106383Hand-holding and the power of free: can a low-cost tailored behavioural intervention carry SMEs over the adoption hurdle?ClosedCollaborative R&DInnovate UKThree quarters of SMEs provide trade credit. They provide the goods or services upfront and get paid later. However, many SMEs have customers that pay them late or not at all. Such 'late payment' is a major problem which increases costs, reduces cash flow, and consumes funds that could otherwise be used to drive business growth. In the extreme, it can force a viable business to close. For the SMEs who are often buying these goods or services, accounts payable is particularly time-consuming, requiring significant time spent in manual data entry. Without rigorous systems in place, overdue invoices get missed and late payment can escalate. Business accounting software provides an accurate and efficient solution for businesses to manage their finances, but it requires much keying of data. Information will come in many formats, both paper and electronic, from which someone needs to identify the essential supplier, address, amount and, crucially, the due-date. Artificial intelligence (AI) tools can read the paperwork with over 99% accuracy and eliminate the need for everything to be manually checked. This saves time and money which could be better spent doing productive activity, and facilitates the timely payment of invoices. This trial tests an idea to speed up mainstream adoption of AI. We believe most SME owners would rather be working on their business than keying in their accounting paperwork. But we understand the natural reluctance to change without proven demonstration of impact within their particular business. Hence we provide a prolonged free trial of the software. This should tackle the cost barriers. And we intend to hold the hand of SMEs through the learning barriers by providing tailored support, webinar masterclasses, accounting advice and tips to make it easy to integrate Evolution AI into bookkeeping practice. We compare this to the effectiveness of a more basic self-service model, where the SME is provided the Evolution AI product for free (but not the 'hand-holding' support package). That is the control intervention of the trial. It remains a financially attractive way to use Evolution AI, but we suspect it will not be as effective at stimulating SMEs to make the change. After the free trial period we have a second experiment that tests SMEs' willingness to pay for continued usage of what we believe will be a demonstrated cost-saver. Such preparedness to commit the SME's own money to the technology is the acid-test of innovation adoption.