Data for people and talent: How can HR Directors improve their hiring, promotion and retention using AI?

Human resources is not traditionally a focus area for the application of data science.  It is sometimes seen from outside as one of the ‘squishy’ area of a business, where performance is measured using value judgements and vaguer criteria such as ‘cultural fit’ and other subjective measures. This can cause a lot of problems. And a lot of wasted effort, wasted opportunity and wasted talent. The bias and frequent lack of rigour in human decision making is well documented. Forward-thinking HR leaders are changing this. There are in fact many area where HR can be, with a little effort, one of the most data-driven areas of a business. I will lay out a few of these here.

First, data-driven hiring practices. Hiring is an enormous expense in both cost and time. Many companies collect top-level descriptive statistics on their employees, perhaps broken down by division or business area. However, more interesting conclusions could be drawn by analysing factors such as the language used in job applications – is it possible to tell, by analysing CVs and cover letters, how successful an employee will be at the company, and what their annual appraisals will say in 1, 2 or 5 years’ time?

Second, data-driven appraisal and promotion. What about the language used by managers in appraisals. Does it vary by gender? There is extensive evidence that women are criticised for being ‘bossy’, while men are praised for being ‘confident’.[i] Does this hold in your company? It is very easy to find out, using text mining techniques. This goes equally for objectives such as under-representation of ethnic minorities at senior levels. At what point do they fall out of the talent development pipeline. Junior levels, mid-levels, or just before they reach the top? Do they tend to leave or to stall but stay with the company. What exact career path have your top executives followed? Do you have enough female and minority employees treading that path?

Third, reducing employee churn. Churn is the flip side of hiring – if the employee is leaving for a ‘better’ job, it often reflects failure by the company to offer the leaver’s talents a suitable outlet. As mentioned above, this can be particularly important for under-represented groups. If your company promotes largely internally, you can’t have many senior minority executives if you don’t have enough senior minority managers. I once undertook a project for a large government department that was undergoing a major HR transformation programme, including changes to employment contracts, which generated substantial discontent among staff. I developed a model that automatically collected and analysed the results of a free text survey that was answered by in excess of 20,000 staff. I communicated to top leadership the issues, sentiment and suggestions made by staff, and the exact reasons for employee dissatisfaction. This allowed the leadership to make minor tweaks and changes to T&Cs that defused tension by addressing many of the issues raised, without diluting the thrust of the reforms. Without this sort of granular information, it is very difficult to improve employee retention.

Fourth, return of investment of skills training, learning and development programmes. Companies can spend very large amounts of money on staff development, often several thousand pounds per employee per year. The ROI of these programmes can often be amorphous and difficult to identify, but there is no reason for this to be the case. Using similar techniques to the second point above, it is possible to track employee appraisals having undertaken different training opportunities and gain an indication of exactly how specific new skills may have improved the ability of an employee to excel in their position, or made them suitable for promotion.

[i] https://www.progressivewomensleadership.com/how-gender-bias-affects-your-performance-reviews/