Building Workflows WITH AI is Different Than Handing Your Workflows TO AI
You shouldn’t hand your processes over to AI. You also shouldn’t keep AI out. Instead, build an automated workflow and see where AI can support the person responsible for getting the job done.
Lucy and Ethel try to keep up at the assembly line. I promise there’s an AI-supported workflow example in this article, but I can’t help but use this image to kick things off. Photo courtesy of Desilu Productions/CBS.
Workflow automation is not a new concept introduced by AI. It’s just finding a way to make a repeated process more efficient, accurate, or simple. When you call customer service and they make you select the number to route your call? That’s workflow automation. The chocolate factory assembly line that Lucy and Ethel couldn’t keep up with? Hilarious. And workflow automation. This isn’t an AI thing. It’s just a new tool you can use to build that assembly line.
There’s always an inherent tension between the work that needs to be done and the resources available to make it happen. Yes, if you had limitless headcount you could achieve anything. But that’s not reality. So you have to figure out ways to get it done with the people and resources you have. That’s why efficiency through automation is so valuable. It saves resources and increases output. Margin expands, returns increase, etc. You know the drill.
When I look at AI, I don’t jump to the conclusion that I can cut headcount and fully replace people with autonomous AI agents. The technology isn’t nearly there yet, and there is a substantial amount of value from the experience and insights of those that do the job. But, many times those people are spending far too much time on the less valuable, more tactical elements of the job. It doesn’t matter how talented your employee is. If they are spending 70% of their time on mundane, low value tasks because, well, someone’s gotta do it…that’s a waste of talent and resources. But, you don’t have the budget to hire someone else on the team. So you’re stuck having your star employee deal with tasks instead of finding ways to add incremental value to your organization.
An automated workflow (via Make.com) that is enhanced by AI but doesn’t outsource the whole thing to AI agents.
That’s where Automation plus AI can be powerful, for your team and business. The workflow in the image above illustrates what workflow automation can look like and the significant amount of time spent by your employee to get that process from point A to Z.
The example is a classic one, where a business receives messages in an inbox and someone on the team needs to respond. In this case, the message comes in, someone has to figure out if it is a Lead, a Question, or a Contact, and then the response needs to be crafted based on the specifics of the message. In the image, you can see the classic workflow. Every red circle represents where a person would give input. The first circle is where someone reads the message and determines what category it belongs in. After it’s routed, it needs to be read and summarized. That’s a person’s judgement as well. And then it is placed in the folder or file where the person ultimately responsible for responding to the message reads the details and crafts a response to send. The process still needs to happen, and generally it needs to follow that flow. Sometimes it’s one person doing everything. Sometimes it’s a team responsible for their specific area of expertise. But it’s a lot of people involved, and the quality and speed of the process relies on their abilities to read and respond to it. Depending on the size of the team and inbox, that can be a longer process than you’d like.
The new value of AI is that you can insert it into those red circles to do the work of the person responsible, while still maintaining oversight of the process. How? By not letting AI actually send the response email to the customer. The process is built to hand the tailored, drafted email to the person responsible for the response. So the first AI agent or step reads the message and decides what bucket it sits in. The next one takes the message and builds the proper summary. The third then reads that and drafts an email. That email draft is sent to the Drafts folder of the human that is ultimately responsible to review and send it along.
The value is that they get to save their time and energy for the most complex and impactful part of the process. The part that the customer actually sees. That’s got to be done right, and in this model the choice was made not to just outsource that to AI.
Now, there will be mistakes. Early on, the person at the end of the chain may be frustrated with incorrect categorizations and low quality drafts. The good news? They can go back into the AI’s instructions and figure out what’s going wrong. Perhaps the prompt wasn’t as clear as they originally thought. Maybe it’s using an older model of AI that isn’t strong enough at writing. Perhaps there needs to be another step in the process to do some research before the final draft is made. The only thing it costs to add AI steps is the time spent articulating what exactly needs to be done. And as you get more practice and time with it, the better the workflow becomes. The employees can rely more and more on the quality of the outputs, and spend their time drawing from the insights to improve products or customer experience. They aren’t bogged down by the tactical pieces early in the process, but they also still minimize the risk of a poor customer experience due to AI error.
Oh, and that process I built above? The entire thing is completed in 10 seconds. 10 seconds from a customer click to a draft in your inbox. Imagine what could be done with that extra time.
This is why I’m not racing to build an army of AI agents that operate on their own. The place to start is your current processes. If your employees are spending far too much time on low value add activities, think about how to define and teach what they do. You can insert AI into the process, include the proper controls, improve it over time, and let the employees spend their time on what actually drives your business forward, instead of forcing them to just do the tasks that just let’s them keep up.
