Setting Up a Bid Experiment in Google Adwords

A Google ‘Experiment’ allows for advertisers to implement a specific change at any level of the account for a certain percentage of the time. These changes can range from keyword additions to bid changes to landing page changes.

The Experiment functionality allows you to utilize both a Control (what you had prior to the experiment) and Experimental variable (the change). For example you can have an ad with Landing Page A show 50% of the time, and then a separate ad using Landing Page B the other 50% of the time. This data can then be segmented and utilized for optimization purposes. This video focuses specifically on using the Experiment tool for bid adjustments: testing bids to see which Max CPC allows for the highest level of efficiency.

In this example, we have a starting bid for a group of keywords (this is our control), but we suspect that more clicks can be gained with a lower bid (we also assume budget is limited here). To see the effects of a lower bid, we can use Google’s Experiment tool to implement this new bid decrease, but ensure it only applies in 50% of ad auctions.

This gives the advertiser a sort of safety net with changes, as opposed to diving head first into new bids, mitigating some of the risk with these changes (lowered conversion rate, worse ad rank, etc.). It’s a more cautious and controlled method of testing bid adjustments.

Video created by Straight North a PPC management company