In a constantly evolving marketing landscape, one technology is emerging as the potential pivot: the AI generation. Known for helping businesses go further, faster and more efficiently, is it also the key to a new way of running creative campaigns? Joyce Gordon, head of generative AI at Amperity, recently discussed this topic and more with Rio Longacre, Managing Director of Slalom and Jon Williams, Global Head of Agency Business Development at AWS.
Gen AI: making sense of the mania
During our conversation, Longacre describes a paradigm shift in the Gen AI landscape. Beyond experimentation, companies are now charting strategic paths, identifying areas where they can truly make a transformative difference.
“Over the last few months there has been a big change, which is very positive. Instead of “let’s just try different things,” it now says “let’s have an AI generation strategy.” They seek to identify areas where generative AI could make a big difference and move things forward. They want to invest in these areas, whether it’s e-commerce, operations or creative; they want to come up with ideas that might work and test them. If they work, great. They will be looking to start marketing them. If they don’t, that’s okay too, so they can change direction and try something else.
AI as a marketing assistant
Williams shares where Generation AI shines as a marketing assistant, of sorts. “Amazon Q is a new type of AI-powered generative assistant that can be used specifically for work tailored to your business to have conversations, solve problems, or generate content. It uses data and expertise found in your company’s information repositories, such as codebases and enterprise systems.
“You can use Q, for example, to:
- • Learn a brand style guide, then
- • Use this information to turn a press release into a blog post that meets these standards, then
- • Analyze the appearance of a brand on social networks, then
- • Create new messages around these releases that will make sense to subscribers, then
- • Analyze the results of these publications, and finally
- • Summarize them for review by the teams
“It’s almost like this self-fulfilling circle of incremental productivity that happens as a result of leveraging some of the generative AI capabilities that are being used through a bot but are connected to the systems and data that your business has . We’re still in the very early stages, which is pretty exciting.
3 ways marketers are leveraging AI generation for greater efficiency and savings, according to Longacre:
- 1. Ecommerce Business: This company wrote descriptions for 10,000 product SKUs using Gen AI in a matter of weeks, saving them months and approximately $1 million.
- 2. Paid Media Campaigns: When it comes to paid media tools, such as those designed for Amazon Marketing Cloud, there is a background image generator designed specifically for creating lifestyle images. The results indicate a remarkable 20-25% increase in conversion rates for products featured with lifestyle images compared to those with a plain white background. Rapidly deploying these features, testing their effectiveness, getting results, and then optimizing based on this information is a game changer.
- 3. Banks and finance: The bank’s creative briefs are now generated by artificial intelligence, reducing the time spent in discussions with agencies by around a week.
Even with segmented strategies, brands often face resource challenges. Accelerating the creation of creative briefs, creative images and product descriptions enables faster personalization of on-site experiences. This progression toward personalization does not require them to move into a “hands-off” mode where the AI generation actually runs the show. Instead, it truly is a true one-on-one chatbot interaction or conversational AI.
Keeping humans in the know with AI
Longacre points out that every use case he shared involves a human. Since we are still in the early days of AI, this is not surprising since most brands are starting with “human in the loop” use cases. This is where the AI generates results that a person then approves and possibly refines. “Human in the loop” use cases enable productivity gains while minimizing risks related to hallucinations or unexpected results.
“Maybe the copy is written by Gen AI, but a human reviews it,” Longacre says. “The image may be generated, but it is not released into the wild.
“We’re starting to see a little bit of that, but generally there’s human oversight. Even with chatbots. I mean chatbots have been around forever. Most of them were based on machine learning. You need to know, “OK, when do you have the escalation?” Where do we go from the chatbot to a real person for certain use cases?’ Identifying this is always extremely critical.
Prepare your brand for success with Gen AI
When developing a generative AI strategy, Williams highlights five key elements that play a central role in ensuring success. They are:
- 1. Technology stack: Your tech stack is vital. You must have the ability to explore models, test use cases and choose the right ones.
- 2. A solid and mature first party database: Generative AI relies on data to function properly, which means you need to have strong storage and data ingestion management capabilities to ensure first-party data is accurate and as close as possible in real time to ensure the accuracy of model output.
- 3. Human surveillance: You still need that human intervention to make sure that what you thought should happen happens and there are no anomalies.
- 4. AI-specific analysis skills: Leveraging AI requires the ability to accurately interpret and apply the results of the AI model. You need to ensure your teams have the expertise to understand how the tools work and how best to leverage the data.
- 5. Process redesign: Consider existing processes or workflows that need to be abandoned to take advantage of general AI.
Start small with AI for big results
My advice to brands and organizations when deploying AI: start small. I would start with a small use case that is highly measurable and doesn’t require a major change. One area that clients we work with have had a lot of success with is simply subject line optimization or body optimization of emails or paid media ads. Since you can have a human in the loop here, it’s a great opportunity to experiment with creating different segmentation strategies and messages.
