Blog  Data & Insights  

A Smart Way to Run Dummy Ads Without Wasting Media Dollars

by Amobee, December 18, 2019

With the New Year quickly approaching, marketers are racing against the clock to make sure ad campaigns pay off as they close out last year’s budgets and prepare for 2020.

The industry rule of thumb for measuring advertising’s effectiveness has been to deliver an ad experiment to 90% of a prospective audience and reserve 10% for a control group, which is typically served irrelevant ad creative like a Public Service Announcement (PSA).

Comparing campaign performance against a targeted user segment and a control group that has not seen the ad provides “uncontaminated” visibility into which ads drove the highest lift against KPIs such as clicks, revenue, offline sales, and footfall visitation. 

But the PSA formula has grown less accurate as advertisers extend reach across cross-device inventory sources and channels. And with one-fourth of marketer ad budgets typically going to waste, experimenting with new channels, creative, and segmentation can quickly become an unjustified expense.

Eliminate cross-screen complexity with ghost bidding.

Whether campaign experiments are created to test specific creative strategies, frequency rules, or segmentation buckets, the Amobee Experiment Engine now offers ghost-bidding capabilities that can run incremental lift experiments without PSA management or ad costs. 

By reducing the complexity and expense of experimentation, ghost bidding makes it easier to understand the results of showing an ad — or not showing an ad — to a prospective consumer.

Through a bidding and control group allocation process, Amobee randomly assigns new and existing profiles to exposed or control users. When a bid request is received, the bidder calculates the effective bid per campaign goals.

Instead of serving PSA dummy ads, Amobee serves an ad from a different Amobee live advertiser campaign to avoid letting test budgets go to waste.

Increase ad experiment accuracy and ease.

Ghost bidding also encourages more routine experimentation within a more accurate, affordable A/B testing framework. And, given that ad experiments are based on a CPM fee, this can mean a cost savings of 80% or more compared to a PSA experiment. 

Ad experiments can be implemented in just a few minutes, and Amobee’s self-service ghost bidding includes reports on ghost impressions and ghost actions through a direct-access console to make it easier to estimate impressions needed for ad experiments.

Bigger cost savings and scale with Amobee.

A more scalable experimentation process combined with ghost bidding can help ensure that your marketing budget is spent on people seeing and reacting to your ad, making every precious dollar count.

Check out how the Amobee Experiment Engine can help you streamline and execute flawless future ad experiments. Contact us at Solutions@amobee.com.

Find us at CES!

Come find the Amobee team at CES in Booth CS-2 in the Mariposa Ballroom, Level 2. Don’t miss our high-impact Future of TV experience, and click below to book your meeting and demo appointment with Amobee from Jan. 7-9. 

Also, catch us on stage at 4 p.m. on Wednesday and learn more about Amobee’s solutions for the converging world — and stick around for Happy Hour from 5-6 p.m. 

SIGN ME UP! 

Read Next

All Blog Posts
Company Updates

Amobee Launches ads.txt Across Platform

On November 13th, Amobee became one of the the first DSPs to implement the Interactive Advertising Bureau’s (IAB) ads.txt initiative. The implementation of ads.txt across the Amobee Marketing Platform…

November 20, 2017

Platform

Amobee Launches Instagram Ads Solution

Today Amobee announced that it is a Facebook Marketing Partner with early Instagram ads API access. The launch of our Instagram ads solution will provide new opportunities for leading brands including…

September 22, 2015

Forward Thinking

Finding Your Audience Starts with Smart Analytics

An analytical framework based on core business questions can be the opening move in extracting more value from your data. One of the most important of those questions might simply be: “Where is my audience?”

March 23, 2017