Most companies I talk to have the same problem when it comes to AI and automation: they don't know where to start.
So they either do nothing, or they automate something visible — like a customer-facing chatbot — when the real time sink is something invisible, like the three hours a week someone spends copying data between systems.
Here's how I think about finding the right place to start.
The Three Questions
Before touching any tool or writing any code, I ask three questions:
1. What's the most predictable repetitive task in your business?
The best candidates for automation are tasks that follow a clear pattern every time. Same inputs, same outputs, same steps. If someone can describe the process in a checklist, it can probably be automated.
2. Who's doing it, and how much of their time does it take?
This is your ROI calculation. A task that takes 30 minutes a week for one person is very different from a task that takes 2 hours a week for five people. Start where the time multiplier is highest.
3. What breaks if you get it wrong?
Some processes have zero tolerance for error (payroll, compliance). Others are forgiving. Start with the forgiving ones. Build confidence. Then work up.
The Three Workflows Worth Fixing First
Based on the businesses I've worked with, these three come up over and over again as the highest-impact starting points:
1. Data Movement Between Systems
This is the most common one. Data lives in your CRM. Also in your spreadsheet. Also in your project management tool. And someone is moving it by hand.
An automation that watches for new data in one system and pushes it to others eliminates this entirely. No more copy-paste. No more "did this get updated?"
2. Report Generation
Weekly reports, monthly summaries, performance dashboards — most of these are built by someone pulling data from multiple places, formatting it, and sending it. Every time.
This is highly automatable. The output is predictable, the data sources are fixed, and the frequency is regular. A system that generates and delivers reports automatically gives your team time back every single week.
3. Intake and Routing
New leads, support tickets, project requests — anything where information comes in and needs to be categorized and sent to the right person. The manual version is slow and inconsistent. An automated version is instant and never misses.
The Mistake to Avoid
The most common mistake is starting with something complex because it sounds impressive.
AI-powered customer service bots. Predictive analytics engines. Automated proposal generators. These are all real things — but they're not where you start. They require clean data, clear processes, and organizational buy-in that most businesses haven't established yet.
Start small. Automate one thing completely. Let your team see it work. Then move to the next one.
The compounding effect of getting three or four small automations working reliably is almost always more valuable than one ambitious project that half-works.
If you want to talk through which of these applies to your business, get in touch. I'll tell you what I'd tackle first.