AI At Work With A Person In Charge
AI is here to stay. I use it every week, in marketing and in the day-to-day processes that keep a business running. But I'm not here to automate everything or hand off the parts that matter.
AI is a tool. A good one. It makes the work faster and catches things I'd miss. It doesn't replace strategy, judgment, or knowing your customers and your team.
Why This Is Different
Most AI talk lands in one of two places. Hype: automate everything, cut the team, let the tool run it. Or fear: avoid it and hope it goes away. Neither helps a business.
I come at it from the work, not the tool. I've run the SEO, the product data, the ERP rollouts, and the sales collateral by hand. I know what good looks like, so I can tell when AI gets it right and when it doesn't.
AI on top of a broken process just makes a faster broken process. The fundamentals come first. Then AI makes them faster, with a person checking what goes out.
How I Actually Use It
I don't know PHP. When a print-on-demand store needed a bulk product upload with customizations, AI helped me untagle the supplier's developers documentation, private server connection, and helped be build the build the API. Without it, that project would have needed a true developer.
When 40 blog post thumbnails needed updating, I didn't do it 40 times. [One or two lines on how AI helped. Link: How I updated blog post thumbnails without doing it 40 times]
Day to day, I mostly use it to brainstorm, make sense of messy information, and move faster on small contracts and projects.
That's where AI earns its spot: the repetitive work, the technical gaps, and the tasks that would otherwise eat the time strategy needs.
Where AI Fits In Marketing
Here's what building and managing it looks like.
Digital Marketing
Faster research, first drafts of ad copy and email, and pulling patterns out of reports. The channel strategy and the call on where budget goes stay with a person.
eCommerce SEO
Bulk product titles and descriptions, schema, and catalog cleanup across thousands of SKUs. Every output still gets checked against how customers actually search.
Local And B2B SEO
Keyword clustering, location page outlines, and translating technical part names into plain-language search terms. Industry knowledge decides what's right.
Customer Journey
Sorting through heatmaps and session recordings faster to spot where people get stuck. Deciding what to fix first is still a judgment call.
Customer Experience
Drafting sales collateral, onboarding emails, and training documentation. The input comes from sitting in sales meetings, not from a prompt.
AI Search Is Changing How You Get Found
Search behavior is changing. Customers are using AI tools like ChatGPT, Google's AI Overviews, and Perplexity to research products before they buy. These tools pull from the same signals traditional SEO is built on: well-structured content, clear answers, credible sources, and proper schema markup.
What that means for your site:
- Structured data: schema markup for products, reviews, and pricing so AI tools can read and cite your pages
- FAQ and comparison content: the format AI tools pull from most often
- Clear, direct product and category descriptions: no filler, no keyword stuffing, accurate information
- Authority signals: links, citations, and content depth that make your site worth referencing
Where It Fits In Operations
SOPs And Documentation
Turning how the work actually gets done into written procedures: training guides, checklists, and reference docs a team will use. AI speeds up the drafting. The steps come from the people doing the job.
Tap to see the checkpointThe Checkpoint
Every draft gets reviewed by the people who do the work before it becomes the procedure.
Process Automation
Repetitive tasks like bulk updates, data cleanup, and report pulls can run on their own. Every automated process gets checkpoints, so nothing moves forward without a review.
Tap to see the checkpointThe Checkpoint
Automations stop at set review points, and a person approves anything that goes out.
Training And Change Management
New system, new process, new hire. AI helps build the training material and answer the repeat questions, so the rollout doesn't stall on one person's calendar.
Tap to see the checkpointThe Checkpoint
Training gets tested with the team first, and a real person stays the go-to after launch.
Data And Systems
Product data, ERP and PIM syncing, and the spreadsheets nobody wants to touch. AI helps find errors and clean data faster. The person who knows the system still signs off.
Tap to see the checkpointThe Checkpoint
AI flags the errors. The person who knows the system decides what changes.
Where I Draw The Line
Some things should never run without a person checking them. Final decisions. Anything touching email sends to customers, banking, or medical information. Even when a process is automated, it needs governance: checkpoints, a stopping point, and someone accountable for what goes out.
AI drafts it. A person approves it.
Contact Cathy Gray
I work best where marketing, eCommerce, and operations overlap.
Whether you're hiring or need help with your business, I'd like to hear about it.