AI automatically optimizes ads and will be the future of ad placement.

[Phoebe｜Oramob]

Recently, I saw a lot of advertisers saying that meta's Advantage+ Shopping Campaign (ASC) is performing very well, so those who have placed ads should pay attention to it.

ASC is based on machine learning to simplify and optimize the ad placement process, and can measure up to 150 combinations of ad materials at a time, pushing the best-performing materials to the most valuable potential consumers.

But currently ASC is not available to all advertising accounts, and should only be open to some of the more mature numbers that buy and convert pixels. After that meta will be open to all.

Through ASC, brand independent stations can just feed advertising material and to meta. meta big data automatically helps you find potential customers who are likely to convert.

I swiped a lot of ad-related Facebook and Twitter today, and my feeling is that ASC is the most valuable change of meta after Apple's ios 14 privacy policy.

If you have an ASC on your ad account, I recommend trying it out. Put in some of your best ad material from past ROAS and see the conversion effect.

I put my 3-4 year old ad material with thousands of likes into ASC, and the material came back to life immediately, with a CPM of less than $20 in the US market in the past two days, with better conversions than other types of campaigns.

The cool thing about doing meta advertising is that as long as you have a strong product + material, you can always recycle and reuse it, and when the results are good, you can open it up to a single.

Good advertising material is worth the continued investment, and in the future, it will be even more critical.

 

[Graybeard]

ML (Machine Learning) is not AI.
Ad Networks' AI claims are laughable.

ASC? WTF is that and it is?
ASC advertising defined - Google Search
Did you just make that up to appear intelligent?


Utilizing acquired data over time in programmatic advertising is really old-hat.

Machine Learning Algorithm

Artificial intelligence (AI) has been with us for a couple of decades. You will come across many terms used as artificial intelligence, machine learning, deep learning, cognitive computing and so on.

You may wonder what the main difference between AI and Machine Learning, or between ML and Deep Learning is.
The term artificial intelligence covers many topics that are not artificial intelligence but many that are the basis of Artificial Intelligent Systems.

Machine learning (ML) is a subset of artificial intelligence (AI). It is a trains the AI deployment to learn from the data provided, without being explicitly programmed to conclude a "correct" result, as a Boolean answer.

ex: IF (sunny_day) AND (water wings) THEN. . . 0 or 1 False or True

What ad networks are passing off as AI (the 'great' leap forward) is just a better Boolean logic leading to an elusive conclusion --better application and ad strategy and tactics --and nothing more. It's a structure that you have to either make known assumptions to from historical data or spend money on acquiring that ad network's traffic to apply their Boolean query logic to. It's useful but also a money pit.

ML is based on the idea that systems can learn from data, identify patterns and make predictions. The goal of ML is to find the hidden structure in data to make better decisions. ML is not AI because it cannot reason or think like a human. ML is a tool that can be used to build AI systems, but it is not AI itself.

Deep learning is a special type of machine learning. Deep learning is a subset of machine learning in artificial intelligence (AI) that is the “wow, that’s amazing” part, involving neural networks and real smart features that may in the future become a part of our daily lives.

 

[Nick909]

I've also recently read about fraud detection machine learning. This is a collection of AI algorithms trained with historical data to suggest risk rules. Then the rules can be implemented to block or allow specific user actions, such as suspicious logins, identity theft, or fraudulent transactions. I should say it sounds impressive.

 

[Graybeard]

I am assuming that this program would apply KPI metrics to advertising segmentation.

In advertising segmentation, AI can be a powerful tool to help affiliates, or other advertisers, better understand their customer base and make more informed decisions about marketing strategies. AI can help these advertisers make better decisions about targeting of their advertising campaigns.

At the end of the day, the use of KPI metrics in advertising segmentation can be a valuable tool for affiliates, or other advertisers, by tracking key performance indicators, an affiliate, or advertiser, will benefit from the Machine Learning insights into which advertising campaigns to continue and which campaigns to abandon as they will never be profitable.

But the insights are only as good as the quality of data. If the ad sources are bad the data will always be bad. If the offers are not seen as a worthwhile proposition by the person that you call traffic they will never succeed often enough to make the advertising worthwhile.

Garbage in --Garbage out is the mantra