I asked Gemini, my friendly neighborhood AI assistant, to tell me the most important thing in marketing. The answer? “...there's a strong argument to be made for understanding your target audience.” ChatGPT answered the same question with less hedging. ”The most important thing in marketing is understanding the customer.” These bots get it.
Marketing IS all about the customer. But ironically, sometimes the use of AI in performance marketing can actually make it more difficult to understand customers.
AI can actually make it more difficult to understand customers.
Why? As with so many other questions in marketing, the answer is data. Great marketing, including performance marketing, starts with understanding consumers. And understanding consumers starts with data. In ancient times (ok, 2004), ‘data’ mostly meant simple things like age and gender and income. But nowadays data includes a host of demographics, financials, transactions, and online behaviors.
Based on the size of the data storage industry, there is over 2,000 megabytes of stored data for every person in the U.S. That’s the equivalent of 465 million words, or 3.1 million blog posts like this one, or listening to an adtech sales rep speak continuously for six and a half years.
Why is this a problem?
Too much dependence on these platforms makes it difficult for marketing to inform product and play a role in overall business strategy. AI can make advertising to feel less like a conversation with customers, and more like playing a video game. Marketers may have a general sense for their theoretical target customer. But all too often, the target customer is simply whoever Meta or Google think it should be.
AI can make advertising to feel less like a conversation with customers, and more like playing a video game.
Over-reliance on walled garden algorithms also makes it very difficult to maximize performance across my entire plan. It becomes risky to reallocate budget between channels or test new channels. This, of course, is the whole point of making the box black. But consumers change and media platforms rise and fall (AOL, anyone?). And if my deepest customer insights are locked inside someone else’s platform, it becomes far more difficult and risky to respond to changing customer behavior. The primary role of media platform algorithms is optimizing their own financial performance, my brand performance is just a means to that end. Meta’s gross profit margin in Q3 2024 was 81%. It pays to be opaque.
So what is a marketer to do?
Do not avoid media platforms like Google and Meta. They may be opaque and greedy, but they work. But there are some things you can do to improve the balance of power and get more value.
- Know Thy Customer – Assemble the data you already have on customers, and start your media plan with an informed point of view on who your audience is. No one should know your customer better than you.
- BYO Data - Bring your own audiences and data into the media platforms. Focus on things you know that the platforms do not.
- Setup to Learn - Setup campaign hierarchies in a way that generates learnings on segments, offers, creative, and customer journey.
- Extract - Media platforms are generally opaque, but they do make some data available via API. Coupled with the right setup, this feed can be a treasure trove of insights.
- Over The Wall - Set budgets high enough that you can test new channels and platforms.
We built ActivImpact to embrace the walled gardens, but shift the balance of power more towards the buyer. Data is power, and there is a lot of data at your disposal. The key is making it actionable. Our platform does just that - use data to enable an approach to media where understanding your customer is first, and media allocation comes second.
We are here to help because, like the bots said, marketing is all about the customer.