At SMX East yesterday, during the Google Enhanced Campaigns: What We’ve Learned So Far session, a collection of the best and brightest PPC professionals came together to discuss insights and lessons learned from Google’s “new” approach to managing ad campaigns.
Larry Kim from Wordstream was the most positive on the pro-enhanced campaigns side. He reminded us that back in February, we all thought the sky was falling. But it’s still up there. He said some surveys companies were showing that CPC was rising, in some cases up to 20%. In actuality, it fell 6% in the second quarter, showing that you have to consider the source when you see “data.” He then showed us some data from Wordstream’s clients that support his hypothesis.
Brad Geddes of Certified Knowledge said he’s less of a fan of enhanced campaigns, and explained the problems they create and how to solve them. The biggest thing is the inability to see quality scores segmented by mobile vs. desktop. His solution (echoed by other panelists) is to segment by device.
Make sure that a low quality score is not a device-specific issue. Look at bounce rates by device, and don’t assume that just because you’re using responsive design you don’t have to worry about this. Remember, if you change your ad to mobile/desktop, you will reset the ad. Brad’s not saying you shouldn’t do this, but you need to keep it in mind.
In that vein, you need to be using mobile preferred ads. You can check a box to show that it’s a mobile preferred ad. But remember, too many ads can be a problem. Remember, if you have 53% good, you also have 47% less optimal. So remove excessively poor ads by device to make sure your users are seeing more “good” ads.
He illustrated this point by showing two ads with their data, and asked which one was best. It seemed obvious that it would be ad 1. But when you segment by device, you see that ad 1 is best for mobile, ad 2 best for desktops. The key takeaway here? Segmentation.
Also on the “enhanced campaigns are wonderful” team was Alastair Dent from Periscopix. He’s not quite as enthusiastic as Larry, but believes that the positives significantly outweigh the negatives.
One thing that is absolutely not true is that enhanced campaigns will save you time. They really don’t. You can now spend a lot more time on targeting and segmenting, and in many ways, you have to. For example, if you want different multipliers on keywords, you have to separate them out. Now you have no choice.
When optimizing campaigns, you should think of context, not device. Are they “in” (sitting somewhere, keyboard attached, ready to do research), or “out” (mobile)? This makes more sense in terms of what Google is trying to do. “Mobile” doesn’t necessarily mean phone, and “desktop” doesn’t mean chained to a desk.
Think first of keyword. But now you can think of it in terms of: proximity, are you open, what device are you using, what time of day is it? Also think of it in terms of persona. And it’s beyond bids: what landing pages can you deliver, what actions might searchers want to take? What ads do you need to present to meet their needs in these different situations?
Susan Waldes of 3Q Digital enlightened us with changes Google just rolled out. They’ve launched estimated cross-device conversions. In other words, they’ll now show you estimates of how many conversions you can expect regardless of what device the person uses to access your site first. For more on the details, read this article.
Susan’s advice to using these new metrics is to implement shallow conversion points, like email collection or app download to determine where visitors ultimately start engaging with your site.
Google also plans to implement click-to-call tracking and in-store-tracking as enhanced conversions. In-store tracking will use Google Places and geo-location extensions when searchers have it turned on. But, Susan asked, is there enough data to make this valuable, and will it freak people out? Scale will be a challenge. Google will need to be able to record at least 50 daily cross-device conversions for the estimates to work.
A couple of additional points on this new feature are that the default will be 30-days latency, but it will be customizable between 7-90 days. It’s expected to be rolled out to all accounts by November 1.
She ended with a series of questions:
Overall, the session seemed to focus on the importance of user engagement. Understanding where and when your users visit your site, and digging deep to find out the specific combinations of elements that work.
It’s interesting to me, as primarily an SEO practitioner, that as SEO is becoming less and less granular (with the loss of keyword data, especially); PPC is becoming much more so. Yet, the end goal for both is the same, and it focuses on the most basic tenets of marketing: define a target market, determine where and when the customer is, and deliver a great experience.