In light of the recent updates to device bidding in AdWords, columnist Kevin Lee suggests ways in which Google might build on these changes to make the platform even more valuable to advertisers.
Hey, Google — thanks for giving us bid modifiers for tablets and smartphones (mobile devices with full browsers). It’s great that we can once again set bids intelligently based on the data around the relative value of the searcher and the relative relevance of our offer to the searcher.
Make no mistake: returning this measure of control to us is a huge deal. For many advertisers — particularly larger ones spending in excess of $100,000 per month — it’s well worth the hassle to calculate and re-calculate the bid modifiers by device type. I’m guessing you knew already that larger advertisers like extra control (because they have data sets large enough to mine information about bid modifiers).
Of course, you also know that this will increase yields (you’ll make more revenue per SERP [search engine results page]), so it’s a win-win. Plus, many of the more sophisticated marketers also use bid modifiers based on location, time of day and audience.
Here’s the problem, Google. I don’t think you went far enough. So please hear two additional, modest requests on behalf of the marketers buying clicks from you.
With operating system available as a bid modifier, device control won’t be needed. Why? Because there are major situational differences between the smartphone, tablet and desktop searchers — and that’s just the beginning of audience differentiators. For tablet, smartphone or desktop, audience differences are often characterized as being due to where the search is located in the “buying funnel” (which is a bit of an over-simplification, even if the buying funnel still exists in its original form).
For many marketers, the differences by operating system are far more useful given that the demographics and psychographics of iOS smartphone or tablet users vary quite significantly from those of Android smartphone or tablet users.
For example, at my company, before the implementation of Enhanced Campaigns, we could target both by OS and device (before there were hundreds of choices). For some of our largest advertisers, we cloned high-volume ad groups into two separate campaigns (since bid boosts weren’t available) and set bids separately by OS. The bid management system almost always bid more for iOS than Android because conversion rates and — in retail — average shopping cart sizes were larger.
So Google, please consider returning this targeting lever to us.
With header bidding and the launch of “native” ad formats within programmatic media, now may be the time to revisit SEM as an RTB (Real-Time Bidding)-based marketplace. We understand that fast SERPs are really important; several years ago, the lag time in programmatic display was far too long to seriously consider RTB.
But a lot has changed in the last five years, even within AdX. With RTB for native ad formats, advertisers — though their DSPs (demand-side platforms)– could use any targeting you provided (e.g., keyword, geography) and combine that with both first- and third-party data to make the most informed choices about bids.
Not ready to make a change of that magnitude? Then why not create a new ad unit in the SERP, particularly in desktop, where the percentage of white space is huge (especially after your right rail elimination earlier this year).
There are several things Google could do on desktop and tablet SERPs that would monetize that space, with all sorts of fun ad formats that consumers would also love (or at least tolerate). After all, Google still uses that space for entertainment and travel queries, for example:
My guess is that Google might well be thinking about video ads, carousel ads, and perhaps even ghosted brand backgrounds as candidates to appear in that white space.
Advertisers will appreciate getting as much control as you can give them. After all, access to each and every one of the segmentation levers is something all search marketing teams should at least think about. Do the analysis and you might be surprised: when the device type bid modifiers are combined with location, time-of-day and audience, you’ll have a lot of control over who sees your ads in their SERPs.
Of course, not all surfers and searchers are equally important. As a marketer, all you need to do is know where to look for the data, and you can confirm that fact. Your business intelligence data will likely tell you that you have a “Pareto Distribution” (otherwise known as the 80-20 rule or “Power Law”) when you measure the value of all your customers based on many different segment comparisons. The same holds true when you buy media.
How do we justify this extra hassle in analysis mathematically? Well, think of it this way: many marketers are given a fixed budget and some marketing KPIs to hit (conversions, calls, store visits, sales ans so on). With a fixed budget, you always want to buy the most valuable clicks (as defined by your KPIs), and as budget increases, you’ll find that the new clicks you’re buying (if you focused on the best clicks first) have a lower value to you.
But by adding additional targeting levers, you have an opportunity to cherry-pick the best from any audience. The bigger you are as a marketer, the faster you can accumulate data about the various targeting segments (device type, location, time of day and audience) and apply this data toward your campaigns. This can provide an unbeatable advantage in the dynamic, always-changing battle for SERP visibility when it matters most.
OK, Google. I’m done. The ball is now in your court — and I hope that you’ll return it soon!