Every December, the marketing industry pumps out prediction pieces for the year ahead. This isn't one of those. This is a look back at what actually moved the needle in 2025—and where a lot of smart people got caught chasing the wrong things.

I've been doing this for 25 years. Some lessons from this year felt new. Most were painful reminders of stuff we already knew.

The year started with everyone convinced AI would automate marketing into the ground. What actually happened? The companies that used AI to produce more mediocre content got buried. Google got better at spotting thin AI-generated material, and audiences got better at ignoring it.

What worked was using AI as an accelerator, not a replacement. Teams that used it to speed up research, refine drafts, and handle repetitive tasks came out ahead. Teams that used it to pump out 50 blog posts a month without human oversight? Their organic traffic cratered.

The winners figured out something important: AI doesn't give you a strategy. It gives you leverage on the strategy you already have. If your strategy was "produce more stuff," congratulations—you now produce more garbage faster.

Google's AI Overviews went from experimental to everywhere in 2025. The data is ugly. Organic CTR dropped by as much as 45% on queries where AI Overviews appeared. Even keywords without AI Overviews saw a 41% year-over-year decline in click-through rates.

If you were still measuring success purely by rankings, you missed the plot. A #1 ranking means less when Google answers the question before anyone clicks.

What worked was getting cited inside the AI Overview itself—being the source Google pulls from—not just ranking alongside it. That meant doubling down on E-E-A-T: real expertise, first-hand experience, structured content that's easy for machines to parse. The "write 2,000 words on a topic you Googled ten minutes ago" approach is dead.

Also: brand search held up. People who knew what they wanted still typed the name directly. Building a recognizable brand mattered more than ever because it's the only thing algorithms can't easily intercept.

I launched Scrollworks this year because the data was screaming at me. Short-form video now accounts for over 80% of internet traffic. TikTok still has the highest engagement rate of any social platform—nearly three times higher than Instagram.

What worked was authenticity over production value. Micro-influencers with 10K followers outperformed celebrities because audiences trust people who feel real. UGC-style content—stuff that looks like a customer made it—converted better than polished brand spots.

The lesson for local businesses was especially clear. You don't need a video team. You need real people talking about real experiences in 30 seconds or less. That's it. The brands still waiting to "figure out video" lost ground they won't easily recover.

Every year someone predicts email is dead. Every year email outperforms. In 2025, AI-optimized subject lines and send times pushed click-through rates even higher. Email remained one of the few channels where you actually own the relationship.

What worked was treating email like a conversation, not a megaphone. Segmentation. Personalization that went beyond "Hi [FIRST NAME]." Emails that solved a problem instead of just announcing a sale.

If you spent 2025 obsessing over TikTok while ignoring your email list, you left money on the table.

Chasing algorithm updates instead of building authority. Every time Google rolled out a core update, marketers scrambled to "fix" their content. The sites that stayed stable were the ones that had already built topical depth and genuine expertise. Quick pivots don't fix a weak foundation.

Thinking paid media could replace everything else. Paid media works. It also got more expensive, and the platforms are increasingly unpredictable. Brands that went all-in on Meta or Google without building organic equity woke up to a nasty surprise when costs spiked or reach tanked.

Waiting for clarity before acting. TikTok was supposedly getting banned. AI was supposedly taking over. A lot of companies sat on their hands waiting to see what would happen. The companies that experimented anyway are now six to twelve months ahead.

Nothing from 2025 was really new. Build trust. Create value. Show up consistently. Use new tools to do it faster, but don't let the tools become the strategy.

The marketing fundamentals don't change. The landscape does. If you spent this year fighting the landscape instead of doubling down on the fundamentals, 2026 is a good time to course-correct.