Right Rail Ads Post-Mortem – Two Weeks Later

In case you missed the first post, I made a few predictions about the loss of right rail ads and really wanted to see if I was right or not. Now that we have over a full week behind us now in the after side era, thought it’d be high time to see what sort of data was kicked up when the sky fell. I took a look at all Elite keywords to see what I could find in terms of position trends, volume shifts and CPC roller coasters.

Once it gets down to conversion and shopping shifts I had to trim my dataset before my laptop started crying. Fair warning, this is only a week’s worth of data so take it with a grain of salt, but trends are pretty crystal clear.

CPCs Increased As A Result

Position 1-3 were largely unaffected, with CPCs across all Elite clients remaining stable when comparing the data two weeks prior to the change vs. one week after. In fact, pure position 1 CPCs (top brand terms) went down a touch.
Keywords in position 3.1-4 were affected the most with an across the board CPC increase of 22%, while 4.1-5 increased 15%.

Screen Shot 2016-03-08 at 5.00.07 PM

To be honest I’m not 100% sure if this is a result of the change, a result of willy-nilly bid changes as I forecasted in the first post, or if that part of our dataset includes the first few days of a month when everyone’s budget replenished. In any case, there is a decent lift in competition for the now coveted 4th ad slot.

Despite The Loss of Ad Slots, Inventory Remained Stable

We saw no material difference in click or display volume across clients, as it essentially remained flat. Broken record here, but this represents a sample of Elite clients where positions over 7 represented less than .5% of impressions. We didn’t forecast losing much, and reality matched our predictions.

The change of impression volume and location from ads moving was noticeable, but not shocking. Top 4 positions now represent 95% of volume vs. 92% in the past. There was a slight increase in top of page volume for position 1.1-2.0, while others remained flat. Impression volume for keywords in positions over 5 dropped 60%, though that may just be us doing our job. Click distribution remained somewhat flat; there was a slight lift to the new below the fold ads, but we’ll get to that later.

Google indicated early on they‘d only show four ads in the top box for “highly commercial” queries. We haven’t dug in to find indicator terms to isolate exactly what that means (maybe good for our next post), but we were able to establish a volume baseline; about 25% of position 4 impressions coming from top of page.

For what it’s worth, there was a net increase of ~15% for impression volume in position 4.

Click Through Rate Increased 17%. That’s A Lot

Yes, that represents overall CTR across everything. Maybe we just had a really great month of SQ (Search Query) reports, or maybe the change from side to top is going to take an SEO’s lunch. Time will tell, but the impressions that we are receiving now (while flat overall) will go further, so you can safely anticipate a lift in volume over time.

Looking at individual positions, we saw a slight drop in CTR for positions 1-2 (down 5%), further reinforcing the need to coat your ads with several layers of durable extension paint. CTRs for position 2.1-3, 3.1-4 and 4.1-5 all increased, with position 3-4 seeing the largest increase.

CTRs For Bottom Ads Are Better. A Lot Better.

Overall CTRs nearly doubled for other ads now that the bottom of page transition is complete. This confirms the ”no we swear this is a good move” rhetoric we got from Google, not that we ever doubted them.

We’re seeing a strange inverse in CTR trends at the bottom of the page now that things have moved. Conventional wisdom says the higher position, the better CTR. Conventional wisdom is sometimes wrong; I’ve read a number of examples/cases that finds organic performance is better at the bottom of a SERP when compared to being position five.

Screen Shot 2016-03-08 at 5.12.45 PM

I suspect there’s some sample bias here, that our “poorer performers” were forced out of the 7 range which is why we see the big bump in position 7. Perhaps I’m drawing correlations that don’t exist. Who know’s.

Conversion Rate Improved (Mostly) Making Up For CPC Lifts

Conversion rate improved nearly 10% in aggregate for bottom ads vs. side ads, mostly making up for the 15% increase in CPCs. We did see a 5% increase in CPAs as a result, but again I expect some bias here regardless due to the beginning of the month.

I suspect the increase in CPC will be mitigated over time and we’ll see net CPA/ROAS remain stable as the dust settles.

No Significant Movement In Shopping. But We’re Seeing Blips

In the first post, I highlighted the vastly superior performance of “Other” shopping ads vs. top of page. Matt from Google agreed with me, therefore I’m the smartest man alive.

We’re not seeing a significant shift in impression volume at the moment. About 0.7% of total impressions shifted to other, though there were swings of about 4% during the post-sidebar era. <1% of total impression shifted to other; I suspect there’s testing afoot.

Screen Shot 2016-03-08 at 5.09.11 PM

TLDR: No Big Deal, Right?

If you skimmed down here, go back to the top and read. It’s even not that long of a post!

No, the changes weren’t dramatic chipmunk worthy, but they still represent a major shift in performance of some pretty popular positions. You can anticipate a nice healthy increase in volume and performance alike from some of your mid-level terms over the next few weeks.

While positions over 7 fell off the 1st page, they didn’t die completely. They only got cut in half. And half of nothing is still nothing. So maybe they did die. Whatever.

An increase in CTR of 16% across SERPs should be pretty concerning to folks in the organic space, and frankly to advertisers as well. I’m not saying these results are instantly stealing 16% of traffic from organic results, but there’s certainly been a migration as a result of this change; however significant or insignificant is yet to be seen.

Ranting and data aside, this is generalized and averaged across a large dataset in a short timeframe. Do not take it as the end-all be-all of all things top vs. side. Evaluate for yourself, and make conscious decisions about what’s best for your business!

Blog Categories
Paid SearchSearch Engine OptimizationConversion Rate OptimizationShopping & FeedPaid SocialPerformance DisplayCompany UpdatesIndustry NewsConferencePressEventsUncategorized
Blog Archives
2016 2015 2014 2013 2012 2011