Aug 62012

More Advertiser Controls, Please

Google’s geotargeting controls have really improved over the years, which makes for a more targeted user experience as well as easier set-up for advertisers.

Namely, advertisers can now pick among several options in Google’s settings, to geotarget the right users:

  • People in, searching for, or viewing pages about my targeted location
  • People in my targeted location
  • People searching for or viewing pages about my targeted location

And, advertisers can now add multiple locations in bulk, which can be a huge time-saver.

Massive Accounts

However, one headache with geotargeting is the need for duplication to achieve certain advertising objectives. To be thorough in a geotargeted build, ads are often launched for each targeted location, which can mean exponential growth of account sizes.

While Google has opened up its limits to accommodate this type of scenario, massive accounts become cumbersome when maneuvering through the user-interface or AdWords Editor, which becomes slow to ‘get recent changes’, to push changes through, and in its search functionality.

Ad Copy Customization

But, as it stands, duplication is necessary for customizing copy by location -- though the nifty location insertion option, (which will dynamically include store locations in the ad copy) can help get-around this for some local advertisers. To take advantage of location insertion, an advertiser must first enable location extensions, which requires inputting their locations into AdWords, or integrating the AdWords account with a Google Places account.

As such, online-only companies are not able to take advantage of this Google offering.  So, if a business is just interested in targeting for delivery purposes: demographics, weather patterns, catalog circulations, top markets, or other variables, re-launching by location is still needed for custom copy-writing. No workaround here, for now at least.

Location Bidding Controls, Please

Another reason to re-launch for each location-target is for geo-performance differences. With different keywords, advertisers can bid locations differently, as the data suggests.

It would be neat to have the option to launch a local ad once and target all the locations in the settings. Then, advertisers could bake in bid pushes and pulls to a base-line max bid in Google’s system. A thought would be a Google bid modification option that would allow advertisers to select specific locations, and then set a push or pull overlay on the default keyword bid. (It would act as a bid subset/tier to be modified, by location.)

So, if New York traffic behaves better than Dallas, TX traffic, advertisers can put in additional pushes to max bid when targeting NY and pull-backs when targeting Dallas.

More Controls, beyond Location Bidding

Another thought, in this same light…it would be nice to have domain bidding controls on the advertisers’ end as well. Having pushes and pulls layered on top of the max bid default would allow advertisers to shuffle funds to spend more on sites likely to convert.

As for matchtype, device, and other targeting options, these are probably more manageable at the unique keyword-level, as launching duplicates shouldn’t make an account massive and hard to manage. At RKG, we are taking advantage of device, matchtype, (rather than network), time of day/day of week, and other differences already. We have adaptive bid management technology to account for these and other conversion-influencing variables.

STOP: Won’t eCPCs Do the Trick?

The enhanced cost-per-click (eCPC) feature in AdWords aims to increase conversions while maintaining or lowering advertisers' cost per conversion.  How does this work? Google’s system tracks click and conversion patterns and compares them to past results. Using this data, Google’s eCPC functionality will increase a max CPC bid by up to 30% when it sees high-converting potential. And, it'll lower a keyword’s max CPC if it determines a conversion is not as likely.

The variables included in eCPC adjustments are: “the user's geographic location, browser, operating system; the time of day that the user was searching; specific content present on a Display Network site; and many additional factors.” So, location is already a variable baked into eCPCs, isn’t that good enough? I’d say, no.  First, eCPCs require conversion tracking to be enabled. Not all advertisers are willing to pass along conversion data to Google.

And, while eCPCs are a step in the right direction, smart advertisers know more about their business than Google.  Marketers know when seasonal changes are going to happen, when and where a catalog drop is planned, and other marketing objectives that influence the paid search space arena. It would be nice to bake these additional influencers into bidding without having a massive and sluggish account with thousands or millions of duplicated keywords.

I’d love to hear your thoughts. Do others see a benefit to additional location and/or domain bidding controls at the advertiser-level?


