The AI implementation gap – how SMEs can avoid falling behind

Artificial intelligence is on the cusp of becoming mainstream.  

Every day, it looks more like the internet did in the 1990s: becoming less of a fringe topic discussed and understood only by academics and the tech-savvy, and more like one that businesses across all sectors and of all sizes can’t afford to ignore. 

The next five years will see more businesses use AI insights and automation widely across their organisation. 

But research shows that SMEs are risking getting left when it comes to adopting and implementing AI.  In a recent study by Opinium, just one per cent of small businesses and two per cent of medium-sized businesses were widely using AI within their organisations. 

With this technology set to take off, how can SMEs avoid falling behind?

According to Opinium’s study, the rate of widespread adoption of AI technology among SMEs is low – under two percent. A little more encouragingly, around a quarter of SMEs are talking about AI and exploring initiatives, while 18 percent of small  businesses and 32 percent of medium-sized businesses  are testing it in a limited way.  Meanwhile three-quarters of big businesses have adopted AI, with one in ten using it widely.

Looking at this data, it’s clear that many SMEs are struggling to move beyond just investigating AI.  Shifting from AI talk and pilots to wide-spread implementation is no easy task. Even for IT-literate, technology-focused businesses, effectively implementing AI so that it makes an impact can be an uphill struggle. For other businesses, especially SMEs, it can feel like a mountain to climb with multiple barriers to overcome. 

One of the first obstacles is understanding what AI is and how it can help businesses. AI uses computers to create systems capable of performing tasks requiring human intelligence, such as recognising images and holding a conversation. Although this can be achieved by manually defining rules, more commonly Machine Learning (a branch of AI) is being used to learn the rules automatically from data. There are a huge number of core business processes, even within small enterprises, that could be improved using AI, as well as enabling changes in business models and practices.

A second barrier is undoubtedly cost. It’s not only the cost of  hardware and software, but the cost of specialists to analyse and manage data. However, implementing AI technology need not be a multi-million pound project or necessitate an extensive, and expensive,  IT team.  Cloud computing has reduced costs considerably. Businesses no longer need to set aside a room full of expensive kit that only a handful of people can use. Simply by bringing automation and intelligence to relatively simple processes, quick gains can be made. 

A third obstacle is acquiring the necessary skills needed to develop and implement AI solutions. This can present big challenges, especially to SMEs, as hiring AI experts may not be possible. Here, businesses could consider upskilling existing staff using online learning resources (many of them free) and low- or no-code AI development products and services. 

SMEs looking to bridge the talent gap could also work with a consultant firm to buy in expertise, perhaps initially working on a small project that will bring obvious value to the business. Collaboration is key, and costs may well be lower than many businesses imagine.

A final obstacle for SMEs, and arguably the most important step towards successful AI implementation, is developing good data governance procedures. Data quality is key here. Businesses looking to analyse trends over time need to bring together the disparate data sources that exist in a company to create a single version of the truth and ensure data sets are complete, coherent and robust. 

Additionally, clear governance policies to produce reliable, unbiased and valid outcomes are vital. This will include guidelines for the responsible use of data and AI that will become vital, as issues around privacy, transparency, bias and security abound. 

The AI constraints on smaller businesses are often practical. SMEs often have smaller budgets and fewer people with the right skills. But with the right approach these challenges can be overcome and, coupled with a focus on data quality and a desire to become a data-driven business, it is possible for SMEs to keep up in the AI race.