Pattern89 looked at the Facebook ad data in their system (they claim they used AI but everyone claims that, it's probably NLP -Ed.) and found some benchmarks and return on ad spend (ROAS) for different content types. While this is not pharma, and benchmarks are difficult at the best of times, it's still entertaining to see what other marketers are doing on the platform.
Some of the top-line numbers from the report:
- 43.62% of marketing budgets are spent on digital
- 15% of brands spend > $1 million
- The average spend was $100k
- Top advertisers have over 40K ads (combination of creative and unique targeting -Ed.)
- The average is 470 ads
- The middle 50% have between 150 and 1650 ads
Then the report provided some ad benchmarks across their entire portfolio:
The Pattern89 system looked at images to categorize them (but the results weren’t that applicable for pharma marketers -Ed.) and emotions to determine which were the most productive.
If you believe that sentiment and emotions can be accurately determined by current algorithms (the author does not -Ed.) then the following emotions provide the best ROAS. Even if you don’t believe that the algorithms are accurate, we can see that the numbers are close enough that marketers should try a spread of messages and experiment with their audience.
This report has other data about image types, emoticons, and beards (4 beards = 3.93:1 ROAS). Grab the PDF to get the full story.