Unlearn 5 Misconceptions about AdWords Campaigns
Published: November 8, 2012
Author: Sean Marshall
Almost a year ago, I wrote a blog on why I prefer simple AdWords campaign structures. I’ve never been a fan of grouping and theming unless it was absolutely necessary, and given I see this type of grouping for all of our new customers, I felt compelled to write about it.
Well, a year later, and I feel like beating this drum one more time. So, I give you: the top 5 misconceptions about AdWords campaigns!
1) Granularity wins. WRONG! Granular analysis is imperative to fine-tuned AdWords campaigns, but this doesn’t need to happen at the campaign level. Product segmentation is usually the first place people start, but geos are most likely to drive the biggest gains. While your account might not benefit from geo segmentation (however unlikely), it’s much easier to build more complex geo-structures if you have a simple structure to begin with.
2) Turn off search partners. WRONG! The one place where Bing is ahead of Google isn’t a total loss for AdWords campaign managers – it’s just really tricky. Too often, I’ve seen people turn off search partners outright on Google. In these cases, SPs are an anchor weighing down performance, and eliminating those stats from the mix frees you up to push more aggressively on Google.com. That said, if the deltas in performance are consistent through your campaigns, a simple bidding hack can help you preserve the volume you were getting from SPs and to push more aggressively on .com. If you had a keyword driving 10 conversions a day w/ an incredibly high CPA, you wouldn’t shut it off, would you? Provided you have the option to control something with a bid and maintain some of the volume, you should.
3) Set budgets by network. WRONG! Some companies have strict budgets, and allocation of budgets between product lines is a requirement (this would be a case where more granular campaign structure is a must), but doing so across networks is incorrect. The number one rule of direct-response (not brand) marketing: spend the money where you can make it back. The holy grail of search budgeting is to run uncapped, and while this is not necessarily the case for everyone, the biggest profit driver should get the lion’s share of spend. There may be a few exceptions (like, say, spending your entire budget on brand keywords), but this one is mostly true across all programs. If you’re running search as an awareness driver (without some way to attribute credit back to it from leads/sales), you should reconsider how you leverage the channel.
4) CPA Optimizer is bad. Partially WRONG! We get it: you don’t trust Google to manage the bids in your AdWords campaigns. For search campaigns, I don’t disagree. That’s not because I don’t trust Google but mainly because search campaigns give you the transparency you need to make decisions on your own (or at least hand-control to more sophisticated algos). Between match types (tight query control), dimensions (geo, day-part insights), and the fact you’re mostly dealing with one site, there’s clarity in knowing what is influencing performance. GDN, however, lacks that transparency. The ultimate black box in search is tailor-made for CPA optimizer. The sheer randomness of placement selection and tangled knot of targeting makes it a prime candidate for automation by Google. Don’t get me wrong – I’m a bit of a control freak and would love to run this on my own, but if I can’t accurately predict what’s going to happen with the information available to me, I’m prepared to give up some control to Google. (It doesn’t hurt that it works, either.)
5) Keep adding more keywords. WRONG! Cannibalization is bad. If your AdWords campaigns are budget-impacted, stop adding more keywords. You’re already getting enough traffic to spend those dollars, so focus on what you’ve got and maximize volume within those constraints. You should still try to research and find even cheaper pockets of traffic, but running experiments like this in the wild, with the risk of cannibalizing what you know is winning traffic, is a risky proposition.
BONUS: Broad match is evil. WRONG! Not so much a campaign issue but worth noting. PPC Associates believes in query control, and introducing pure broad match is a scary prospect. Reality is, there’s no better query discovery engine than Google itself. Now, you’ll need to set aside the proper resources to monitor query quality and quickly move winners to exact/phrase/broad match modified, but if you’re needing to tap into volume, make it happen.
– Sean Marshall, Director of Client Services