Two weeks into 2025 I wrote that I was wondering how the year was going to play out.
Specifically, would AI Agents fully change marketing as we know it.
As I sit here, taking a holiday “break”, from a South Florida rental.
I’m reflecting on that post and what this next year might look like.
There was progress on the AI Agent front but I’m not sure I’d say they’ve fully changed marketing.
Not for me so far.
Not in my consulting work or even at the agency where I help lead a paid search team.
Has AI been involved?
Absolutely.
But not in the form of true AI Agents.
In side coding projects, there’s been some, if we want to count Claude Code as an AI Agent.
In the beginning Claude Code was so hit and miss.
But good enough to subscribe to the max plan.
It just seemed to gradually go downhill though and I decided to cancel the max plan right before they announced the new Opus model.
I still use Claude Code multiple times a week – just on the regular pro plan.
Now, Opus for plan mode and Sonnet 4.5 for execution has been wild.
It’s incredible when you really stop to think about how good these models are getting at coding abilities.
Where Claude Code is now, going into 2026, we’re getting a glimpse into how agents will make their way into more and more things we do each day.
But I’m not really using Claude Code for marketing.
So I still can’t say Agents have changed marketing for me personally yet.
Just my experience.
That doesn’t mean its not progressing though.
This Forbes article highlights brands like Adobe, Saleforce, Braze, Hubspot and others are all investing in and building out agents.
Google is putting agentic capabilities in platforms like Chrome, Google ads and GA4.
Meta just purchased Manus AI for a supposed two billion to help them build out agentic capabilities.
There is an “unofficial” Google ads MCP server that I still need to tinker with.
I can envision using that with a local model agent possibly.
So things are moving along.
Not always in the best ways though.
Let’s see how it plays out in 2026.