Palantir and Zeta Global announce strategic partnership.Read News
Palantir and Zeta Global announce strategic partnership.Read News
Predictive AI is already baked into many of your marketing tasks. But when it comes to generative AI, most of us are still dipping our toes in the waters.
What, exactly, is the difference?
Jessica Liu, Principal Analyst at Forrester, explains it simply:
Christian Monberg, Chief Technology Officer & Head of Product at Zeta, explains that as we see generative AI in marketing advance:
“…we’ll finally get to some of the promise that we’ve all been talking about for the last five or 10 years, which is the simplicity that marketers want.”
That simplicity will largely come from increased efficiency and better performance.
GenAI reduces a lot of the time-consuming work in your daily marketing tasks — from brainstorming content, to writing product copy, to optimizing ad campaigns, to answering customer support inquiries. By reducing manual effort and streamlining processes, you can achieve significant cost efficiencies and allocate resources more effectively.
Stitch Fix offers the perfect example. The online personal styling service uses generative AI to craft all Facebook and Instagram ad copy. The brand explains:
“Prior to GPT-3, it would take roughly two weeks to plan and strategize each advertising campaign and draft the asset copy. Now, on average, it takes less than a minute for copywriters to review each asset and they receive a 77% pass rate.”
Related content: Why Marketers Need Generative AI for More Human Marketing
By eliminating or reducing more mundane tasks, generative AI allows you more time to think creatively and focus on higher-level strategy.
Consider a typical email campaign.
You may spend a decent amount of time drafting and testing at least two subject lines. Test, wait, see results. Hopefully, at least one is decent and your campaign can move forward (otherwise you’re starting from scratch!).
But generative AI can audit all of your campaign history and the segments you’re sending to — and quickly create several subject lines that will perform well. DONE.
Now the team members that were spending days or weeks optimizing your emails have the bandwidth to think critically about your overarching email strategy:
Generative AI can crunch the data better than ever before, identifying and executing on opportunities humans never could.
Take segmentation—generative segmentation tools analyze all sorts of structured and unstructured data to automatically create new segments based on predicted behavior and outcomes.
Traditionally, building an audience relies on using a bunch of conditional statements, like:
Instead, with generative AI, you can just provide a prompt like, “I want to target East Coast sneaker heads”.
“Gen AI takes this workflow that requires a lot of clicking and being super meticulous about all your fields, and instead allows you to say, ‘Hey, this is what I want’ — and it’ll do it for you.” explains Roman Gun, Vice President of Product at Zeta. “You can even then say, ‘Hey, how else could I segment this audience’, and it can offer suggestions to optimize for outcomes.”
It should come as no surprise that combining efficiency gains, more strategic bandwidth, and better marketing performance results in the kinds of business outcomes CMOs are looking for: lower costs and higher ROI.
For example, Zeta’s marketing team has trained an AI agent to writing the HTML code that turns an image into an email. The input is the image, and the output is the code.
This saves us about 400 hours a month with a single generative AI use case.
Adoption of generative AI for marketing is growing—but most marketers are still playing around and testing it out.
“Predictive AI is certainly very much large and in charge and used day-to-day, but generative AI still in experimental mode,” says Liu. “Many, especially in regulated industries, are quite frankly scared to actually use it for various reasons.”
Liu therefore sees more adoption in low-risk, high-reward scenarios, such as:
Brandon Purcell, Principal Analyst at Forrester, agrees:
“What I find when I talk to marketers is there’s still not an adequate level of trust to allow these systems to interface with customers on the fly. They’re great for spinning up first iterations of marketing content and iterating upon it, but there still needs to be a human in the loop at this point.”
Related content: The Next Frontier for Data & AI
But there are more innovative and impactful ways to use generative AI in marketing today that deliver more value, such as:
Janet Balis, Marketing Practice Leader at EY, explains that while technology is not there yet, siloed generative AI use cases will be integrated. “Today, [AI is] around optimization within a function. Over time, we’ll create business value and growth by looking at how those decisions interact with each other to create more value together.”
In fact, Zeta already allows for this “agent chaining”, or orchestrating how generative AI agents work together.
What can this interaction of generative AI use cases look like?
For years, marketing has been trying to make a paradigm shift away from legacy marketing campaigns to more of an always-on, autonomous customer journey. Instead of marketers blasting you with predetermined campaigns and offers, you as a customer have the power to determine how, when, and where you want to engage.
Liu explains the vision:
Related content: How Generative AI Impacts your Digital Transformation Strategy
Using generative AI for marketing isn’t without its risks — including accuracy, bias, privacy, and copyright infringement, for example. As adoption of the technology increases, responsible AI practices, like transparency, trust, and human oversight, are also gaining importance.
Purcell wrote Forrester’s first report on responsible AI in 2016. He explains that while he always had a steady trickle of requests for the information, demand skyrocketed with the launch of chatGPT in November 2022.
The good news is that businesses are concerned with building generative AI systems in a trustworthy, ethical, and responsible way.
The question is how to do it? Moving forward, it will be important that AI vendors are transparent with:
And regulators will be taking note.
The European Parliament is leading the pack, passing the Artificial Intelligence Act in March of 2024. In addition to setting transparency requirements, the AI Act also bans certain AI applications (e.g. social scoring), provides for law enforcement exemptions, and outlines obligations for high-risk systems (e.g. critical infrastructure, education).
As we saw with the onslaught of customer data privacy legislation, we’ll undoubtedly see more generative AI legislation in the future.
As you begin to imagine how generative AI might improve your marketing, our best advice is to be curious and to be explicit. Generative AI is likely able to help you with any given marketing task, but you’ll have to be specific with the technology to get the outputs you’re looking for.
“Just play around, but be really clear about what you need out of it,” says Gun. “That’s how you’re going to be able to figure out what works, what doesn’t work, and then you can tweak the prompts.”
It also helps to work with marketing partners platforms that are integrating generative AI natively into their tools.
AI agents can accomplish many different marketing tasks, but building them and learning how to use them can be daunting. That’s why Zeta has been rolling out a library of pre-built agents that marketers can easily use within the Zeta Marketing Platform. The goal is to allow marketers to use AI more easily in day-to-day work — and demystify generative AI in general.
As Ramon Jones, EVP and Chief Marketing Officer at Nationwide, explains, “The tools available are phenomenal: how we use them going forward will dictate who’s going to win and lose in this space.”
Learn More: How Zeta is Harnessing Generative AI to Help Marketers Succeed
The Zeta Marketing Platform empowers businesses to offer highly tailored experiences driven by AI.