• At a recent ASI-hosted webinar, three distributors discussed how AI is helping automate tasks, boost efficiency and free staff for client-focused work.
• Panelists stressed that AI requires human oversight: Users should refine prompts, verify outputs for accuracy and protect sensitive information.
• Experts recommend meeting employees at their comfort level, identifying tasks they’d like to off-load, learning from peers and continuously experimenting as AI capabilities evolve.
For Bryan Goltzman, AI is not unlike a Halloween pumpkin.
Each year at the beginning of fall, the founder and president of Liquid Screen Design (asi/254663) carves a giant pumpkin.
“Last year it was 1,200 lbs. This year will be 800 lbs,” he says. “Every house on the block has a pumpkin on Halloween.”
Some are beautifully crafted jack-o’-lanterns, while others are basic pumpkins with little effort put into their appearance. Most of the houses have a pumpkin to work with, but some are more intricately decorated than others.
“Like AI, we all have access to the basic tools,” he explains. “It’s about the skill of creatively putting your ideas into it.”
Goltzman shared this comparison at a recent ASI-hosted webinar, where he was joined by distributor peers Tom Goos of Image Source (asi/230121) and Lois Jones of Ironmark (asi/169127) to share their tips and recommendations for successful AI usage in promo.
The conversation comes as businesses are already seeing gains when implementing AI thoughtfully and securely. According to Forbes Advisor, 64% of business owners believe AI will improve customer relationships, and a recent Nvidia report found that 88% of business professionals saw at least a modest increase in revenue because of AI.
But what does AI adoption and implementation look like in practice? Goltzman, Goos and Jones discussed how they’re using the technology, what roadblocks they’ve faced on the path to adoption and how others can start their journey toward AI success.
Use AI To Amplify Human Capabilities
For Tom Goos, AI does the grunt work so that his team can focus on building key relationships and, ultimately, serving clients. The president and CEO of Image Source says he and his colleagues turn to AI as a strategic partner for a laundry list of tasks, including but not limited to prospecting, due diligence, coding and tech, presentations, creative development, product ideation, summarizing data, process improvement and problem-solving.
For example, instead of reviewing an entire 60-page contract himself, Goos asks AI to assist.
“I ask AI, ‘based on what you know about Image Source, what are my risk factors?’” Goos says.
Interestingly, Goos adds, many chatbots provide better results when the user is kind and says “thank you” after the prompt is complete. But regardless of manners, prompting AI takes time and effort up front.
“You have to train it,” Goos explains. “Sometimes it takes five to seven times back and forth to get the results I want.”
For Goltzman of Liquid Screen Design, one of his best use cases for AI is email management. “Going through my inbox is a big time suck,” he says.
Through an email automation solution called Fyxer, Goltzman off-loads specific aspects of email management, including categorization, and trusts it to draft responses. “It’s not perfect, [and] sometimes you have to edit, but usually it’s good enough,” he notes. “And it gets better and better every time I use it.”
“If you’re not paying for AI, you are the product.”
The technology works on Gmail and Microsoft Outlook, and while it’s not a free tool, Goltzman says Fyxer is well worth the price.
“It saved me 8.5 hours of time last week for my personal email,” he said.
In addition to giving him back time on arduous administrative tasks, Goltzman and his team have embraced a software development practice called “vibe coding,” which relies on AI to generate code. The result? An agentic AI tool that functions autonomously and has saved his staff 25% of their time since its implementation.
The Dangers of Overreliance on AI
Of course, AI’s answer is only as good as the human overseeing it. Goos, for example, has seen his LLMs hallucinate several times.
“You should challenge it, because it’s trying to please you,” he says. “Use your brain to make sure what it is sending is factual and true.”
This advice is especially apt for those who use the free versions of these LLMs.
“If you’re not paying for AI, you are the product,” explains Goos, who pays for both ChatGPT and Copilot. “Be careful about putting personally identifiable information or financials into the chat.”
And even if you use AI to draft an email or create a concept, “don’t ever make it obvious that what you’re sending is based on AI,” he advises. “It can sound inauthentic, and authenticity in this world is so important.”
Meet People Where They Are
Not everyone is going to be equally ready or excited to use AI, and that’s alright, says Jones, who serves as the AI adoption lead for Ironmark.
“It’s about meeting people where they are in their AI journey,” she says. “Some are earlier on. I work with all different types of people and give them the support they need.”
She adds that her strategy for working with colleagues depends not only on AI maturity level, but also job function.
“Working with someone in finance versus marketing is a very different process,” Jones notes. “I get myself oriented with who I’m talking to, listening and understanding their workflow, and letting them rant about frustrations and frictions.”
At the beginning, it’s all about building trust, she says. “You don’t have to start with the most sophisticated tool first.”
64%
The share of business owners who believe AI will improve customer relationships
Source: Forbes Advisor
But AI isn’t always the answer to a problem. “Often it is, but sometimes we’re missing education, or people aren’t trained on a tool, or maybe it’s a process change,” Jones says. “[AI] solutions should follow the needs of people, not the other way around.”
For those who might be unsure where to start, or are feeling a bit overwhelmed by the prospect of doing so, Jones recommends starting with a wish list. “Write down the things you wish you could hand off if you could hire someone else,” she says, then figure out which of these tasks can be delegated to AI.
Goltzman adds that many people in the promo industry are especially skilled at putting their differences aside and working together to solve common business challenges. AI is an opportunity, he says, to do just that. “Find a mentor and pick their brain” about how they’re using AI, he recommends.
Goos adds that he’s begun blocking time out for an hour every Friday to do AI research. Educating yourself on the topic is especially important, Goltzman says, because capabilities and best practices are evolving so quickly. “What’s best today [might not be] best in the future,” he says. “Figure out what you like right now and evolve as the models do.”
