If you don't have the ability to track view-through information (otherwise known as post-impression information) on display ads, it's difficult to optimize display campaigns.
Historically, display optimization has been done through A/B testing. You run the display campaign, you look at how that affects your KPIs (profit, revenue, ROI, CPA, etc.), and if you see lift in the KPIs you're focused on, you can "attribute" that to the campaign working. If your numbers stay flat or go down, you can conclude display isn't adding value.
With more A/B testing, you can get more granular. You could run one specific creative, look at your bottom line results, then test another creative and compare the results.
You could do the same kind of testing to evaluate whether one network, size of creative, specific website, set of demographics, psychographics, etc., is better than another, and then compare the results.
But is there a better or quicker way to get these answers? This type of testing is costly and takes a long time to finalize.
New Tracking Technology
The other variable that comes into play is that you aren't running these tests concurrently, so the results you receive may be indicative of the timing of the test versus the specific element you were trying to test.
More companies are leveraging technology to optimize their display campaigns in a more exact and efficient manner. Rather than using a whole host of A/B tests spread over a period of time, companies are leveraging new tracking technologies that allow them to track view-throughs versus just clicks. This tracking technology can provide specifics of performance for each network, psychographic attributes, behavioral attributes, demographic and geographic attributes, size of creative, specific website, etc.
Some of the largest ad agencies and advertisers are even going a few steps further by combining view-through-level tracking with attribution management technologies. A combination of view-through and attribution allows a marketer to assign a value (profit, revenue, etc.) to every ad a consumer views or clicks en route to conversion. For these marketers, optimizing display just got a whole lot easier.
With view-through tracking and attribution tracking in place, a marketer now has the ability to view all their ads in terms of profit and loss. When they want to evaluate whether a display campaign is working, it's as simple as if the ad has profit attributed to it, it's working; and if it has a loss attributed to it, it's not. Of course, they wouldn't stop there.
They also have the ability to determine things like which network is most profitable, which creative within that network is most profitable, what is the average frequency of impressions for profitable paths versus not, which media partners are adhering to campaign guidelines such as frequency caps, etc. This allows a marketer to be able to use all this information to understand how much they'll be able to spend on display in terms of CPM and/or CPA.
Display Optimization Rebirth
Ultimately, the optimization of online media is about running ads that work while paying a price that allows you to produce the maximum profit.
With A/B testing, optimization of display likely came down to "turn-it-on-or-off" decision making. This type of testing provides no insight into the proper CPM/CPA you should pay for this media.
With view-through tracking and attribution, optimization now gets far more granular and exact. It also takes far less time to know what works.
Display advertising, with the advent of view-through tracking and attribution technology, is going through a major rebirth. No longer are companies asking display to perform the same as paid search (click to close). They understand that the online marketing world is similar to the offline world.
Consumers still need to become aware, they need to engage, and they need to close. Display advertising is one of the best awareness-building vehicles in the marketing world. Without the ability to track view-throughs and apply attribution to measure performance, the real value of display advertising is missed.