A lot of people hate the Google Display Network (GDN), and for good reason. I think there has been no greater squandered potential within the Google Ads sphere than GDN and the Mobile App Networks. What I mean by this is, the failure to more effectively filter poor quality placements and apps. This would help prevent advertisers from "trying" the GDN and failing due to all the crazy crap out there on the web.

I think this is one of the reasons AppLovin has done so well, by they way. It seems to be like all they've done is take good audience targeting and place those ads on good potential app placements. Google absolutely has that capability, and it's pretty frustrating that they have only more recently begun to take significant steps to clean up the GDN.

Unfortunately, it already has a bad name... but this doesn't mean it can never work! Here are some questions you may have about the GDN, and our ZATO compiled answers as to how we think about this.

‍

‍

1. Should your GDN offer be materially different than other advertising channels?

This depends a lot on what your GDN goal is for that specific campaign. I’m a fan of segmenting the Google Display Network between remarketing (warm audiences) and prospecting (cold audiences). This means you may want to have different offers for those different groups, but IMO even more importantly, you will likely want different landers.

‍

2. If the offer and lander are different, how should they be different?

Over-simplified: think of the prospecting lander as "what is the info a person with no details about our brand needs to know and how do we funnel them to take that action " and then the remarketing lander “what is the info we want to remind people of, or determine why they haven’t purchased and then funnel them to the action”.

So while you may want lots of social proof in both of them, perhaps in the remarketing lander you have a couple of video case studies like “I really didn’t believe this would work at first!” whereas in the prospecting, perhaps you have more of your primary branding video plus some basic social proof.

Also, it could be that you find value in testing different discounts on your landers as well for prospecting. For example, on Search—where users are actively looking for something—you are simply connecting your primary product and pricing with an actively interested user. But on Display, where people aren’t necessarily shopping yet, throwing in a free sample with a minimum spend could be a better way to grab their attention.

The trick is making sure these offers from other channels complement each other rather than clash, and work for your margins—so customers don’t feel like they’re missing out or getting mixed signals. I think discounting on landers is so individual brand dependent it’s really hard for me as an agency to make broad suggestions, but I think having some sort of “wow hurry buy now!” offer on a lander, and thus on the GDN ad as well (for attracting the eyeballs and clicks) really is important for helping GDN work.

‍

3. How do new visitor/new customer %s look VS other channels? ‍Does it have a tendency to maintain exclusions?

NOTE: I haven’t had enough exposure to GA4’s new visitor metrics with Display to confidently compare it to other channels. However, if tracking new customers is a priority, looking into GA4’s audience breakdowns or segmenting first-time visitors could help provide a clearer picture.

Overall, if you have customer exclusions in place, and you are uploading your customer email list regularly, you should be able to fairly accurately exclude your returning customers. THAT BEING SAID, I’d highly suggest pondering this recent video by Taylor Holiday at Common Thread Collective on not over estimating your loyalty from returning visitors:

‍

‍

This may be a case where (1) exclude excluding as best you can (2) RETARGETING those returning customers in a separate campaign with unique (higher) ROAS targets so you have confidence that even if they’re not totally incremental, at least the requirement for profitability is higher so you’re not spending on returning customers… and then (3) just not caring as much about whether it’s new or returning if the math maths out.

By the way, for the sake of definitions, please keep in mind that "retargeting visitors" and "returning customers" are not the same thing to Google.

• Retargeting/Remarketing visitors are people who have been to your website who may or may not have ever purchased from you.
• Returning customers are those people who have actually purchased from you in the past.

This is an important distinction because these two groups should be treated differently in your targeting and offers.

‍

4. How would you recommend starting on the Google Display Network for someone who wants to test it?

Here's the ZATO playbook for starting out on the GDN in almost every account who wants to try Display:

1. Create (at least) 2 Demand Gen campaigns, which are part of the Display Network but offer more control over placement quality.

• a) remarketing: Begin with remarketing strategies to re-engage high-intent users.

