There’s a lot of noise about AI in marketing right now. Half the industry is acting like it’s going to replace everyone, and the other half is pretending it doesn’t exist. The truth is somewhere in the middle, and we figured it was worth sharing how we’re actually using it day-to-day at Humes Design.
Content drafts and brainstorming
This is the big one. When we need to write a blog post, social media captions, or email copy, AI is a solid starting point. We use it to generate rough drafts and then rewrite them in our client’s voice. It cuts the “staring at a blank page” time in half.
The key word there is starting point. We never publish AI output directly. It’s a drafting tool, not a writer. If you’ve ever read something that sounds vaguely informative but weirdly generic, that’s probably raw AI output that nobody bothered to edit.
Research and competitor analysis
Need to understand a new industry before a client kickoff? AI can summarize competitor websites, pull out positioning themes, and help us get up to speed faster. It’s like having a research assistant who works at 3am and doesn’t complain about it.
We still verify everything — AI gets facts wrong more often than you’d think — but it’s a huge time saver for initial research.
SEO keyword clustering
AI is surprisingly good at taking a big list of keywords and organizing them into logical groups. What used to take a couple hours of spreadsheet work now takes about ten minutes, plus another twenty minutes of us checking the groupings and adjusting.
Image generation (sometimes)
We use AI image tools for mood boards and concept exploration. It’s great for showing a client “something like this” before we invest in a full photoshoot or design. We don’t use AI-generated images in final deliverables though — our clients are paying for original work, and that’s what they get.
What we don’t use it for
We don’t use AI for final copy, brand voice development, design work, or anything that requires understanding the nuance of a client’s business. Those things require human judgment and years of experience that a language model doesn’t have.
We also don’t use it to inflate our output. Some agencies are using AI to pump out ten blog posts a week for clients. That’s not quality content — it’s noise, and Google is getting better at recognizing it.
The bottom line
AI is a tool, like Photoshop or Google Analytics. It makes certain parts of our job faster and easier, but it doesn’t replace the thinking, strategy, and creative judgment that make the work actually good. The agencies that treat AI as a replacement for talent are going to produce mediocre work. The ones that use it to augment good people are going to pull ahead.
We’re in the second camp, and so far it’s working out pretty well.