Despite growing investment in artificial intelligence (AI) across the hospitality sector, many restaurant operators have yet to see measurable returns. Deven Desai, head of AI products at iTradeNetwork, explores how connected data and unified workflows can help AI protect margins and deliver greater operational value.

Optimism around AI remains high across the hospitality sector. Many organisations agree that AI has the potential to improve workflows and around 90% of surveyed CEOs believe that by 2028, AI will redefine what success looks like in their industry.

Yet, despite the overall AI sentiment, many enterprises in the hospitality sector have failed to achieve measurable returns. Nowhere is this more prominent than in restaurant operations.

As of now, around 73% of operators are actively investing in AI or planning to, but only 9% report that AI has produced measurable operational value or guest impact. Restaurant operators need a way to reduce this disconnect between AI ROI and confidence in its implementation. This starts by identifying where current deployments are falling short and how operators can tailor AI applications to protect margins.

Operational blind spots are behind AI’s underperformance

The traditional approach throughout restaurant management is to focus AI investments on customer-facing tools, forecasting, or reporting. These areas are strong candidates for AI integration, but only if the data supporting them is aligned. This is often not the case.

The core operational layer across restaurant systems is usually fragmented, and 37% of restaurant brands point to this fragmentation as a key barrier to delivering better guest experiences. Pricing often differs between systems, and procurement platforms can fail to apply contract pricing at the point of order. The result is AI systems that function with minimal business context, leaving operators scrambling to resolve discrepancies after the fact.

Most of these small inconsistencies lead to margin leakage. Small pricing errors, off-contract purchases and substitutions compound quietly, and by the time a restaurant manager checks the profit-and-loss statement (P&L), the damage is already done.

Recapturing some of this lost margin and laying the foundation for successful AI adoption requires operators to unify their workflows and link their data. This would create the operational backbone AI needs to identify discrepancies earlier and prevent margin loss before it reaches the P&L.

The infrastructure behind successful AI

Many organisations that successfully implement AI follow a similar approach. This framework includes:

• Step one: Creating clear product data records. Restaurant operators should synchronise product and supplier data, update pricing throughout systems, and reflect contracts at the point of order.

• Step two: Breaking down operational silos. Organisations should ensure that all data is integrated across internal and partner platforms.

• Step three: Centralising all information into a single area. After enterprises eliminate operational silos and update data across systems, they can consolidate this information into a single, unified view to create a clear picture of performance.

While all organisations will have different requirements, this framework serves as a valuable starting point for a connected business architecture. It provides operators with a single source of truth and the visibility needed to make clear operational decisions, while giving AI the operational context required to optimise workflows.

This means pricing, supplier and inventory information are available in one place, and manual reconciliation efforts are reduced. In short, enterprises could incorporate this technology in a way that actually adds business value.

Practical applications that protect margins

With a secure and synchronised infrastructure, operators can deploy AI to automate critical back-office tasks, including contract compliance, price enforcement, purchase order management, and inventory visibility.

Over time, these applications will allow AI to act as a continuous support layer. It could shift from a tool used for standalone functions to an always-on part of restaurant management. This would allow operators to reduce repetitive work, proactively surface areas where workflows are likely to fail, and alert management before issues escalate.

The value of these capabilities is even clearer when tied back to margin protection. By proactively identifying areas where exceptions occur, AI could address the profitability drain earlier in the process and reduce the revenue leakage that often shows up on the P&L.

Unlike manual workflows, AI can complete all these functions in near real time, allowing operators to monitor and respond to live market signals before margins are affected. Taken together, these capabilities would help restaurants transition to a predictive business model.

From reactive operations to proactive management

If AI is going to transition from stunted pilots to a system that can work alongside team members to deliver value, then restaurant teams must address fragmented workflows. Operators cannot rely on AI while critical operational systems remain disconnected.

Once organisations unify this architecture and the data within it, restaurant teams can use AI to shift their operations from reactive problem-solving to proactive management. These systems can operate in real time to identify discrepancies and automate price enforcement across workflows.

Ultimately, this will translate into less margin leakage and stronger returns from AI investments across restaurant operations, positively impacting the overall hospitality sector.