This is a guest post by Sam Engel. Sam is the Marketing Manager at BrandVerity, where he enjoys exploring trademark issues that come up in paid search.
Brand Considerations from Analyzing Over 6,000 PLAs
Product Listing Ads, or PLAs, can be a powerful tool for retailers to attract shoppers. Not only do they provide an added visual element to encourage clicks, but they also target buyers who have a strong purchase intent. The potential value of PLAs has been well documented, particularly on this blog.
They’re a new and very successful promotional platform. But what are some of the brand considerations to be aware of when it comes to PLAs? What kinds of issues can come up?
Inspired by CPC Strategy’s post about how affiliates can hijack your brand in PLAs, we decided to investigate potential PLA brand issues a bit further. In particular, we chose to inspect the PLAs that appeared on a set of branded keywords. Our keywords represented a variety of clothing brands. Each keyword combined a brand name with a clothing category (for example, “Nike shoes”).
Once we had our keywords set, we were ready to start collecting ads. Overall, we pulled 6,652 PLAs from a variety of advertisers. Let’s walk through some of the findings:
Good News: Ad Identity Theft May Not Be Terribly Common
One of the first things we looked for was any sign of affiliates (Comparison Shopping Engines in this case, also known as CSEs) appearing as the retail brands themselves. CPC Strategy had already identified some useful indicators of this behavior so we tried to follow their guidelines. Specifically, we looked for uncapitalized seller names as a signal of affiliate PLAs. For example, if the displayed seller name were “nike”, we would flag it as an affiliate’s ad.
“Re-marketing ads are also easy to spot in that they often use lower case seller names and are not accompanied by shopping reviews.”- Mary
Of course, some brands display their names as lower case anyways. eBay and several others sellers were therefore ineligible for this type of analysis. They were a small minority of the total PLAs, but had to be removed nonetheless. So, after removing those ads, what were the results? How common is the problem? Surprisingly, we found zero instances of uncapitalized seller names among the remaining thousands of ads.
Even with that encouraging find, we were still curious about what might turn up. So, we decided to take a new approach and look for a different signal. This time, we looked for anomalies in the tracking links associated with the PLAs.
Sellers each tend to redirect their PLAs through a specific tracking platform, making it simple to identify anything that deviates from their pattern.
For example, noticing that Nike always uses a specific tracking link structure (within the pixel.everesttech.net subdomain), we looked for any Nike PLAs that didn’t match the pattern.
Overall, we tested each individual PLA from the nine most popular sellers. These sellers accounted for 4,269 of the total PLAs in our study, well over half the total we collected. So, what did we turn up this time? Again, to our surprise, we found a whopping zero anomalies. That’s some good news for brands who use PLAs!
Brands’ Ads Don’t Appear Very Often for Their Own Keywords
We also looked for how strong brands’ PLAs coverage was on their branded keywords. Since all the brands included in our monitoring have their own online stores, we expected them to appear quite regularly with PLAs for searches related to their own products. For instance, if someone searched from “Calvin Klein pants”, we predicted that we’d find a Calvin Klein PLA most of the time.
We were surprised by the results. Interestingly, even though we exclusively monitored a set of branded keywords, this set of brands only appeared 47.8% of the time on their own keywords:
While we don’t know the inner workings of each brand’s distribution plan, it’s still somewhat shocking to see these brands show up on less than half of SERPs related to their brand names. It may not be as profitable for these brands to sell direct in the short term, but wouldn’t branded keywords be a great way to help build a brand’s online store? By appearing less than half the time, these brands could be missing out on some valuable sales.
Resellers’ Ads Outnumber Brands’ Ads by a Wide Margin
Brands are not only appearing infrequently on their branded keywords, but they’re also well outnumbered by their resellers—even when they do appear. Of course, some disparity is to be expected, given that a brand is only one advertiser and its resellers are numerous. But we thought that the numbers would be a little closer. Considering that there’s no double-serving restriction in PLAs, we thought brands might be able to control the SERP by showing up in multiple PLA slots.
Our hypothesis proved wrong again. The average number of these brands’ PLAs per SERP was only 0.76. By comparison, the number of resellers’ ads per SERP was 3.26 (That’s over four times as many ads from resellers!):
Even if you correct for brands’ poor coverage on their own keywords, you still find the resellers outperforming the manufacturer brands. If we divide the 0.76 brand ads per SERP by the relatively low coverage of 47.8%, we see that they averaged 1.58 ads on SERPs where they did appear.
That doesn’t sound too bad at all, initially. But let’s compare it to the resellers again. Let’s even assume that resellers appeared on 100% of SERPs. That assumption prevents any increase in their adjusted ads per SERP, leaving the figure at 3. Compare that to the mmanufacturers brands’ 1.58 and you find the resellers more than doubled the performance of those brands.
Brands Maintained High Positions When They Did Appear
Position may not be as significant in PLAs as in traditional paid search, but it can be a useful means of ensuring more exposure. PLAs that average a higher position are more likely to be shown, since they’re still eligible to appear on SERPs that deliver fewer PLAs. For example, considering that many PLA groups include only four ads, an average position of five would make it hard for an advertiser to appear.
Could this explain the low PLA coverage by brands? After all, maybe these brands were constantly on the bubble—limiting their exposure on the much more common four-PLA SERPs. Interestingly, that’s not what we found at all. On the contrary, the average position of brands’ ads was 2.32. Even when you consider the fact that the vast majority of SERPs in our data only held four PLAs, this is still a relatively high ranking.
Despite being outnumbered by resellers’ ads, brands held the #1 position nearly as often as resellers when they actually appeared on the SERP. Here’s a quick breakdown of the brands’ share of positions 1-4:
- Share of Position 1: 46.2%
- Share of Position 2: 47.2%
- Share of Position 3: 31.7%
- Share of Position 4: 32.7%
So, if position doesn’t seem to be holding their coverage back, what is? One possibility would certainly be budget. Typically, manufacturer brands won’t have quite the same budget as their resellers. A brand’s online store may be a fraction of the size—preventing it from spending large sums here. While it would be strange to see them simply running out of budget instead of adjusting their bids (and thus appearing in lower positions), stranger things have probably happened in the still-uncharted world of PLAs.
Of course, these are only some initial questions and metrics. There’s still plenty more to explore with PLAs—in their current format and as they continue to evolve over time. During the course of analyzing this set of data, here are just a few ideas that popped into my head:
- In terms of pricing, how do brands’ PLAs compare to those of their resellers? (For the same product, that is).
- How different would our results be on generic keywords, rather than the branded keywords we monitored here?
- Should brands specify Negative Keywords for their resellers’ PLA campaigns?
- Are prominent brands more likely to opt out of allowing affiliates to place PLAs on their behalf?
I’m sure you have some interesting ideas to contribute as well. What branding issues do you see with PLAs? What additional visibility do you think advertisers would benefit from? I’d love to hear your thoughts, and hope we all continue to ask demanding questions as our collective knowledge of PLAs evolves.