When clients of a business consulting firm were asked what they wanted, an app to track applications, documents and renewals emerged as one of their most requested features. The firm spent months designing and developing it. After launch, fewer than 20 per cent of clients used it.
Most still preferred to call or send a WhatsApp message for an update from a real person, recalls Neha Thomas in conversation with Campaign Middle East . Neha, who is an independent strategy consultant to government and private institutions, worked with the firm.
The experience exposed a gap between what clients requested in a survey and what they found useful in practice. It is a distinction Neha believes marketers can overlook when building strategies around consumer research.
“If a strategy is built mostly on what people say, it’s built on intentions, and intentions are a weak predictor of behaviour,” she says.
Neha sees surveys as a useful starting point for understanding how people perceive themselves and what they aspire to. Those insights can inform a brand’s positioning and tone, but should sit alongside evidence of what customers actually do.
Purchase histories, CRM and loyalty data, website and app activity, search trends and social listening can help marketers test whether stated preferences translate into action.
The consulting firm’s app offered one example of that disconnect. Clients had requested a digital tool, but most continued to choose direct contact with a person once it became available.
“For marketers, the practical takeaway is to treat what people say as a starting point and test it against what they do before building a strategy around it,” Neha says.
That principle also informs her approach to GenAI-powered digital twins designed to simulate consumer personalities and responses.
Neha sees potential in using these tools to generate ideas and stress-test assumptions early. However, she cautions that models reproducing stated attitudes may also reproduce the limitations of the research feeding them.
“In other words, they risk automating the same say-do gap we already struggle with,” she says.
She also raises questions about representation, particularly when applying models trained heavily on English-language and Western data to audiences across the UAE.
The value lies in supporting exploration, with decisions subsequently tested against customer behaviour. “The final decisions should still be grounded in real behaviour and real conversations with customers.”
The distance between brands and their audiences can begin within strategy teams themselves, Thomas argues.
“We work in similar offices, follow the same industry voices, and test ideas on colleagues who think the way we do,” she says. “It’s easy to mistake our own world for the market.”
In the UAE, differences in income, language, culture and lifestyle make that a particular concern. A campaign that feels relevant to a marketing team in Dubai Marina may mean little to a family in Sharjah or a worker in Al Quoz, she says.
Neha identifies a lack of diversity as a key contributor to this disconnect. Teams representing a narrow slice of their market risk overlooking experiences and priorities outside their own.
Underused first-party data compounds the problem. Brands may hold valuable information across CRM systems, loyalty programmes, websites and customer service teams without consistently bringing it into strategic decisions.
Research can then reinforce ex…