In Part 16, we talked about how to broaden PPC campaigns beyond the obvious high-intent search inventory. This week, I’ll reopen an age-old conversation: can the Google Display Network perform for your business?
I won’t bury the lede. I’ve seen Display perform well—so much so that the ROI on a carefully curated list of Display placements is comparable to Search.
For real? Yes. For one of our larger clients, I just sorted their Google Ads Placements file for all time, filtering out Remarketing and YouTube. Conclusion: based on a fairly specific type of retail intent, and diligent effort in discovering, curating, and bidding on managed placements, the ROI was fantastic – equivalent to what we realize in Search. It’s just the volume that’s a bit disappointing: about $1.5 million in attributed revenues over seven years (a drop in the bucket for this client). The true revenue picture, and the Lifetime Value (LTV), likely look much stronger, $4 or $5 million, I’d imagine. A good result – if you’re a performance marketer and you prefer concrete proof of performance rather than fairy-tales about brand lift.
The average conversion rate was pretty insane – from a Display ad to a purchase, over 4%. That’s highly atypical for Display.
But that hardly ever happens. Most of us will never enjoy performance anything like that.
And unfortunately, that volume and general interest level have waned over the years – in this account and others we manage. It can be more effort than it seems to be worth.
Figure 1: Efforts were made over a three-year period, for a total of 607 conversions, for a travel client. But ultimately Display became unreliable, and was a pain in the arse to boot, so it was shut off.
But – sorry, publishers – that doesn’t mean we’re going to capitulate and “just throw money” at the channel. It’s no fun searching for the needles in that haystack when they’re jostled and moved around, given the strong and ever-changing performance picture across our total paid media spend.
That the Display component of paid media accounts tends to revert to the non-performing norm shouldn’t surprise us. It’s probably just a matter of a daunting haystack:needle ratio. Sometimes, we solve that problem. Much of the time, we can’t.
But as the saying goes, the time to buy is when blood is running in the streets. Since Google’s many recent changes seem to have killed many performance marketers’ will to live in the GDN, I’m optimistic that there are deals to be found here.
Think small
How should we start? Since this is science, and we’re supposed to be learning: at first, think small.
In many ways, not much has changed. Publishers are publishers, advertisers are advertisers, and bid control is a good thing. If you’re good at PPC generally, you’re really good at bidding.
Every time someone talks about how big something is, it may distract you from how small bits of intent work for your business. So, again: think small. SMALL IS THE NEW BIG.
Remember what I said about granularity in Part 15? Even as we think about expansion, we need to keep the “small bits” in mind, as they ultimately drive the aggregate performance. Two seemingly opposed ideas, fused into one powerful driver of performance: that’s integrative thinking.
Managed Placements—where are they now?
After starting it, Google took years to refer to their own display network as Display. Initially, it had been known to all and sundry as the Content Network. Until YouTube and Doubleclick were part of Google, Google essentially identified as the anti-Display company. “We’re engineers; we care about the user experience; only extremely granular and relevant advertising is morally OK” – that type of thing. “Don’t be evil,” eh?
The principles are evergreen, but the workflow currently lies in a broken state. For some advertisers, carefully-curated lists of “Managed Placements” went away when the new Google Ads interface rolled out. They’re now empty, and you have to click on “Where Your Ads Showed” to see, er, where your ads showed.
In other cases, happily, your curated Placement lists remain intact. But over the years, the value of different Placements shifts and typical CPC’s may change as well.
Figure 2: From deep in the archives: an old Display campaign with oodles of placements. Pictured: a rash of nonperformers. Rather than pausing them, we *removed* them. For spite. Fun cure for insomnia over a peanut butter and raisin sandwich at 4:00 a.m.
It also seems harder to manage placements now. It used to be pretty seamless to add them from the so-called “Automatic Placements” side, where they’d become Managed. Now, you’ll need to make a little more effort to manage (add) placements in the interface. Was this change by Google thoughtless or part of a plan? I can’t decide.
(Note to smarty-pants advanced people: you absolutely can build software tools to sweep your favorite placements into a managed list! I intend to do just that.)
Keen advertisers might have to rebuild their Managed Placements lists from scratch (although certainly, you should have access to old data in your Google Ads account or in Google Analytics). They may not perform the same. But aren’t you just a bit curious to see whether they do? I certainly am. In an upcoming part of this series, I’ll share a current case study – good, bad, and/or ugly.
Bottom line: if everyone else is demoralized, not trying, and proceeding on the basis of “let’s not optimize like that anymore, who the hell knows if it’s going to help our results, but let’s do this anyway,” there very likely lies opportunity amid the apathy.
Next week, I’ll share a concrete 10-point plan to set up a performance-based Display campaign.
By the way, it helps if you have a lot of time on your hands, and rock-solid Analytics abilities.
Read Part 19: Statistics & PPC: There's More You Can Do
