How are you using AI and automation to grow your agency? That was the question that a group of 20+ senior leaders and I discussed recently at Mirren Live.
My personal biggest takeaway? Most advertisers use AI and automation in some capacity, but use cases vary widely and are often limited. Many leaders struggle with how to make them work in their organization in a structured way that will drive measurable impacts. Others are challenged with implementation and adoption—both from a technical standpoint and a change management perspective.
Here are some of the biggest AI and automation implementation challenges that the group discussed. I’ll also offer some suggestions on how you can start laying a strong foundation for these technologies starting today.
Breaking down data silos: establishing a data source of truth
Before even dreaming of AI or automation success, you need clean, centralized data. This was a recurring theme echoed by multiple leaders at Mirren Live. A single source of truth—that holy grail of unified data sets—is the bedrock on which all automation and AI rests.
“We spent the past year across our media data doing what we most needed to do: build a single source of data truth,” someone explained during our conversation. “It sounds so basic, but you can’t do anything unless you have data. Anybody who tells you otherwise is lying to you.”
Some agencies in our session have already taken the plunge by creating standardized data frameworks that enable seamless integration with automation platforms. Others are still in the data consolidation phase at their organizations or aren’t sure how to move forward.
“What's happening [in my company] right now is our systems are at a point where they're becoming more disconnected,” said one person at our Mirren Live roundtable. “AI and automation and connecting data sources may give us the business intelligence that we need, but there’s not a silver bullet. It’s just spinning wider and deeper as opposed to establishing a center of gravity.”
Breaking down silos, standardizing inputs, and getting all your data into one place are massive, but necessary, hurdles to getting the most value out of AI and automation. Why? Your different systems need to talk to each other if you’re ever going to use technology to simplify and streamline workflows from start to finish.
If you’re not sure how to improve your company’s data readiness, start by thoughtfully looking at your end-to-end operational processes. Where is there an overlap in tech features? Where are there gaps in different data sets talking to each other? For example, if you’re manually moving things from one place to another 75 times a day, there’s a huge opportunity to save keystrokes, time, and sanity with technology.
But the first step, for sure, is connecting your data together in one system.
Reducing tech fragmentation: replacing single-point tools for holistic systems
After years of adapting to different digital advertising evolutions, many agencies are plagued by disconnected tech stacks and technical debt. That’s because advertisers have often been forced to rely on single-point solutions. (A single-point solution is a software or service that addresses specific needs within your business rather than an entire workflow.)
“After using multiple project management systems, we finally learned how important it is that the project management system and accounting and finance systems all freely and openly communicate with each other,” said one participant. “They need to transfer data from one another. We’re on the tail end of converting our systems, but we finally feel we’re moving in the right direction.”
By reducing the need for disparate tools and integrating (or shrinking) your existing technology stack, you can allocate resources more effectively and deliver greater value to customers.
Fixing messy data sets: improving data hygiene for accurate AI output
If you can’t tell, data was a huge topic of conversation during the session. Why? As our participants mentioned, data lies at the core of all their operational bottlenecks.
Reporting, staffing, expenses, figuring out which team members have capacity and knowledge to work on specific projects…running the numbers or getting the data together to perform these operational and administrative tasks wastes the most time and causes the biggest headaches.
“You have so much data,” one participant emphasized. “Having all that data in one place is vital so you can use AI to ask questions, get information quickly, and run your business. Otherwise, you’re running a report in one system and comparing it to something in another system, all just to analyze something. That’s where I see AI coming in in a big way.”
As you go through your technology inventory and look for data gaps, determine what needs to happen for your data systems to pass clean data from one step to the next.
For example, if you’re manually entering budgets into five different places as part of a launch workflow, there’s a high likelihood that one of those entries will get flubbed. Automation can disperse a single data point of entry to multiple systems, in the right format, with minimal manual effort from your team.