14 Responses to "More Advertiser Controls, Please"
Great stuff, Joy. I'd throw in a couple of other challenges with eCPC: First, all conversions are not created equally. Lead values differ based on context and those value differences may not be clear at checkout, margin structures vary, lifetime values vary, etc. Second, all the value isn't transacted on websites; call-center spillover and store spillover need to be factored in in ways that eCPC wasn't designed to do. Third, many websites have multiple success metrics with complex real value relationships. Last, "conversion" isn't the goal for all advertisement so "letting the machine figure it out" isn't a good solution. What you're really getting at here, is a need for a fundamental structural change in AdWords, and you're absolutely right.
Yes, great additional points on the limitations of Google's internal conversion tracking driving the eCPC. We'd love to get those controls on the advertisers' end so we can bake in all known variables, including things like life-time value calculations, call-center and in-store sales.
Hey Joy, These are great ideas and it would be great if Google would in fact go in that direction. Duplicated campaigns are a pain to manage - you either have to rely on software to push changes to all duplicates, or you better get paid by the hour. Especially when it's only done to take control of bidding, I find that duplicating campaigns often defeats the purpose. If you get too granular, individual campaigns gather so little data that isolated bidding hurts more than it helps. An advanced bidding software could evaluate data on a more aggregated level, but that's not exactly available to everyone. For bidding purposes, some sort of bid adjustment option, as you outlined, would be perfect. We already have that for dayparting, so why not for locations? And while we're at it, why not for the other targeting options as well? About eCPC, or Conversion Optimizer, which basically offers the same features, I actually like the basic idea here. If the only thing you care about is your CPA goal, these tools are perfect to account for regional variations. However, aside from the fact that a CPA goal isn't exactly state of the art anymore, eCPC and Conversion Optimizer only act on data that the individual campaign has gathered so far. With location based bid adjustments on the other hand you could initialize these adjustments using all the data on consumer behaviour you've ever gathered. I guess that would be a nice selling point for a big agency ;) Anyway, I hope you guys use your influence with Google and tell them about these ideas...
BienTek says:
To our business, the only solution to the problem of massive duplication has been to create customer software to manage many locations. Internally, all variables are properly accounted for better than Google would be able to do, since our software has access to our proprietary sales data - not available to Google, of course. Still, everything that we do could be solved by Google through an improved interface on their end, or some special API capabilities.
Martin, when RKG opens its first office in Germany, we'll be giving you a call!
I'm looking forward to it, George :)
@ Martin - right on!! @ BienTek - yes, Google can certainly make this game easier.
Kenny says:
At least one of those is on it's way... ;)
Mark says:
Mark from Google search ads marketing here. We definitely appreciate the challenge and opportunity around optimizing locally targeted campaigns at scale. We're listening closely here. Please keep the good ideas coming! In that vein, Joy, can you please clarify what you mean when you talk about domain bidding controls? You said: "Another thought, in this same light…it would be nice to have domain bidding controls on the advertisers’ end as well. Having pushes and pulls layered on top of the max bid default would allow advertisers to shuffle funds to spend more on sites likely to convert." What do you mean by domain? And are you talking about bids or budgets when you talk about shuffling spend?
Thanks for listening, Mark! Per the domain piece, I am talking about partner network sites, since these behave differently (often converting at a lower rate) than the core traffic. And, I am talking about bids, as we try and avoid budget-based spending --- in order to maximize return and keep the control in the advertisers' hands. The hope would be to layer in bid pull-backs on those syndicated sites that are not converting, and push harder on those that are showing a strong return. Does this help answer your questions?
Thanks for listening, Google!
My thoughts exactly :)


Check out what others are saying...
[...] More Advertiser Controls, Please, RKG Blog [...]
[...] a high level, Enhanced Campaigns is step in a direction RKG has requested a number of times in the past.  Instead of having to duplicate campaigns and keywords multiple [...]

Leave A Comment