Tuesday, March 26, 2024

Can AI Improve Ad Effectiveness by 50% or So?

One of the verities of advertising has been that "half my investment is wasted, but I don't know which half." Artficial intelligence should help, in that regard, allowing use of greater analytics to improve precision. Cookiers were supposed to help. But other alternatives are coming.


And it does matter, as so much content access is subsidized by advertising.


Market research firm PQ Media estimated 54 percent of total media consumption time was spent with ad-supported media in 2022. 


Of course, much user engagement with content happens in social media or within the context of search. Google in 2023 earned about $238 billion from search, according to Statista. Social media booked about $199 billion globally in 2023, according to The Business Research Company. 


Social Media

Revenue $ Billions

Source

Meta (Facebook, Instagram, etc.)

132.1

Insider Intelligence

Alphabet (Google)

85.7 (Social Media segment)

Alphabet Investor Relations 

TikTok

6.1

Insider Intelligence 

Twitter

5.8

Insider Intelligence 

Snap

5.1

Insider Intelligence 

Search

Revenue $ Billions

Source

Google

237.9

Statista 

Microsoft (Bing)

12.0

Insider Intelligence 

Baidu (China)

31.0

Insider Intelligence 


So to the extent that advertising enables citizen use of such information and content sources, advertising actually does support media and content access. 


Cookies have been an essential building block for online advertising, but are going to  be replaced. Replaced “by what” remains an issue, though. 


There isn't one single expected replacement for "cookies" as a tracking mechanism for advertisers. Instead, the industry is moving towards a multifaceted approach. 


First-party data is likely to become more important, where data is directly gotten from users with consent.


More efficient is contextual targeting, where advertisers target ads based on the context of the web page or app they are using. This can involve factors like the content itself, user demographics, and browsing history within the specific platform. Artificial intelligence is likely to help with that. 


Efforts also will be made to target based on cohort data rather than individual information. 

FLOC (Federated Learning of Cohorts) is a Google-developed proposal where users are grouped into cohorts based on similar browsing behavior without revealing individual data points.


Unified ID 2.0 is an industry-backed initiative that aims to provide a privacy-preserving alternative to third-party cookies by using a single, anonymous identifier across different platforms. 




No comments:

AI's "iPhone Moment" Will Come. We Just Don't Know When

Some observers might be underwhelmed with the current state of smartphone AI use cases, as they might see somewhat-limited value for other ...