The Pulse Check is a series sharing opinions, lessons, and guidance from real marketing leaders. 

Business leaders make countless daily decisions, many of which can change the business trajectory for better or worse. It’s a lot of pressure so, increasingly, leaders are leaning on data to inform the decisions they make. 

But generating good insights from large data sets in a time-critical situation can be difficult or even impossible, especially when you’re talking about decisions that involve a level of personalization — like how or when to contact a prospect or customer. 

Many marketing and sales leaders have turned to AI to help their decision-making process, with mixed results. Others have steered away from entrusting early-stage technology with their decision-making, choosing instead to learn from the mistakes of their peers or wait for the technology to mature.

Whether you’ve adopted, dappled, or steered clear to date, there’s little doubt that decision-making with AI will become a part of your process in the near future. Thankfully, the marketing and sales leaders using it now already have a lot to share about how to make the most of AI-powered decision engines  — and with new and improved AI-powered tools set to land in 2025, there’s never been a better time to hone your skills.

Lesson 1: Define roles for AI tools and human teams

Overwhelmingly, marketers agree that AI is best when humans play to its biggest strength in the decision-making process: number crunching. 

It can be helpful to consider the role or job title your AI tools are playing in any given project. For example, think of your AI tool as a consulting data scientist or strategist and you will set an expectation that a clear, thorough brief will be required for success. You will also have more room for forgiveness and intervention when the AI doesn’t understand the nuances of your audience, brand, or business model.  

As Tasha Tadi, Marketing Manager at eBallot shared, “AI excels at crunching data and finding patterns, but it doesn't know our audience the way we do. We always step in to interpret the results and make adjustments. For me, it's a tool that sharpens decision-making, not replaces it. There's still plenty of human judgment and creativity involved, and that's where the magic happens.”

Peter Lewis, Chief Marketing Officer, Strategic Pete agrees. “The strong points of AI are data crunching and pattern recognition, not understanding human nuances of a brand's voice or long-term goals. It's great for segmenting customers, creating personalized experiences at scale, and running through decision trees faster than any human could. But sometimes it starts generating strategies that look good on paper, but when you step back and look at the bigger picture, it's lost the story.” 

Once you get that balance right, you can have insights that lead to huge marketing wins. For example, Garin Hobbs, a martech expert from InboxArmy saw a huge 25% jump in email open rates after letting AI refine audience segmentation. 

Of the experience, he said, “It saved us time and delivered better outcomes than manual efforts ever could. But AI doesn't always get it right. It's great with numbers but can miss the mark on nuance. That's where the human side kicks in. We use AI to generate insights, but it's up to the team to interpret them and adjust strategies to fit the context. It's a partnership, not a replacement.”

Lesson 2: Learn to write specific and clear prompts

Prompt writing is the skill to hone if you want to get the most out of AI, especially when it comes to decision-making. 

As Adnan Jiwani, Assistant Manager Digital Marketing at PureVPN, shared, “One key thing we've learned about writing good prompts for AI tools is to be clear and specific. Instead of saying, 'Show customer insights,' we'd ask, 'What are the top three products customers aged 25-34 purchased last month?' This saves time and gives us better results. It's all about asking the right questions to get meaningful answers.” 

The prompt perfection process may help you crystalize your goals as a business, too — it forces you to get specific about what you expect to learn and the data that should, or should not, inform the decision. 

Aimie Ye, Director of Inbound Marketing, Centime gave the example of finding the top-performing content across their site. She said “Instead of asking, 'What content performs best?' we use prompts like, 'Analyze blog performance by conversions for TOFU versus MOFU content.' This precision not only enhances the AI's output but also ensures the recommendations align with our business objectives, enabling data-driven, strategic decisions.” 

Tasha at eBallot shared a similar sentiment, “One thing I've learned is that AI thrives on specifics. If you're vague, the results will be too. Instead of asking it to 'analyze voter behavior,' I drill down into the details—what time frame, what audience, what outcomes I want. The more precise you are, the better the output.

It's not perfect, and we're still finding our way to use it most effectively, but it's shown me how much potential exists when you blend AI with a human touch. It's like having a really smart assistant that still needs your guidance to shine.” 

Lesson 3: Pay attention to your data

Garbage in, garbage out. We’ve all heard it before and it rings true for even the most sophisticated tools. AI can only work with the information it has, so you need to trust the data before you can trust the results your AI tools generate. 

Janelle Warner, Co-director at Born Social takes a three-fold approach. “First, to get good results, we have to feed the AI quality, structured data. Cleaning up data pays dividends later. Second, we monitor outputs and have staff override bad recommendations - that feedback further refines the AI. Finally, we keep the human in the loop. AI augments our team's skills rather than replacing them.” 

Clean data is important, but equally so is understanding the data you have and how it is structured. As we covered in the section above, writing clear, detailed, and specific prompts is crucial to getting results with AI. The closer you are to your data — including naming conventions, how you track activities and data retention limits — the easier it will be to write specific prompts that your AI tools can instantly understand. 

Again, it can be helpful to think of your AI tools as a team member. When you start working with an AI tool, you come with a set of skills and experiences, and, over time, knowledge is passed over. Soon enough, you develop a sort of shorthand with each other — and that’s when harmonious, efficient work begins. 

Working with you, not for you

Even as AI tools and assistants get more sophisticated, decisions will be made with the help of AI, not by AI. Marketers who work to input specific and clear prompts for decision-makers to work with, add a human lens to the AI-generated insight, and ensure data inputs are clean and complete will set themselves up for success.  

As Peter at Strategic Pete put it, “AI is a fantastic tool—if you know how to work with it. It cannot replace the subtle thought, the creativity, or the strategic oversight that comes from people.”