The Display Grid
Mike Rhodes has an interesting approach to the Display Network using a Display Grid as seen below.
PPC experts can use this grid to combine targeting options and ad types for Display Network campaigns. The grid is a visual representative of the possibilities with display advertising and helpful for PPC experts to see which strategies are already in use in a campaign or where there are additional opportunities. It also simplifies the process of creating appropriate ads to use in each part of the grid. Not all businesses need to run on every network or with every ad type, but each square is a potential campaign for PPC experts who are ready to scale their Google Display Network campaigns.
Be cautious with keywords in Display Campaigns since they are broad match. Yes, they can help scale a campaign, but it may scale it in the wrong direction! Although it is helpful to view top keywords in Search Network campaigns to potentially use in Display Campaigns, PPC experts are better off thinking about concepts and themes rather than aiming for specific long tail keywords.
Also to consider is whether keyword terms overlap industries because it could result in targeting an audience that is not relevant to the business. Although one keyword per ad group may seem like a lot to manage, it will provide the most control when choosing to target by keywords so PPC experts can see if they are on the right track with a particular phrase.
We walked you through targeting for each stage of the funnel in an earlier post and have additional detail on this below demonstrating how to use data to further influence path to purchase. Take the example of someone planning a vacation and how you might reach him based on the below path. The target persona is “John”, a man in his 40s enjoys travel and reads blogs about different places to take a vacation. After doing some reading, John then watches a video about specific trips of interest so he has started to get more focused in his searching.
As he compares prices, in-market strategies might be used because we now know that John is getting closer to making a decision. Since John is actively researching, Google checks site content visited by John and the recency of visits to differentiate interest from intent in real time. In-market ads serve to people who are close to conversion, with real time updates.
Once John starts visiting hotel websites, PPC experts can re-engage John with dynamic tailored content on the sites he visits. All of this is assisted by Google’s machine learning that uses context around what a person is viewing, what device they are using, and physical location to maximize ROI.
A custom affinity audience is an audience PPC experts create. It is created by entering multiple interests and/or URLs to see what that audience looks like based on Google data for a target market. This feature can also be utilized to reach your competitors’ audiences through the Google Display Network.
To use for competitor audiences, PPC experts create a new Display Campaign for a custom audience but use it only to build an audience, not to start a new campaign by entering the domains for competitors along with competitors’ names. When the audience is previewed, Google will provide an overview of what competitors’ audiences look like which can then be added to a campaign.
Successful audience targeting through any of these methods is key to success and is described further in the below video from the AdWords team.
Before scaling a campaign, PPC experts need to measure performance of advertising efforts thus far. In the early stages, brand impact is key and measured by viewability, reach, brand lift, or Google consumer surveys. The emphasis is more on quantity, such as how many people viewed a YouTube video. As the searcher continues searching and interacting with multiple channels, success metrics include search lift and other indicators for the stage of their journey.
Google Analytics offers data that not only indicates the success of your Display Campaign but can be used for even more sophisticated audience targeting with insights based on acquisition channels, stage of the funnel, engagement, technology used, sequential conditions and demographics. As advertisers learn about the audience and segment them appropriately, there is opportunity for more personalization with messaging. And of course, the real power comes from combining these options to reach the right customers.
Takeaways and Best Practices
- To truly test performance, it is good practice to use the same ad size to be able to truly test, rather than a single ad in different sizes. Different sites on the Display Network allow for different size ads so you want to be consistent when comparing performance.
- Take a quick look at your placement report and sort by cost. If there are expensive placements on what seems to be irrelevant or poor quality sites based on the domain name, such as .ru, pause it. You may end up excluding sites that could be good performers eventually but if it’s questionable, just exclude it because there are so many more sites to choose from.
- If an existing in-market audience is close enough to what you sell, test a small campaign targeting those searchers.
- Affinity audiences are broad so it may not be a targeting method to use by itself. Instead, layer it on top of another targeting method.
- Take time to experiment and utilize multiple combinations of keywords and topics
- Consider using in-market audiences after first maximizing the potential of your own data from Google Analytics or your CRM.
- If you serve a very niche market and want to find new users who look like existing customers, you can do some targeting with Similar Audiences
Creating and managing high converting campaigns on the Google Display Network is a time-consuming process so do not spend too much fixing ads that are not doing well. Instead, pause the strategies that are not producing results, keep learning from data, and expand on the winners. And take advantage of automation such as bid strategies to allow more time for the creativity and strategy pieces that cannot be replaced by a computer.