• b) prospecting: Use placement and contextual targeting for mid funnel. Contextual targeting can be quite broad, so be mindful of what keywords you would use. Start with your top performing terms on search.
  • Expand with lookalike audiences (available only in Demand Gen) to find new customers similar to your existing ones.
  • If your account has robust conversion data, you can also test optimized targeting, but keep audience selection tight, so the system will have a good seed sample.

2. Create 2-3 GDN Display campaigns,

• a) remarketing: Build a campaign with individual specific audiences targeted in individual ad groups.
  • This allows you to (1) set individual bidding targets for individual ad groups/audiences (2) it also allows you to try different landers and discounts or offers for those different audiences.
  • For returning customers, as an example, instead of excluding them, why not compile a list of customer emails (and upload to Google as a Customer List) who have purchased Product A, and then serve them ads and land them on the page for Product B that goes along with Product A? Rather than avoid returning, upsell them with remarketing!

• b) prospecting:
  • Placement targeting is a good way to start with more control over ad inventory. My suggestion if you have enough volume is to have a dedicated placement campaign where you target key placements of value you have researched in each ad group. This will allow you to control specific ad creative, landers, and bids for individual placements and you can get SUPER granular in targeting a specific website/forum audience. For instance, when I managed a coffee equipment company, we could target specific home barista websites and design specific coupons and ads and landers FOR THAT specific website, and really connect on an individual level with those specific forum users.
  • Audience-based targeting, create a separate prospecting audience campaign, and I recommend using combined/layered audiences—a custom mix of audience segments that better define an ideal customer persona. For instance, you can create a custom audience where you target people who have searched for specific keywords:

‍

‍

5. How do you determine success in the Google Display Network?

Ultimately, focusing too much on in-platform metrics is not realistic, especially with prospecting. So good ole holdouts and incrementality testing are important, but this is pretty tricky without a set plan like Haus or other long-term providers offer. Taylor Holiday (again, I like the content he puts out :) ), has a great podcast episode on being cautious with poorly done incrementality testing:

Engagement Metrics can help give you directional insight on the prospecting side of things – Look at things like time on site, bounce rate, and page depth to gauge traffic quality. If Display is driving a lot of low-engagement visits, it might not be reaching the right audience. This is absolutely tricky, however, and should be taken with a grain of salt since if you are using gated landers, you may have very successful sales from that page but poor engagement metrics. I see them more as “additional pieces of the puzzle” rather than the sources of truth.

Other questions as you gather data: Are you seeing an increase in net new customers? How does that percentage compare to other top-of-funnel (TOF) channels? And just as important—what’s the cost of acquiring a new customer through Display versus other TOF channels?

Since new customers often take time to convert, it’s a good idea to exclude the last 3-7 days (depending on conversion lag) when evaluating performance. Without knowing the specifics of the business, the best way to measure success is by comparing Display to other TOF channels. It won’t be a perfect 1:1 comparison, but it can still give a clear picture of whether Display prospecting is performing around the level of satisfaction you have with other TOF channels. Industry benchmarks are helpful, but always secondary to your own data.

Conversion Rate – How well is Display traffic converting compared to other TOF channels? If it’s significantly lower, it could mean adjustments are needed in targeting, messaging, or offer strategy.

View-Through Conversions– Since Display is more of an assist channel, tracking indirect conversions helps show its impact. Even if someone doesn’t click the ad right away, Display might still be influencing later conversions.

NOTE: When setting targets for Display, especially for New Customer Acquisition, consider these Google-recommended benchmarks—they tend to be fairly accurate from what our team has found:

• Display Target CPA = Avg. CPA (Non-Brand Search) × 5
• Ideal Display Budget = Display CPA × 5

If you have a lower risk tolerance, you can use a multiplier of 3 instead of 5, but keep in mind that this may extend the learning phase.

‍

So there you have it, I think the GDN can be successful for Ecommerce businesses, but you definitely have to have your landers and offers nailed down, and a smart testing of placements and audiences is important.

I hope this was helpful in at least giving you some additional things to think through!

‍

‍