Calculated Risk
Filed Under (Adwords Conversion Optimizer) by Jeff Hudson on 19-02-2008
If you’ve been hesitating on experimenting with the Adwords Conversion Optimizer now may be the time to try it. The Adwords blog team is pushing their case and educating everyone on the process and setting expectations. 2 key points I’d like to point out from their post:
# Keep in mind that if you choose a CPA lower than the recommended maximum CPA bid, you are likely to get less traffic than you did with your old CPC bids. It’s a good idea to start with this recommendation and adjust based on the results you observe.
As we all should know what our acceptable CPA range is, you will get a good idea right off the bat if this will work for your campaign or not. If your recommended max CPA bid is out of line with your metrics, you risk losing traffic if your actual max CPA bid is too low.
However, if those 2 are fairly close and you decide to give it a shot, you do have something to fall back on:
If you choose to opt out of the Conversion Optimizer, your campaign will revert to the previous CPC bids you were using. (So, there’s nothing to stop you from giving it a try!)
Even if it didn’t reset itself to your previous bids, you should be backing up your campaigns via Adwords Editor before any major changes, so you could simply override the campaign changes.
So, there’s the case for giving it a shot, as well as 1 item to look for when you’re going through the setup process.
Providing Google with your conversion data is like showing your hand in a poker game.
Christ, find a programmer to do the job and save yourself and your industry.
There is a reason why Google is pushing this so hard. Please don’t fall prey to their rhetoric; having Google Analytics is bad enough.
Anonymous J -
Is it bad to show your hand in a poker game? No wonder I keep losing.
Seriously though, my skepticism with google’s intentions are well documented all throughout this blog. No doubt allowing Google to see and then make decisions based on your data is a dicey proposition.
I guess since you won’t identify yourself you tend to fall on that side of the paranoia fence?
Shhh..I hear someone coming.
[...] Calculated Risk - Quick thoughts from Jeff on giving the AdWords Optimizer a chance. One key takeaway is the reminder that before testing any optimizer, “you should be backing up your campaigns via Adwords Editor.” Share It: These icons link to social bookmarking sites where readers can share and discover new web pages. [...]
I love the idea, but didn’t get good results. The campaign I tested gets about 1800 conversions a week. I set the max CPA above the recommendation, and still wasn’t getting enough traffic. Just turned it off after a week, and conversions are back to normal.
I think they need to switch from max CPA to avg. That’d work much better. I’m sure they’ll eventually get it right, but I’m staying manual for now.
Thanks for the feedback Adam. Always good to hear about real results, good or bad.
I’ve talked to at least 7-8 different people who’ve tried C.O., and have tried to find every instance of folks who’ve posted results online after trying it. I’m eyeballing here, but no more than 10-12% saw improved results and at least 60% saw decreased results.