For the last couple of weeks, I’ve had my AdWords Display Network lab coat on. I needed volume and more conversions for my Google AdWords account, so I decided to set up and run a variety of keyword-driven campaign experiments.
What I learned can save you a lot of time, money and frustration.
What Are Keyword-Driven Campaigns?
Keywords are the most common method for targeting in the Google Display Network. In a keyword-driven display campaign, the main factor in determining which landing pages appear for your ads is your choice of keywords inside each AdGroup.
Keywords can be combined with managed placements, topic targeting and audience targeting to refine the targeting for those types of campaigns. For my experiments, I focused on AdGroups that only use one or more keywords and no other methods of targeting.
Experiments, Challenges And Results
Google’s display network has changed a great deal over the years and only continues to change and improve. I started by setting up my keyword (only) driven display network campaigns and AdGroups like I have in the past, using small groups of keywords.
Challenge #1: Relevant Traffic. Google was doing a very good job at sending me relevant traffic, so at the beginning, I only had to add a couple of topic exclusions to refine the traffic. It didn’t seem to matter to Google if I used one keyword or multiple keywords in an AdGroup, the traffic was still very relevant.
Challenge #2: Optimizing Bids. Google has only recently allowed advertisers the option for keyword-level bidding in the display network. Since this was a new toy to play with, I started my bid optimizations by using the AdWords User Interface (UI) to do bid changes at the keyword level. *Currently, that is the only way you can see performance at the keyword level for display advertising. If you go to AdWords Editor, your only view for performance in keyword-driven campaigns is at the AdGroup level. Even though you can change bids at the keyword level in AdWords Editor, you just can’t see impressions, clicks, cost or conversion data.
Challenge #3: Reporting Delay. Google initially reports some data, but not all data, at the keyword level in your UI. You have to wait about 48 hours before you get the complete picture. I have talked to other advertisers who are also seeing keyword-driven campaigns that have some reporting (impressions, clicks, cost, conversions) that can only be attributed to the AdGroup theme and not to any specific keyword. The reporting delay complicates determining if cost and/or conversions were attributed to a specific keyword or to the AdGroup as a whole.
Keyword Bids May Not Work Like You Think They Do
As an experiment, I set up new campaigns with AdGroups that only contained one keyword each. My thought was that I could now optimize at the AdGroup bid level. As I was setting up the new campaigns, I made a very big mistake.
I set up the new bids at the keyword level so the old (much higher) bids at the AdGroup level were still set. When I checked in on my new campaigns, I quickly noticed that my cost-per-click was much higher than the keyword level bids that I remembered setting.
What I learned from my bidding mistake was that in my case, Google defaulted to the higher AdGroup level bid even though there was only one keyword in my AdGroup and that keyword level bid was lower than the AdGroup bid.
Challenge #4: Optimizing Conversions. Although it originally appeared as if I could optimize for conversions at the keyword level, the delay in reporting left me with a dilemma. Because of the volume and spend, I wanted to optimize within hours, not days. I knew I could use a combination of data from Google Analytics and the UI to make quick adjustments to my AdGroups, but if I wanted to leave the AdGroups themed using multiple keywords per AdGroup, it would be a minimum of 48 hours before I could make any adjustments.
Challenge #5: Keyword Cannibalization. I inherited a very old display network campaign that performed well in spite of how it was originally set up. I quickly saw the traffic and conversions drop when I created new AdGroups that contained the same keywords used in the original campaign. As soon as I deleted those keywords from my new AdGroups, the traffic volume and conversions went right back up.
Disclaimer: After discovering the cannibalization issue, I knew there were more experiments to run, but I still haven’t had the opportunity to experiment with fresh campaigns that use AdGroup-level bids and keyword themes where keywords overlap across multiple AdGroups.
Themes Or Individual Keyword AdGroups? The Answer
Unfortunately, the answer to the question is not which strategy is best. It seems to be more like: how much time do you have to spend trying to find the right balance between keyword bids and AdGroup bids? It also may depend on how much money you are willing to spend and how much patience you have with the delay in reporting for themed AdGroups.
Because of the competitive nature of my industry and sheer volume of keywords I was working with, for right now, I have chosen to move toward the one keyword AdGroups strategy. My choice was based more on efficiency. I get volume and relevant traffic and I can now focus my optimizations on the AdGroup bid and not worry about themes or keyword-level bids.
Share Your Experience. Please share your own experience with the Google display network’s keyword bidding options, with one keyword AdGroups and with themed AdGroup strategies.
Comments