt what might seem to others to be experimental methods or even "worst practices."

Choosing keywords appropriately, testing ads, sending searchers to tailored landing pages, and tracking conversions to sales, leads, or desired actions; most of us could do better at tackling these pay-per-click basics. But increasingly, advertisers are looking beyond the basics to a variety of advanced tactics. Here's just one example I recently came across.

The CTR-High-Position 'Lock-In' Hypothesis

This is a technique that some of you may do very well with in the coming year. I only regret not studying it more closely in the past, but truth be told, our ability to study the impact of ad position (how prominently a pay-per-click listing appears on the page of search results) on return on investment (ROI) is limited, since companies like Google and Overture don't pass through rich detail like ad position or cost in the referring URL attached to each discrete click (rather, they provide averages and aggregates in their reporting).

(In the interest of complete accuracy: my sources tell me that Google's "unofficial" AdWords API has just been released in a new version, and that this version does allow pass-through of information such as the cost of a discrete click, whether a click was generated by an AdSense partner or not, that click's ad position, and a variety of other rich data. However, unless you work for one of a handful of software companies who have a close working relationship with Google in this regard, or unless you manage your campaign through an agency that is able to hook you up with such sophisticated analytics, you'll probably be using the more basic metrics that Google makes available to 'the general public'.)

Recent studies by companies like Atlas DMT speculate on whether higher ad positions are better for ROI, or whether it's those lower positions achieved with much lower bids that work better. These aggregate findings are inconclusive; likely, it varies by industry and keyword. But there are special situations in which one can gain unique insight, as I'm about to show.

As any intermediate-level advertiser knows, higher clickthrough rates (CTR's) can help make a small bid bigger on Google AdWords, because rank on the page is determined by your maximum bid multiplied by CTR. Advanced advertisers have long speculated that in some industries, grabbing top spot - especially the 'premium' sponsorship spot above the search engine results (bold, with a light blue background) - gives you a disproportionately higher CTR in and of itself. This was a theory I might have encouraged in early versions of my Google AdWords handbook as I recommended bidding high at first, to "establish position," and then walking bids down gradually to save money.

Some advertisers believe that bidding very high on their core terms to keep themselves in first spot creates a snowball effect that locks them into the #1 position and would force competitors to bid counterintuitively high to dislodge them.

Case in point: a client who recently called me with a tough question.

He had recently acquired a competitor, and for the time being, he was operating both accounts competing head-to-head against one another for AdWords position. What he noticed and did not understand (he had forgotten Google's AdWords position formula, evidently) was why he had to bid so awfully high ($12 or more!) just to maintain ad position on core keywords for the recently-acquired site, whereas his main site was able to maintain a hammerlock on #1 spot without bidding any more than $5.

I asked the obvious question. "Let's take your most popular keyphrase on your main site and look at the CTR for the last seven days." OK. It was 6.1%.

"Now, let's look at the business you just acquired. You say it's stuck in about fourth position and having trouble maintaining that position. What is its CTR over the same period?" As it turns out, it was 1.7%. One might expect a drop-off in CTR as one goes down the page, but the impact of this drop-off was severe. The top-position ad, substantially similar to the ad running in third or fourth spot, enjoyed a CTR 3.6X higher than the lower-position ad. We don't know what the second-place advertiser was doing, but one assumes he might have been bidding quite high (let's say $8) to maintain his second spot. For my client's newly-acquired site to get out of fourth position and up into second place, he might need to bid double what the higher-placed advertiser was bidding (let's say $16). To be sure of gaining first spot, he might need to bid as high as $19 per click, a bid he had not even considered! (He had gone as high as $12 and was baffled as to why Google still put him in 3rd place even at that high bid.)

So there was no particular anomaly here. It was simply that the first-place ad was so prominent, and the display URL and ad compelling enough to build on that prominence, that it became nearly unstoppable due to the high CTR that resulted. The second-place advertiser was either bidding very high or also enjoying some positive effect from high ad position. Sure, these advertisers probably could have been dislodged if another advertiser had thought of bidding $20 per click for a few days. But who would have even considered that? Many would have rejected the idea out of hand.

It's become commonplace to hear of the strategy of "nibbling away" in expensive keyword fields to ensure that one does not lower one's ROI by attracting expensive clicks by "compulsive clickers." Many advertisers "nibble" at second, third, fifth, or ninth ad position rather than going after the costly top spot. I have noticed that this sometimes works quite well. On Overture, one of my clients gets 40% of their clicks (after many efforts to diversify!) from two main phrases. These incredibly competitive phrases are not worth entering into bidding wars for 1st or 2nd position - those wars are just crazy. I personally am fond of fifth spot. If you have a look at a page of Yahoo results, you'll know why. Fifth Overture position is good bang for the buck.

But as my client's example showed in graphic detail, a go-for-it strategy might pay off in Google AdWords if you can get unusually high CTR's from sitting in 1st or 2nd spot. As with any counterintuitive pay-per-click tactic, if everyone starts doing it, it won't work. And unless you have proof, you shouldn't put too much stock in it. Needless to say, the rare situation my client had - owning and having access to two separate businesses that were competing head-to-head for attention on the page - gave him the proof he needed that his main site had for some time been benefiting from "lock-in" in first ad position.

The connection between different ad positions and ROI is something you need to be constantly aware of, but finding the balance is far from easy. Another client works in a competitive B2C high-tech niche. Predictably, their core words are overpriced. We were able to justify hanging around in ad position 8 or 9, but even here, the daily spend was high and the ROI on these words far below that on less-competitive words. To reduce the drain, I dropped the bids even further, and ran the ads for a month in ad position 11 or 12 (on the second page of search results). Although click volume was still puzzlingly robust, the ROI got even worse. My only remaining option would seem to be to actually bid higher, to see if ROI improves in 5th or 6th spot. Unfortunately in such situations, with bids at nosebleed levels and competing against conglomerates which have a very different idea about the lifetime value of a customer, there are no good answers.

To sum up, speculation on whether ROI can improve with higher or lower ad positions is just that: purely speculative. I'll stick to the assertion I made in a very early piece of AdWords Handbook advice. If you're finding yourself dependent on low bids and low ad positions to survive financially - or worse, if your ads are falling farther and farther down the page because they're getting poor CTR's to go with your low bids - you may need to re-examine your fundamentals. Healthy companies can afford to go for broke occasionally with higher bids. Healthy companies are frequently found high on the page, and that reflects well on them, thus creating a snowball effect (or is that a halo effect, or both?). Well-thought-out, carefully-tested AdWords ads will pull higher CTR's and won't fall down the page. Smart advertisers also find untapped relevant keywords that others aren't competing for.

Low (less visible) ad positions can be a symptom that some aspects of your business model or campaign just aren't right. You won't always be able to fix those problems with rock-bottom bids. Sometimes, you can fix them with high bids, since a high bid is a signal to the world (and hopefully, to your prospective customer) that you really want and can handle their business.

I'd meant to review a number of other breaking developments in the paid search world, but I see I'm already out of words. We must do this again soon.

About author Andrew Goodman - Andrew Goodman is Founder and Principal of Page Zero Media, a Toronto-based company that helps clients build, track, and iteratively improve paid search campaigns. His forthcoming book, Winning Results with Google AdWords (McGraw-Hill), is near completion.