Bridging the human adoption gap: fostering a culture of human-technology collaboration
Technology (or data, for that matter) isn’t the only piece of the puzzle. Resistance within teams often presents one of the largest implementation obstacles. As one participant eloquently described, there are two camps on every AdOps team:
“People who want to use AI and people who want to run far away from it.”
“I'm surprised to find myself being more active with AI than some of our younger employees,” one person added. “I thought our younger employees would just jump on it, but a lot of them say, ‘I don't even know where to start with this.’”
Tensions between team members can also arise when some people use AI to “shorten” or expedite their workload while others still use manual processes. If people are excited about AI’s potential, find ways to encourage their exploration rather than condone them—as long as they’re not breaking clear AI company policies or security practices.
As one participant said during the session, “You can start sniffing out when somebody's using AI to do parts of their job. However, we all have to get used to that and not judge somebody for not doing their work the same way.”
One way to start bridging the gap between AI enthusiasts and AI skeptics is to leverage AI within a system that contains your unique data sets. This provides peace of mind that your AI is learning based on your strategies and that the data stays securely within a closed system (an important aspect of AI governance).
You can also reserve part of your team meetings to discuss ways that people use or want to use AI in their work. This could include sharing prompts and successful use cases to inspire others about what’s possible. Even the smallest wins from early adopters can spark interests around AI and automation’s capabilities.
Make sure you’re taking the time to address fears and misconceptions about AI, too. For example, AI agents aren’t here to replace your human teams. AI can complement what your human teams do best and take the most tedious, time-consuming tasks off their plates so they can focus on what they love to do.
Practical next steps for advertisers exploring AI and automation
AI and automation are both useful technologies for advertisers but they are not the same. They play well together, though, and are both worth considering adding into your operational workflows.
AI acts like a creative engine for your company. You can ask it questions, get real-time responses based on real-time events or data, and use it for content generation. Automation helps you replicate AI’s actions—and so, so many AdOPs tasks—at infinite scale in a fraction of the time. It can also keep AI’s outputs in check, acting like a bouncer that keeps noncompliant text or images from going out the door.
When we talk about the relationship between AI and automation in advertising, we emphasize that automation enables you to get maximum value out of AI by making it scalable, safe, and repeatable. (You can get the full breakdown of AI and automation in advertising in our exclusive guide, including where they shine and how to maximize your ROI.)
If this article has sparked ideas for using AI and automation, here are three steps you can take to continue preparing for (or evolving) their use within your advertising company.
- Look for immediate implementation opportunities
Identify one or two “time vampires” in your existing AdOps processes, such as budget pacing or making campaign updates. Focus on implementing automation and AI where it can make an immediate, noticeable impact. Once those quick wins are in place, you’ll have proof of concept to justify bolder investments.
- Identify existing process flows for process patterns
If it’s in a flow chart, it can be automated. No matter how unique it seems, every process mapped out in your playbooks has patterns somewhere—and where there’s a pattern, there’s potential for automation. Start by drawing up your current process flow charts. Look for tasks that show up again and again: those are your goldmines for efficiency. Automating these processes will save time and free your team to focus on the work that fuels creativity and strategy.
- Build an AI and automation roadmap
Now is the time to assess your agency’s readiness. Where do your tech tools overlap? Where are there gaps in data seamlessly moving from one step to the next? Is your team equipped to manage the change? What’s your company’s 6-, 12-, and 24-month vision for automating workflows? Answering these questions honestly will set the stage for sustained success.
Laying the groundwork for tech-enabled ad teams
AI and automation aren’t here to replace what makes your advertising strategies unique. They can help you amplify what you and your teams do best.
Think about it like this: If 80% of the work your most creative team members are doing is not creative work, then you’re wasting their talents. Offloading mundane tasks to robots frees your best talent to focus on what they truly love and excel at.
Of course, this shift to automation and AI requires intention, effort, and patience. But make no mistake, those who fully harness AI and automation will see the benefits—not just in their bottom lines but in employee satisfaction and strengthened client trust. (You can see the proof in every one of our customer stories.)
