Feeding the Performance Marketing Algorithm: Why Creative is the New Media Buying

Anyone who was involved with some digital ad work in the past two years would have felt the whirl of a complete shift, to say the least. You know the mad scientist media buyer, the one with a handful of monitors sitting behind him or her, tweaking ads by the milligrams, and creating dozens of hyper-detailed audience lookalikes, may be on their way out.

These major changes in the game include Meta's Advantage+ and Google's Performance Max (PMax). The current landscape is one of practically complete dependence on automated targeting and bidding in ad platforms by the reliance on AI. Now nearly the entire world of ad  platforms is going through a revolution: automated AI bidding and black-box targeting. The models are used for machine learning and they're very intelligent. They can reach your customer, and more often than not, know who they are better than you do, and they can reach them at the least expenses.

But there is a major drawback of these algorithms: They are as good as the fuel that you put in them. With the full-automation of technical ad management, marketers' main interest will need to now be in creative input quality as well as management of distribution. The first thing you don't have to do anymore to win is to play with the dials, in today's world it's feeding those pigs.

If using Meta, or Google, to target, they're going to use your creative assets as the signal to find your audience. The algorithm not only scores on the content of your videos, but on the images too, as well as the hooks, and dials up the users who are most likely to be interested in your videos, based on that.  Just feeding one type of creative in the machine will keep it confined from learning and optimizing.

Say you are a retailer of an ergonomic chair that you make for the office. If you only target ads towards back pain relief, it would identify people who are suffering from back pain. What about the remote worker that plainly just would like a stylish chair when you consider to their new work space? Or the B2B company procurement manager who's seeking a resilient B2B solution?

Without delivering the creative contents for those views, then audiences won't be found by the algorithm.

The New Playbook: Psychological and Creativity

That's because it's distributing the content, and as a marketer, you'll need to provide the system with very unique, psychologically different creative assets. The AI must be provided with raw material to better test, learn and optimize if it's to be successful.

A well-functioning creative pipeline involves three areas that you need to work to "feed" the algorithm with:

1. Diverse Video Hooks

The first 2-3 seconds of a video are the deciding factor for whether or not the user will scroll past or stay on the video. Don't test minor variations of the same video, test entirely different hooks and the variation does not need to be of the same video. Make a statistic statement ("84% of remote workers have back pain!"), a contrary statement ("Don't waste your money on a fancy office chair!"), or use a raw user-generated content (aka UGC) comment. These are all made to manipulate different psychological digit.

2. Varied Copy Angles

Avoid variations of the same benefit with the help of varied adjectives. Translate the whole angle.

• Angle A (logically-driven): Technical requirements, materials and warranty.
• The emotion-driven (Angle B): Individual wants to feel energised at end of a long workday.
• Prompt C (Social proof): Use all-star reviews and from expert endorsements.

3. Distinct Visual Assets

The Ad Placement Cross-Fit algorithm requires a variety of formats to show ads in many placements (Reels, Shorts, Search, Display, Stories). Submit clean studio product photos and photos of them in real life settings, infographic-style diagrams and honest video content.

If you've tried something and it didn't work out, it's not for usually last about half a second. Here is a quick thing that happens most often for a lot of people: If a performance marketing campaign doesn't work out, it will not be the case that you've set up a bad targeting. It's nearly always due to insufficient data points on which to draw a winning combination because those creative missed it.

Want to know how to develop compelling strategies through algorithm that can scale? Look for our services on Bitwise Branding, Call our team Now.