top of page
  • Writer's pictureJohn Bell

TRENDS: Marketing AI Systems Start Delivering Value

If you are a marketer or a small businessperson, you have probably been targeted with ads pushing artificial intelligence companies who can create your ads, your blog posts or more. They want to help you create the most efficient ads quickly or reduce the cost of producing long-form content like blogs.

The ad testing and creation side of AI is a bit more measurable. The companies who promise to help you write blog posts – like - have a harder time demonstrating improvements as organic search engine results – the most relevant performance metric – take months to demonstrate movement.

How It Works

Companies like Persado, offer to significantly improve content and ad performance by making sure you are using the most effective words in your copy. Their natural language processing (NLP) and “Experimentation Engine” reveals which words, phrases and messages drive the best actions. Keep in mind, their target customers are big, enterprise businesses who advertise at scale. Most often, AI requires significant quantities of data to drive the improvements. AI works best when it is optimizing against an existing, mature advertising practice. That’s where it can show worthwhile improvements like a 66% increase in app downloads for a health insurer.

Pencil caters to fast growth DTC brands. Again, these are ecommerce companies that spend an appreciable and consistent amount in paid media. That gives them a lot to learn from and something to improve upon. This is mostly about testing online ads far faster and more efficiently than human-evaluated A/B testing. Brands need to upload key brand assets like images, video and copy to give services like Pencil the raw materials to create and test ads.

Small businesses may find it tougher to reap the benefit from these new AI companies unless they spend enough or their performance data is aggregated by the AI company with many other brands to come up with the scale of data necessary for machine-learning to do its thing.

Still, now is a good time for those brands with certain assets to start experimenting with creative AI to start learning.

What Brands Need to Learn and Succeed

Whatever the size of your business, there seem to be three attributes of a good AI experiment and program in ad creation and performance.

Brand Assets: All that work you have done to define your winning brand value proposition, key differentiators and product descriptions is important to fuel the AI engine. There is a limit to the “magic” and the truism of ‘garbage-in, garbage-out’ goes for AI, too.

Minimum Ad Spend: AI is a numbers game. It needs enough performance activity to learn and have something to improve upon. When you think about DTC brands and small businesses, the logical question is what is the smallest spend possible to have a strong outcome. No doubt, there are lots of variables by market size and business type, but I am going out on a limb to just get specific and say that about a $10K per month ad spend is somewhere around the minimum.

Good Performance Metrics: This likely should go without saying, but let’s be clear, if you don’t have a strong measurement model and system in place, you will likely not learn as much as possible with an AI investment. That's why DTC brands are such a good category to use and learn from AI. They have a direct conversion metric to optimize against. Even if AI ad vendors promise to deliver all the reporting necessary, you will want to corroborate what happens through your own marketing measurement. Get this foundation right before proceeding.

Lots of Data: Let’s just say that the more good data available, the more machine learning can deliver for you.

Marketing AI will get better and better. The performance marketing use case is one to start experimenting with now. The benefits will be immediate and could actually significantly improve advertising ROI now.


bottom of page