The “small trader” asks artificial intelligence about today’s sales before tomorrow’s forecasts: Gulf firms’ trust comes first
When the story of artificial intelligence in the UAE is told this year, the same characters recur: sovereign strategies, institutional transformation programmes, and ministries preparing to deploy generative AI in 50 percent of federal services and operations within two years. But an article written by Ahmed Samah, chief marketing officer at Fortis, and published by Wamda, argues that this narrative does not explain how the work of most Gulf companies is changing.
The figure on which the author builds his argument: small and medium enterprises account for 94 percent of businesses in the UAE, according to the Ministry of Economy, yet they receive little attention in AI discussions. He says that what he hears from traders rarely resembles the narrative of large corporations: their concerns are operational and their resources are limited.
The issue here is trust before technology.A 2025 study that surveyed 315 participants from small and medium hospitality establishments in the UAE found that senior-management support was one of the strongest indicators of intent to adopt AI, alongside competitive pressure, government regulation, and a weight greater than the perceived benefit of the technology. The question is not whether the tool works, but whether anyone inside the organization trusts it enough to adopt it and finds time to use it.
The author relies on field research his company conducted with food-and-beverage traders. They described software decisions in terms of risk rather than features: fast payment, easy menu setup, less staff training, reliable support, and clear pricing. Their conclusion:Point-of-sale systems are chosen to reduce risk, not to maximize features.Any AI transferred from enterprise software may demand what the small establishment lacks: a dedicated tech employee, lengthy onboarding, and patience for features designed for problems that are not its own.
On the consumer side, change is already happening, on the shopper’s terms. The 2026 Tabi shopping survey, which included more than 20,000 participants in the UAE and Saudi Arabia, found that 43 percent use AI to decide what to buy, and that 51.8 percent of those aged 18 to 29 used it at some point in the shopping journey. Yet only 30 percent place complete trust in its recommendations. Thus usage precedes trust.
As for the traders themselves, the author’s research shows they use tools they have chosen themselves, such as ChatGPT, Gemini and DeepSeek, mainly for marketing tasks: writing social-media copy, generating campaign ideas, and suggesting product names. Some also rely on them for pricing, such as setting a new item’s price, bundling, or discounting. Sales analytics and forecasting, however, were not tried by the traders surveyed. The reason is not merely hesitation; an independent assistant cannot see the actual performance of the establishment unless it is linked to sales and operational data.
The most common request from traders is simple: “Tell me what’s happening in my business right now.” They want answers about today’s sales, what has changed since yesterday, the best-selling items, any unusual activity, who sold the most, and which customers have returned. Not a new dashboard. The author summarises that the trader needs to trust the tool for today’s events before entrusting it with next-week forecasts.
What actually changes for someone operating in this market? For the trader, the purchase decision becomes a test of three questions: how long does setup take, does the tool require a dedicated employee, and do I read my own data. For a vendor selling software to small establishments, a stripped-down version of a platform built for a 500-employee organization will not suffice, and the closest competitor is one that provides instant answers without lengthy onboarding. For a marketer in retail, the disparity in usage and trust among customers means an AI recommendation may reach the buyer before they accept it. This is an inference drawn from the article’s data, not a reported fact.
The author adds that the 2025 McKinsey AI State of the Union report found that organisations classified as high-performing, about 6 percent of respondents, were three times more likely to report strong ownership and senior-leadership commitment. He notes that these results pertain to organisations of various sizes, not specifically small Gulf enterprises, but the lesson is the same: adopting tools is one thing, extracting measurable value from them is another.
In his view, a restaurant that anticipates a quiet Tuesday and adjusts its staff schedule, or a store that re-engages customers who have stopped returning, may operate more efficiently than a competitor relying on manual decisions, and this advantage compounds over time in retail and hospitality where profit margins are thin. The next phase of AI adoption in small Gulf establishments may depend more on how well the technology fits them than on how advanced it is. Those who find the right tool will be better positioned to advance.