By: Maher Ghazal: Chief Growth Officer, Reach MENA

From Guesswork to Ground Truth: What 2025 Taught Us and What 2026 Will  Demand 

As 2025 comes to a close, one change stands out clearly in how marketing and media  decisions are being made. We moved away from relying purely on inferred intent and  began modelling intent using real, connected signals. 

For years, marketing strategies were built by classifying people into abstract audience  types: the luxury seeker, the trend setter, the impulse buyer and inferring behaviour  based on who we believed they were. These frameworks helped organise thinking, but  they were ultimately proxies for intent rather than proof of it. 

In 2025, that approach continued to evolve. Across both Reach MENA and PiWheel, we  saw further shifts toward using deterministic signals, actions such as searching,  booking, purchasing, and travelling and, critically, connecting them rather than viewing  them in isolation. 

In ecommerce, more clients began relying on purchase behaviour not just to optimise  conversion, but to inform how and when to engage audiences earlier in the journey. In  travel, there was growing focus on in-market travellers, historic travellers, and travellers  to competing destinations; people already expressing intent through action, not  assumption. What had traditionally been treated as lower-funnel data proved invaluable  in shaping higher-funnel decisions. Upper-funnel marketing became less about broad  awareness and more about relevance, showing up at the right moment, not everywhere. 

This shift required more than better data; it required new ways of working. In 2025,  companies like ours had to invest in data science to help connect these signals and  build more meaningful personas, and in engineering talent to translate insight into  process. Understanding intent became less about defining audiences and more about  designing systems that could interpret behaviour at scale. 

This evolution was reflected on the human side as well. The people who stood out most  weren’t those following established playbooks, or even those simply skilled at  prompting AI. Single prompts didn’t scale. What made the difference were leaders who  began building. In one case, a team lead went as far as learning how to code using AI,  building an internal MVP to solve a real problem, and proving its value before developers  were brought in to turn it into a full tool. Over time, this became a pattern, teams didn’t  wait for new processes, they created them because the work demanded it. 

Looking ahead to 2026, intent modelling will become even more layered. Human  actions will remain central, but they will increasingly be combined with signals  generated by AI systems working alongside people; tools that compare options, narrow  choices, and shape decisions before a final selection is made. The challenge will be bringing these human and machine signals together into a single, coherent  understanding of the individual. 

In 2026, competitive advantage will come from access to connected data including sales data, platform data, and media data, and from the ability to link these sources so  AI models can meaningfully predict, model, and support decision-making. Without this  foundation, AI remains theoretical; with it, it becomes practical. 

The organisations that succeed will be those doing the deliberate, often unglamorous  work of building systems, both technical and human, that allow them to understand  intent clearly and act on it consistently.