From the rise of AI explainability to the return of ‘old school qual,’ here’s what’s been shaping the insight space this month.

🔁 Quick Pulse: What We’re Hearing

💬 “Senior stakeholders are buying into AI… but still want to see the workings.”

💬 “There’s fatigue around dashboards, people want decisions, not data dumps.”

💬 “Clients are coming back to qual, they miss the ‘why’ behind the numbers.”

After two years of tech hype, there’s a pivot back toward insight with narrative… not just metrics.

🧠 One Big Trend: AI Explainability Hits the Insight Agenda

2025 saw mass adoption of AI for concept testing, coding qual, and predictive modelling. But now in 2026, trust and transparency are the key concerns, especially when insights are used to shape customer experience or comms.

🧪 Example: Samsung worked with a research agency to pilot an “AI explainability layer” – a human-readable narrative that outlines how and why the model reached its recommendation. This helped align insight with compliance, brand tone, and stakeholder buy-in.

“Black box” AI won’t cut it anymore… clients want to see behind the curtain.

🔍 Method Spotlight: Vox Pop Reels for Instant Storytelling

Short-form video isn’t just for TikTok anymore. Researchers are embracing 60-second vox pops as fast, rich, human-centred data capture.

🎤 Brands are using:

• Selfie-style responses via mobile
• On-street intercepts with social-style editing
• Quick-turn showreels for stakeholder buy-in

👟 Example: New Balance used vox pop reels to test cultural relevance of a brand refresh across Gen Z consumers in the UK and US. The results? Stakeholders described the output as “more convincing than a slide deck”.

It’s qual, but creator-style… human, raw, and powerful.

🚀 Brand to Watch: Pret A Manger

Pret has quietly become a testing powerhouse, using CRM and loyalty data to trial new product lines, formats, and messaging at speed.

In Feb, it launched a micro-test of new vegan breakfast options via app-exclusive offers with instant feedback loops via 2-click polls and purchase behaviour tracking. 4 variations, 1 week, and a decision made.

It’s a lesson in using what you already have to test smarter, not bigger.

📊 Smart Stat

📈 63% of insight teams say they’re “over-reliant” on past data, but only 27% have a proactive plan for live testing in 2026. (Source: Insight Leaders Barometer, February 2026)

The insight edge this year? Being braver in the present, not just smarter about the past.