When Google transitioned Google Shopping over to Product Listing ads back in October 17, 2012 (or on June 11th of this year for a lot of countries outside of the US) many advertisers saw a serious lift in performance. We’ve talked about how to setup and optimize PLAs around here a lot since the transition.
For eCommerce advertisers, PLAs have been great. We totally love them, and their return has been incredibly strong in almost all cases. For eCommerce small businesses, though, it’s been more of an onerous process (as moving from “free” to “not free” can often be). Those used to getting all of this valuable Google Shopping traffic for free have had to make some adjustments to paying for these clicks.
One such small business owner got in touch with PPC Hero about his experience with PLAs and he brought up some really interesting points. The full thrust of his thoughts on the matter is contained in this AdWords community post. I wanted to take a moment to recap the conversation that the two of us had about PLAs and some possible strategies that could be leveraged by businesses both small and large as they make the most of PLAs in their own accounts.
Here are some of the more salient points that Burt Shulman brought up in his post:
I am charged for every click on my product listing ad. However, some Googlers click on my ad 5 times in a minute.
All of these clicks are occurring while customers price/benefit compare across the different sites that are showing PLAs alongside him, and it’s easy to imagine the sort of toll that all of those extra clicks could take on an advertiser that’s closely watching each and every click.
Consider that the advertising benefit I receive diminishes with each repetitive click, and so should the charge.
Mr. Shulman took a look at each of the clickers to his site and broke down their performance. Here’s what he came up with:
It’s clear that first clicks are driving a majority of his sales, which means those are the most valuable clicks to him. A solution to this setup that Mr. Shulman proposes is a cascading fee for each subsequent click:
[Advertisers could be] charged 100% for the first click, 10% for the second and 0% after that and also, the click “counter” would be reset, let’s say, every hour.
I think it’s a great idea to have diminishing subsequent click costs. But it’s a great idea in the same way that I thought the hover board in Back to the Future II was a great idea. It’s awesome, and I love it, but I think we’ll be waiting quite a while (forever, maybe?) for it to actually happen. That movie came out in 1989 and was set in 2015. I’m guessing functional pink Hasbro hoverboards won’t be coming out within the next two years.
Google gives us plenty of discounts (such as the basic setup of their auction with the AdWords discounter and their monitoring of invalid clicks). As they just transitioned to this paid PLA model in the last year, I think it’s a long shot they’ll adopt such an aggressive discount structure for repeat clickers, which has to be disappointing for smaller advertisers. The hope, though, is that PLAs will still be profitable for all advertisers in the long run.
What would be great, though, is if they rolled out remarketing for search for PLAs. Advertisers could create an audience list for previous site visitors and bid those audiences down accordingly. Since Mr. Shulman’s main concern is clickers within a very short time frame, there’s a possibility that the AdWords system wouldn’t even have time to add that user to your audience, though. As RLSA continues its transition from “former beta” to “thing that we’re all used to and take for granted” I hope to see it spread out even further. One logical place for it to arrive would be PLAs, so here’s hoping that it arrives sooner rather than later.
What I think is some good news for Mr. Shulman and other smaller advertisers like him, though, is the users who continue to come back via multiple clicks are still making sales. Of those click-happy users within PLA ads, they are still those that convert. This surely will vary depending on how competitive you are with your pricing and benefits, but it’s definitely good to see that those users that click a lot can also convert a lot. I don’t have access to all of his data, so it’s possible that the conversion rates are low, but the presence of conversions at all is heartening.
One final thing that I’d also like to address from Mr. Shulman’s post on the AdWords community blog is the inability to optimize PLAs. What may be surprising to some companies new to PLAs is that you can add negative keywords (even though you don’t target them by keywords). Matt Umbro has posted about this strategy before so please refer to that for more specifics.
Have any of you seen similar behavior on your PLAs? Within Analytics I don’t know of a way to identify users by unique IP addresses, so this problem may be one that is somewhat difficult to track and could require some gymnastics on the back end. How do you feel about a tiered pricing strategy within AdWords for repeat clickers (which even happens in regular old search campaigns)? When Burt Shulman contacted me he mostly wanted to get a conversation started about this phenomenon and I’m hoping to do just that. Let us know your thoughts!