AI in MarTech: what’s real vs hype right now

There is a pattern that repeats itself every time a new category of technology arrives in the enterprise software market. Vendors rush to rebrand. Procurement teams get excited. Consultants write whitepapers. A wave of implementations follows — some brilliant, many quietly shelved six months later.

AI in MarTech is going through exactly that cycle right now. And as practitioners who work inside these stacks every day, we feel it is time to be honest about what we are actually seeing deliver results — and what is still, despite the impressive demos, either immature, overpromised, or simply not ready for production use at scale.

This is not a sceptic’s piece. We use AI tooling ourselves and we see genuine value in specific, well-defined applications. But the noise is deafening at the moment, and our clients deserve a clearer signal.

The most dangerous thing about AI hype in MarTech is not that the technology is bad — it is that the wrong implementations undermine trust in the capabilities that actually work.

The Honest Split: Real vs Hype

Let’s start with the clearest possible view. In our experience across enterprise CMS, campaign platforms and analytics implementations, here is how the current AI MarTech landscape divides:

Delivering real value

• AI-assisted content tagging & metadata
• Predictive send-time optimisation in email
• Automated A/B test analysis & recommendations
• Semantic search within DAMs and CMS
• Anomaly detection in analytics data
• First-draft content generation (with governance)
• Chatbot deflection for common support queries

Still mostly hype

• “Autonomous” campaign management
• AI-generated personalisation at full scale
• Predictive CLV models without clean data
• One-click creative generation for brand use
• AI content strategy (vs AI content drafting)
• Real-time intent scoring in B2B contexts

What to Actually Do Right Now

If you are a digital or marketing leader trying to make sensible decisions in a noisy market, here is our practical guidance:

Start with the productivity wins. AI-assisted content tagging, metadata generation, subject-line testing and send-time optimisation are low-risk, measurable, and available today in tools you likely already own. Focus there before reaching for anything more ambitious.

Fix your data before you buy AI. A three-month data hygiene programme will deliver more value than a six-figure AI personalisation platform on a broken data foundation. This is unglamorous advice, but it is the right one.

Demand governance frameworks from vendors. Any AI feature that touches published content or customer-facing communication needs a human review step. If a vendor is pitching “fully automated” with no human in the loop, treat that as a red flag, not a selling point.

Run genuine pilots, not demos. Ask to run a pilot on your data, in your environment, against your actual use case. Demo environments are optimised for demos. The only way to know if something will work for you is to test it with your constraints.

The AI revolution in MarTech is real — it is just arriving in stages, and the stages that are ready right now are more modest than the hype suggests. Work with what works. Build the foundations for what is coming. And approach the “autonomous” promises with productive scepticism until the data proves otherwise.