How Generative AI Is Improving Performance Marketing
In 2023, we saw generative AI transform the creation of content ranging from video to flat images. This was just a warm-up for what is in store beyond content generation. McKinsey reckons that the productivity of marketing alone due to generative AI could increase between 5 percent and 15 percent of total marketing spend, worth about $463 billion annually. One area in particular that will be transformed (and already is) by generative AI is performance marketing, from strategy to execution. I have said it before, and I will say it again: smart decision makers need to embrace gen AI but do so carefully, managing its downsides (such as their imperfect performance) to maximize the upsides.
By now marketers have either cracked the code for using gen AI in content formulation or they have been left behind. Content creators who understand how to use technology wisely stopped fussing about being threatened by gen AI a long time ago and started to use gen AI to create better content. Like A/B testing headlines and copy tone or ideating on topics as good content marketers have been doing for years with techniques such as Google keyword search. All good marketers adapt to technology. Gen AI is no different.
Gen AI, though, is just beginning to flex its muscle in the realm of performance marketing in ways that the industry is just beginning to appreciate. Let’s look at using first-party data to do audience segmentation, for example. Google is finally making good on its threat to do away with third-party cookies on Chrome, the world’s most popular browser. This puts more pressure on businesses to lean into first-party data more effectively to align their performance marketing strategies with audience intent. Generative AI can help them do so boldly.
Generating Better Audience Insight from Always-on Personas
For example, businesses are using digital twins (audiences generated with AI) to A/B test new campaign and content ideas with more realistic test audiences, resulting in more personalized content and messaging for current audiences and the expansion of content to new audiences without wasting effort. Advertising Age recently shared many examples of how businesses are doing this, including a marketer who feeds audience profiling data into ChatGPT to test marketing messages for different product price points with AI-generated customers. By feeding ChatGPT anonymized customer data from zip codes across the United States, the marketer can interrogate different computer-generated personalities.
And that’s just the tip of the iceberg for AI-generated audience research. AI-generated consumer profiles can be used by brands to create digital focus groups that act like real people, or they can be used to hyper-personalize products and services.
In effect, marketers are using generative AI to create always-on personas, or composite representations of audiences that businesses can use to tailor all forms of performance marketing ranging from pay-per-click advertising to email marketing.
Leveling up Digital Advertising
AI has been shaping all forms of digital advertising for years at this point. For example, AI acted as a crucial catalyst in the rise of programmatic advertising and continues to be the driving force behind its evolution. Generative AI is and will permeate every conceivable form of digital advertising, from workflows to data analysis and campaign optimization.
Consider social media advertising, which has become even more complicated with the explosion of apps such as TikTok and the expansion of ad tools on platforms such as LinkedIn. One of the challenges of managing a sustained social media ad campaign is monitoring shifts in behavior and sentiment before they bubble to the surface. Generative AI does not “solve” this challenge, but it sure can help. A beverage company can use generative AI to analyze current trends on social media, say the rise of health-conscious drinking in a target market, which can inform the nimble creation of a campaign for its low-calorie drinks.
And that’s just one example. Generative AI really transforms social media advertising in every possible way. For example, consider a company launching a new fitness app on Facebook and using a tool like Google Bard or ChatGPT to create and adjust dynamic ad variations. Gen AI can personalize an ad campaign by analyzing user data from Facebook (interests, demographics, past fitness-related searches) and create personalized ad variations, say ads featuring workouts aligned with a user’s specific interests (running, weightlifting).
Transforming Good Old Fashioned (and Valuable) Email Marketing
According to a report by Mailmodo, email is ranked the top-performing marketing channel by 78.29 percent of B2B and B2C marketers. When emails seem personal and authentic, they work.
And this is where generative AI can help. Not by writing your emails. But by inspiring you to write better ones via A/B testing. For example, you can use Bard or ChatGPT to test whether subject lines directly referencing past purchases or interests (e.g., “New arrivals you’ll love based on your last purchase of [item]” or “Did you forget these in your cart?”) will achieve significantly higher open rates and click-throughs compared to generic subject lines (hint: they will!). McKinsey reports on the example of Michaels Stores, which has gone from personalizing 20 percent of its email campaigns to personalizing 95 percent. This has lifted the click-through rate for SMS campaigns by 41 percent and email campaigns by 25 percent.
None of this is rocket science. It’s the outcome of experimenting with the tools available to you.
(First-Party) Data, Data, Everywhere
I’ll close this the way I opened: the value of first-party data. It’s not the end-all and be-all. But as first-party data becomes more valuable in 2024, marketers are going to be scratching and clawing to wring more value out of it for their performance marketing. Generative AI will help them break on through.
Click here to learn more about our performance marketing expertise, including client case studies.
