Believe the hype? The role of AI in staff engagement

The discussion over the role of AI in the HR has become very polarised between enthusiasts and sceptics. Enthusiasts believe AI will be transformational and help to automate large amounts of administrative tasks, as well as offering enhanced analytical tools for understanding data. They point to successful implementation in recruitment activity. 

Sceptics say the claims for AI boosting productivity are based on questionable evidence, that AI analytics are as yet unproven, and in HR implementation it has been largely confined to early stages of recruitment. 

A similar divide can be found around the potential use of AI within the field of employee engagement. Some commentators see a big role for AI in supporting analysis of employee views in the future, potentially reducing the role of employee surveys and offering real time feedback. Others argue that existing tools are not yet sophisticated enough to rely on and that AI analysis needs to be supplemented with human insight. 

The jury is still out on the wider claims, but there is one important area where AI is already proving its worth: in analysis of free text comments. Free text comments can be a rich resource in helping to understand employee views. They allow individuals to express themselves in their own words rather than having to select a tick box in a survey question. 

Free text can be analysed using sentiment analysis to get an idea of employee mood. Comments have been collected as part of the NHS Staff Survey and in many local surveys for many years, but until the advent of modern AI tools, analysis of responses was time consuming and could mean using tools developed for other purposes. 

An AI tool allows comments to analysed quickly for key words and statements to be categorised using locally generated bespoke thematic classifications. As the tool is working with text supplied to it, it should be reasonably reliable and not usually require any specialist training. 

Commercially available free-to-use AI tools are currently not very nuanced in analysis and may not be familiar with NHS terminology. Trusts have sought to address this via prompts/theming and in some cases developing NHS specific tools with local partners.

There are potential benefits of being able to analyse free text comments more effectively and NHS Employers is seeking to share learning on this topic.

A number of trusts have begun to make use of AI to analyse free text and shared their learning in the most recent NHS Employers staff engagement network call. In the call, most organisations had used commercially available tools to analyse free text by designing prompts to classify responses. These prompts were designed around the NHS People Promise themes and enabled them to see more easily where there the balance of comments were focussed. It did require work to design the classification framework and review all the comments to ensure correct allocation and get an overview. Most trusts have used generic tools.

NHS Employers is keen to share learning on this topic as it develops. Please get in touch if you would like more information on the approaches that have been developed or have experience of your own to share. Hopefully, on this aspect at least, the NHS can lead the way and use technology to ensure it makes best use of the views shared by staff.