2025 in Hindsight: What We Learned

By Luke Brewin, Operations Director at Force24

At the start of 2025, it felt like the whole industry was standing on the edge of a new era. AI was going to change everything. Personalisation would finally live up to the hype. The cookie would crumble. And somewhere in the mix, someone promised we’d all have robot assistants by now.

Twelve months later, the assistants I know are still very real people, and the only robot I own is a hoover that keeps trying to eat cables.

That’s not to say nothing changed, plenty did. Just not always in the ways we expected. The progress was quieter, more operational, and a lot more human than any keynote slide suggested. Working with hundreds of brands this year, we saw that success rarely came from chasing the next big thing. It came from teams willing to rewire how they work, test properly, and put trust before technology.

So, here’s what stood out to me looking back on what we got right, what we overestimated, and what it says about the year ahead.

 

We Thought: AI Would Change Everything

Reality: It changed a lot, but not everyone changed with it.

Tools became smarter. Workflows smoother. But most teams were still figuring out what to do with them. We saw a rush of enthusiasm early in the year with prompt libraries popping up, clever use cases being shared, and a few brave experiments. But as things settled, the gap between experimentation and integration widened.

Accuracy improved, but discernment didn’t always keep pace. Bias didn’t disappear; it just became better written.

The teams that saw real value were the ones that built structure around it with governance and by reviewing processes and human sign-off. They made AI part of their workflow, not their strategy.

At Force24, that became our biggest learning too. Making AI useful isn’t about novelty, it’s about reinforcement. When we built Nexus, the aim was never to replace the marketer; it was to reduce the noise while making decisions cleaner, not just faster.

AI matured in 2025. The industry, for the most part, caught its breath and tried to keep up.

 

We Thought: Email Had Peaked

Reality: It quietly had one of its best years yet.

If you work in marketing long enough, you get used to hearing that email is dead. But 2025 proved otherwise. When Google and Yahoo rolled out their new sender authentication rules, it forced a reset. SPF, DKIM and DMARC went from being IT acronyms to boardroom talking points.

For the first time in a while, deliverability wasn’t treated as hygiene, and it became strategy. Marketers stopped chasing inbox placement and started earning it. Authentication, permission, tone and timing became the new creative brief.

We saw it first-hand. The brands that focused on reputation and trust, not just volume, outperformed almost every other channel. The DMA’s benchmarking showed delivery rates holding strong at 98%, with open rates sitting around 36%. In an industry obsessed with shiny tools, that’s a solid pulse.

It reminded us that the most dependable channels still work, especially, if you treat them with the respect they deserve.

 

 

We Thought: The Channel Mix Would Get More Experimental

Reality: It got more practical.

After years of chasing trends, 2025 was the year marketers rediscovered what works.
Short-form video dominated attention spans, but when boards asked for evidence, the dependable trio of email, SMS, and post delivered the results.

SMS made a real comeback. And we saw this particularly in automotive, retail, and recruitment, basically, anywhere timing mattered more than aesthetics. Direct mail found its rhythm again too. According to JICMAIL, interaction rates hit their highest level in five years. In a screen-saturated world, something physical landing on a doormat felt refreshing, not old-fashioned.

And email continued to quietly outperform both in measurable ROI. It was precise, predictable and trusted. When budgets tightened, marketers turned to the channels that offered clarity: value per pound and value per person.

Digital can be noisy, and we saw that the quietest channels can do the loudest work.

 

We Thought: Automation Would Replace Effort

Reality: It only replaced repetition.

Automation had a solid year. It did what it was meant to which was deliver consistency, reduced manual errors, and kept journeys flowing when teams were stretched. But consistency is not the same as connection.

We saw that in campaigns across every industry. The workflows ran flawlessly, but the human moments like the empathy, timing, and context are still decided by how well a message landed. Automation can take care of how and when, but it can’t always tell you why.

The teams that did best understood that distinction. They built automation to scale care, not to skip it. They didn’t just automate responses; they designed better conversations.

That’s where the real power lies and it’s not in doing more, but in doing the right things at scale.

 

We Thought: AI Could Build the Future for Us

Reality: It still needs us to make sense of it.

The idea that AI could code, design or even campaign for you was tempting. “Vibe coding;” the idea that you could describe an outcome and have the system build it became the new buzzword.

In reality, it rarely worked that neatly. AI could generate a thousand lines of code, but 900 of them still needed human cleanup. Developers didn’t disappear; their jobs just changed. The work moved from writing to refining.

We saw the same pattern in marketing: the illusion of instant creation gave way to an appreciation of iteration. AI can build fast, but only people can make it fit.

So, the best teams stopped chasing shortcuts and started improving feedback loops by showing ideas earlier, testing faster, and designing before they developed. It wasn’t glamorous, but it was progress.

 

We Thought: Speed Would Always Win

Reality: Sincerity caught up.

In the rush to automate everything, a few brands learned the hard way that faster doesn’t always mean better. Automated replies went out before complaints were resolved. Personalisation crossed into intrusion. Efficiency became the enemy of sincerity.

It was a reminder that automation works best when empathy stays in the loop. The aim isn’t to move faster than your customers but it’s to meet them where they are, in a way that feels considered.

When we talk about “making automation more human,” this is what we mean. It’s not soft talk; it’s strategy.

 

We Thought: Authenticity Would Be Obvious

Reality: It became an operational challenge.

Synthetic content multiplied faster than anyone could track. Feeds filled with AI-generated text and imagery that looked convincing until people stopped trusting what they saw.

Thankfully, 2025 also brought a few solutions. Adobe, Google and others began embedding provenance tools like Content Credentials and SynthID into their ecosystems, helping trace content back to its source. But these tools only work if marketers use them.

For us, authenticity isn’t just a creative value; it’s a compliance one. The brands that will win in 2026 will be those that can prove what’s real like where their content came from, when it was created, and how it was checked. Transparency is no longer a nice-to-have; it’s part of quality control.

 

What 2025 Really Taught Us

It wasn’t a year of revolution. It was a year of realignment. We learned that progress doesn’t come from louder ideas, but from cleaner ones plus cleaner data, logic, and expectations.

AI is only as good as what you feed it. Automation is only as strong as the empathy you build in. And the marketers who’ll thrive in 2026 will be the ones who keep both in balance while pairing trust with technology and creativity with control.

The future rarely arrives in a headline. It shows up quietly, in the small operational decisions that make your next campaign smoother and your customer experience stronger.

That’s what 2025 reminded me. Progress doesn’t always look dramatic, sometimes it just looks dependable.