As a business owner, the pressure to keep up with advances in technology can feel relentless. No sooner have you got to grips with cloud-based IT systems, or unifying operations across on-premise, online, mobile and more, than along comes something the size of artificial intelligence (AI) to rip up the rule book and make you rethink everything.
Like everyone else, restaurateurs are embarking on their own journeys exploring what AI can do for them. Interest is high – according to Square, 85% of decision-makers in the UK restaurant industry are planning to invest in AI and automation tools.
But there is always danger in rushing into the adoption of any technology simply because you feel pressure to keep up with a trend.
Faced with spiralling costs and labour shortages, it’s easy to understand why restaurant owners are drawn to AI and its promises of ‘smart’, ‘intelligence-based’ efficiency gains and productivity drives.
But in a sector which, from the kitchen to the table, still relies so heavily on human labour, how much can you realistically automate away? With legacy IT systems and data processes, how ready are most restaurants to jump straight into AI? Are there not other priorities to address first?
The engagement conundrum
Industry insiders are well aware of this tension. According to one recent survey of restaurant owners and leaders, the number one priority for tech investment is improving customer engagement (61%), followed by reducing costs (42%) and enhancing business analytics (39%). When asked what the biggest barriers to achieving these goals were, the top two answers were measuring the ROI of new tech (50%), and integrating with legacy systems (44%).
Given the squeeze on margins the sector is facing, restaurateurs want about what new tech will be worth to them. There’s little headroom to take risks. But as they are only too aware, quantifying impact, especially in financial terms, isn’t easy, especially for complex goals like improving customer engagement.
This leaves decision-makers between a rock and a hard place. On the one hand, there’s comfort and security in aiming for relatively easy wins like cost-cutting – wins that can be easily measured. But there’s only so far you can cut costs, even with the cleverest technology. Goals like boosting customer retention and loyalty, raising average order values through personalisation, and creating the kind of word-of-mouth buzz about your restaurant that no investment can buy are more unpredictable. But get them right, and the sky’s the limit.
Laying the data foundations
It’s revealing that improving data analytics follows hot on the heels of cost-cutting in restaurant leaders’ list of tech investment priorities. Advanced data analytics – and, moreover, advanced AI-powered analytics – does promise the kind of sophisticated pattern and behavioural analysis that makes engagement more predictable and measurable. In simple terms, it lets you see what works at an incredibly granular level. And let’s not forget that it throws the door open to personalisation.
But AI and enhanced data analytics come with their own demands. Again, restaurant owners recognise the limitations of the legacy POS hardware and siloed back-end systems they’ve run for as long as possible to get the maximum returns possible from their last investments in new tech. Advanced AI models require high amounts of compute power to run efficiently. An AI that takes a minute to ‘think’ before executing every task won’t wash when customer expectations are for instant, fluid performance.
Similarly, AI and AI-powered analytics are only as good as the data they have to work with. Siloed legacy systems where there’s no effective pooling of data resources across a business actively impede these tools from doing their job.
The priority for many restaurant owners in the short term, therefore, is not a full AI overhaul of their business. It’s putting in the changes to infrastructure and data management that will smooth the path to AI in the future. That involves better integration of systems, making use of cloud and API-first architectures so all data paths lead to one place. It involves gradually upgrading on-premise compute capabilities with newer, faster POS hardware and other endpoints. And it involves exploring what AI can do in small, incremental ways, being rigorous in how you measure impact against business goals, and only building from proven successes.