Places to Check for Waste in Your Display Advertising Spend
Waste. Advertisers hate it, but it’s an inevitable part of any online display advertising spend. It’s a bit like the old Wanamaker quote “Half the money I spend on advertising is wasted. The trouble is I don’t know which half.”
Luckily, we are in an era of significantly increased measurability for marketers and companies who are looking to use online display ads to grow their business.
Here are key questions to ask when checking for waste in your display advertising spend:
Where are your ads being served geographically?
This is an easy one. If you’re looking to target specific geographies, and your ads aren’t being served there, it’s likely you’ll find places to fine tune the targeting and save more money.
Who are your ads being served to?
When you’re doing a look-a-like model for online ads, periodically check the demographics your ads are being targeted to. Do they reconcile with what your analytics tells you are high-converting segments? Or are there significant discrepancies? If there are discrepancies, it’s a good idea to hunt them down. It could be your advertising is targeting the right people, but your website the wrong people. Or the reverse. Either is extremely important to know!
Is there overlap in the people you’re targeting?
Especially if you’re running with multiple publishers, try to get a read on what your overlap is likely to be. If the overlap is likely to be high, then you’re likely paying multiple publishers to target the same person.
It’s incredibly difficult to manage overlap if you don’t have a data management platform. Even then, it can be tricky with people using multiple devices.
Generally, it’s a good practice to make sure your publishers aren’t competing against each other on your behalf. I recommend using specific tactics with specific publishers to reduce this. Don’t have two publishers each running a look-a-like model on the same conversion – they’ll ultimately be targeting the same people.
Are you managing to a cost per conversion?
Bots exist, and there can be all sorts of fake or accidental clicks on your ads. That’s why it’s generally a bad idea to have media measured only on a result like “new visitors to the site.” Yes, that’s good for awareness, but you’ll want to check at least a second-level conversion to make sure that the awareness is translating into some sort of interest.
It could be “new visitors to the site who viewed more than 2 pages” or “email signups” or “cart additions” or any number of things – just make sure it’s not easily fake-able through bots or accidental clicks.
What’s your viewability like?
Viewability is one of those new key terms that references whether or not your ad had the opportunity to be viewed by a prospect. Most media companies and publishers will have a “viewable” cost per thousand (CPM) that guarantees a certain percent of your impressions will be considered “viewable.” This CPM is likely to be more expensive than the standard one you’re quoted. I’ve seen CPMs for 100% viewable be more than double a 60% viewable CPM.
The key is to keep an eye on your viewability and what that might mean for your conversions. If your viewability is low and your conversions are high, it’s likely your creative is working well.
How have your cost per conversions trended over time?
One area waste can enter a campaign is when you’re running a publisher who’s always worked, but then doesn’t work as well. If your cost per conversions are consistently going up, it might be time to re-think your model or targeting.
There are always cost increases for things like inflation, competition in bidding, and other marketplace conditions. However, if your cost per conversion is increasing significantly week over week, it’s time to do a deep dive into your campaigns. There might be some waste to uncover and fix.
Ultimately, eliminating waste in display advertising is all about maximizing your spend and your efficiencies. Keeping on top of it can keep every dollar invested in advertising working as hard as possible for your business.