Why You Need To Understand Your Problems Before Building AI Solutions

Forbes

5 hours ago

Kelly Starman is the chief marketing officer for MasterControl.
So, you're ready to create an AI-first marketing program. The time is right, the technology is advancing quickly, and building for the future is a smart decision that will set you up for success in the transformed landscape ahead.
However, as with any marketing initiative, getting the first step right is critical. It's also where most leaders get it wrong. To build an AI-first marketing program, the first step is addressing the problem, not embracing the tech. It's an easy mistake to make.
Marketing teams get enamored with the latest AI demo, sign up for three new tools and, six months later, wonder why their efficiency hasn't improved. But the issue isn't the technology; it's the approach.
Why Starting With Problems Sets You Up For Success
The AI landscape feels overwhelming because it is. With virtually unlimited software options and build-your-own possibilities, it's easy to lose focus in the noise. Starting with a tight focus on parameters helps shape your approach so you can actually get somewhere productive.
More importantly, starting with your familiar business challenges gives everyone an avenue to offer meaningful contributions. When you're trying to drive change across teams, not everyone is at the same comfort level with AI technology. Rather than jumping straight to LLMs, agents or retrieval-augmented generation (RAG), talk about improving sales development output by 30% or reducing campaign deployment time by half.
This approach also helps you avoid getting excited about impressive technology that might not actually solve your problems. There's a huge organizational cost when you implement new tools, including training, integration and budget allocation. If the technology doesn't deliver measurable impact, it can become a disappointment and resource drain.
This lesson was made clear for me during a recent vendor demo for a content generation tool. It looked incredible, and the webinar presentation was compelling, but when we sat through the actual sales process and saw how it would work in our specific workflows, it didn't deliver what we needed. The gap between polished demo and practical application was significant enough that we passed.
Understand Your Needs, Through An AI Lens
It can sound daunting, but defining your problems for an AI-first marketing program is the same methodology marketers have always used, just applied to AI opportunities.
To do this, start with the following:
Start with the necessary leaders at the table in order to eliminate cross-functional inefficiencies and capture every important detail. Each person should represent the challenges that are vital to that area of the business—and have an equal voice in identifying what needs attention. Assign a dedicated project manager whose job is to maintain AI implementation and problem-solving momentum.
Carefully discuss each function and brainstorm down to brass tacks. No problem should be unworthy of consideration or excluded from the whiteboard. Then, narrow your list to repeatable processes with clear beginnings and ends. In our organization, quarterly campaigns work well for test cases because you execute them predictably. Annual events involve similar challenges but aren't repeatable because every year is different.
Target processes that are measurably inefficient. A common example is content workflows where projects bounce between demand generation, product marketing, creative, back to demand gen and then to product marketing again. These are the workflows where everyone acknowledges the inefficient back-and-forth that wastes time, but nobody has had the framework to address systematically.
Simply finding individual use cases that benefit single functions or just one person won't create truly AI-first marketing teams. Focus on processes that span multiple departments and create positive organizational friction.
Building The Right Approach
You can't tackle everything at once, so your team needs to decide what you are capable of addressing at the time. A solid approach might involve kicking off one well-crafted initiative and expanding from there, based on results. Trying multiple approaches and hoping one succeeds runs the risk of diluting your focus.
Meet with your leaders weekly to maintain momentum, track progress and establish a productive cadence. I find daily standups can be a valuable tool, especially when we are actively implementing a new program. Above all, the key to success is consistent forward motion and accountability.
Don't be afraid to do a little internal marketing, as well. When teams see leadership prioritizing these investments, they are much more likely to engage meaningfully with the process. Your involvement signals importance and helps drive adoption.
Keep It Practical, Not Perfect
There are formal methodologies and structured approaches for problem selection, like Six Sigma. These are especially valuable if your team needs a more comprehensive framework or a general overhaul. The benefit of the informal approach outlined above is that it can help marketing leaders get started quickly and see results without extensive training.
Ultimately, this first step is about identifying options and prioritizing which problems offer the best AI opportunities. Once you've selected your top targets, you'll move to detailed process-mapping and success metrics.
The difference between AI success and failure isn't the technology you choose—it's whether you build to address real problems or go right to theoretical solutions. Teams that begin with concrete business challenges create systematic frameworks for transformation. Teams that start with cool technology accumulate tools without impact.
Forbes Communications Council is an invitation-only community for executives in successful public relations, media strategy, creative and advertising agencies. Do I qualify?