When automation is the best fit
- The process follows a structured, linear set of rules.
- There are minimal variations, or only a few well-defined ones.
- A consistent, repeatable output is required.
How Should Your Business Actually Use AI?
AI is in every product demo, every conference agenda, and every sales pitch right now. So if you’re a business owner wondering how it actually applies to your business, you’re asking the right question. The hype is loud, and most of it isn’t useful.
The short version: AI is powerful, but it’s not where you start. It sits at the top of a stack, and if the layers underneath aren’t solid, AI just becomes another expensive tool that doesn’t deliver.
People, Process, Technology
Innovating the right way requires businesses to look at technology solutions as more than just the purchase of a product. Think of it as a three-legged stool: people, process, and technology. If one leg is missing, the whole stool is certain to topple. Replacing a piece of aging technology without revisiting the people and processes involved is a surefire way to end up with a shiny, new system that shows up as an expense to the bottom line with no business benefit.
Why do technology implementations keep failing? Most of the time, the failure isn’t in the technology — it’s in what’s missing underneath it.
Process Before System
As a technical person, I love systems. I always want to build something that can automate a task, streamline a workflow, or solve a problem. But over the years, I’ve learned a hard lesson: building too soon can lead to failure. Many times, I’ve tried to build systems or automation where processes weren’t fully developed or a use case was not yet clear. The result is a great idea that never gets off the ground.
This is a layer deeper than the common mistake of implementing a solution without defining the needs and requirements. Even when requirements exist, they may be missing a well-defined process. A classic example is project management. Many small businesses search for the perfect project management software to solve their project delivery issues. But do they have a robust project management process in place first? If not, no software will ever seem to work. The failure isn’t in the technology. It’s in the missing foundation.
Some things don’t require a highly structured process, and you can get away with jumping straight to the system. But in most cases where multiple people are involved, process is essential for the technology to succeed.
Not sure if your foundation is ready? Start with an assessment.
How do you get people to actually use new technology?
If systems require a strong process foundation, what underpins the process? People.
Managing those people, motivating them, and ensuring they adhere to the processes is an art in itself. How do you get people to follow procedures? Do you incentivize them? Penalize them? The field of people management, coaching, and HR exists because getting the best out of people while making them happy and productive is a science.
It’s not just about hiring the right people. That’s one challenge. The real challenge is working with those people to ensure they succeed while the business succeeds. No process, system, or automation will work if the people involved are the wrong fit, poorly managed, unmotivated, disengaged, or unsure how to interact with the technologies and the processes.
Technology isn’t just about tools. It’s about the people using them. And the most successful technological implementations start with the human factor.
In practice
A client of ours invested heavily, in both time and money, in a CRM and marketing automation platform that promised the world. It was nothing but trouble. Nothing worked as it should, and extra costs surfaced at every turn. But as they worked through that painful experience, new requirements emerged that hadn't been considered upfront. In the end, we helped them land on a different platform, one they've been happily using for years since. That first solution wasn't just a bad fit, it surfaced a lack of process readiness and a deeper gap in defining requirements and expectations. The fact that the product itself was also a dud just made the lesson more expensive.
When should you automate, and when should you use AI?
Once you’ve established the foundations of people and process, systems often have their own layers. Most apps we use are driven by a database. Collecting and structuring data properly opens up a world of opportunities for analytics, workflow automation, and AI.
In today’s world, everyone is looking to implement AI in some way. But AI is not always superior to automation. In fact, sometimes automation is the better choice.
When AI is the best fit
- There are too many ambiguous variations to manually code.
- The logic for automation would be too complex or verbose to be practical.
- The goal is intelligent adaptation and decision-making, not just consistency.
A good rule of thumb is to start with automation wherever possible. If automation becomes too complex or too rigid to handle real-world needs, that’s a signal that AI may be the better solution. The right tool in the right situation is the key.
The Technology Hierarchy for Success
This is the stack of success with technology — each layer built on the one below it:
1
People
The foundation. The right people, properly engaged and managed, make everything work.
2
Process
Well-defined workflows ensure consistency and efficiency.
3
Technology
Technology solutions support and enhance people and processes.
Data Collection
A structured approach to gathering meaningful information.
Automation
Streamlining repetitive tasks and reducing human error.
AI
Filling in the gaps where automation alone isn’t enough.
AI plays a role throughout the journey, assisting in different stages. But it’s most effective when layered on top of a well-structured foundation, rather than being rushed into as a standalone solution.
Continuous improvement: duct tape and pathways
Technology is a never-ending cycle of continuous improvement. Failure is essential to progress, so failing fast and failing forward is ideal. Even if you follow this framework in the right order, nothing is perfect on the first attempt.
Sometimes you can’t define something properly until you’ve tried a few approaches. Instead of aiming for the perfect finished product upfront, it might be better to accept an imperfect but functional version first, then refine it over time.
Rem Koolhaas famously designed the McCormick Tribune Campus Center at IIT by waiting to see which paths students naturally took across an open field before paving sidewalks. The result? Highly irregular but intuitive pathways that aligned with real-world usage rather than a pre-imposed design.
This same concept applies to systems and processes. Instead of rigidly designing workflows upfront, observe how work naturally happens, then build the system to support that. This may seem at odds with the idea of defining needs upfront, but it isn’t — it highlights the importance of both. You need to establish the core requirements clearly while also recognizing that many finer details will only emerge as you move forward.
Where does AI actually fit in your business?
Want to explore how AI and automation apply to your business? Join one of our workshops — or start with an assessment to see whether your foundation is ready.