All companies would like to avoid or minimize employee turnover costs. Various evaluations have estimated the cost of a lost employee at about 1.5 to 2 times his annual salary, also counting the money spent on recruiting and education. The shorter the time an employee spends in a company, the higher these costs are since it takes a few months for any new hire to start producing value. 

In this situation, there is a tool that can change the game completely. Artificial intelligence (AI) is already used by other activity sectors like retail, finance, and production with astonishing results. The techniques and ideas that yield returns in these areas could be adapted for HR.

Lessons from retail

The retail world is all about matching the client with the product and selecting from an ocean of possible leads those that fit the target market to channel the advertising efforts. HR follows a similar pattern in creating profiles, selecting the best candidates and creating a connection with those who matter.

Creating a candidate profile

Online retail chains have systems in place that scan their existing clients and create profiles of the most profitable ones based on demographics, behavior, and cross-purchase. These systems can be adapted by HR in scanning their best employees and extracting those features that are exhibited by the best and who have been onboard the longest.

Attracting and sorting candidates

E-stores have also perfected the art of attracting the most promising leads and guiding them along the sales funnel. This is done by analyzing the users’ social media profiles and interactions with competitors. Similarly, the HR department can use these techniques by scraping information from dedicated platforms like LinkedIn and using third-party data channeled into neural networks to see if the candidates are also applying to competitors. Sorting resumes based on keyword analysis, and natural language processing is another way of saving time for human recruiters.

Evaluating the fit with the organizational culture

When hiring, Google does not only look at schools, GPA or previous employers, but also “googliness” and this is different from one company to another. AI can scan their personal records and highlight the patterns of what makes a good employee for each sector of your organization then see if the new candidates belong to the same clusters and in what percentage they match the needs of the organization.  The AI can perform dynamic matching of candidates with positions.

Lessons from finance

Finance is all about getting the best return on capital. To ensure this, there is a constant evaluation of performance, followed by adjustments and portfolio creation. The same tool used to value stocks can be repurposed for evaluating employees and creating a “portfolio” of satisfying benefits.

Evaluating performance

Financial environments are already using AI to assess the performance of stock exchanges, currencies, and commodities. With the appropriate changes, the machine learning algorithms could be taught to evaluate the employees on different levels. Depending on the industry, if there is a substantial set of KPIs, the deep learning algorithms can be trained to recognize the connections that lead to the desired outcome. A relevant example is given by assessment of sportsmen, where a baseball team acquired a supercomputer.

Creating the perfect benefits mix

For stock investments having the right combination of assets in your portfolio ensures a winning strategy. HR can borrow this idea and treat salary and other benefits as portfolio items. Generating the best combination of money, flexible program and medical care, increases retention.

Lessons from the production environments

Data has been of paramount importance in production environments to minimize waste and increase quality output. It’s all about well-defined systems. The HR component in a company is a system itself and can be treated as a distinct silo, which needs to hit certain performance metrics.

A predictive approach to staffing

In the past decades, the recruitment process was set in motion once a need appeared in the company. Due to the velocity of the modern world and the scarcity of the human resources in some areas, this approach is no longer viable. A good lesson from production is to predict, and stock before the actual need arises. AI can do this by creating development models for the company and identifying the areas where more talent will be required. This can set the recruitment and selection process in motion long before there is an immediate necessity, avoiding a vacancy.

Creating a system

Production facilities are all about replicable and coordinated systems. AI is already used on the production line with excellent results in quality, testing, and validation. The same ideas can be transferred to validating employees and their work. By defining performance metrics and keeping the system running in the background, each member of the staff can receive alerts or congratulations based on their individual performance daily or weekly. This is an agile approach to work and can prevent not reaching the target or struggling with work for weeks at a time.

Avoiding bias

One of the top management problems is making biased decisions. While some might argue that intuition yielded some of the best outcomes in business, data offers constant and consistent top results. Using a computer reduces the bias a human recruiter could have towards a candidate. Even in equal opportunity environments, personal judgments could interfere with real talent.

Making the transitions from psychology to deep learning

Traditionally, managing human resources was done using motivational factors and even the people hired in this department have a background in psychology. Using AI solutions to identify, select, recruit candidates and retain workforce will be a paradigm shift. However, these things are not exclusive, but complementary. The future of work is a collaborative one between man and machine, or, how the creator of Siri calls it, humanistic AI. Learning from the healthcare department where AI is used to help oncologists detect cancer, HR can take a similar route. Instead of thinking about giving full control to deep learning algorithms in making new hires, just use the machine as an automated assistant to humans. After all, no matter how sophisticated, AI has yet to learn about common sense.