HR & People Analytics: First 3 reasons to embrace Natural Language Processing in Human Resources
Updated: Jul 1, 2020
Did you know that 80% of business-relevant information is made up of text? This type of unstructured content is growing much faster than structured data. Think about the millions of words written every day on social networks, company forums, blogs, resumes, or comments in surveys.
There is a great opportunity for those who can leverage that type of data. For example, analyzing exit surveys and acting upon the results, will enable you to shorten recruiting time. It will streamline the hiring process or reduce absenteeism. You could also reduce the risk of possible litigation.
But, there is a problem. The information available is massive. When human beings have to read and extract insights from text-based sources, it is tedious and expensive. If you’ve ever done qualitative research, you know how time-consuming it is. Sometimes, it is even just impossible.
This is where Natural Language Processing comes in! NLP is defined as the ability of machines to understand and interpret human language the way we human beings do.
NLP is a sub-discipline of Artificial Intelligence. Artificial intelligence has been around since the 1950s. It has become increasingly popular over the past few years. Personally, I can hardly hide my enthusiasm about Artificial Intelligence. I share this excitement with an increasing number of people analytics practitioners. We see that AI is transforming HR departments from top to bottom.
Reason number one: We, humans, are rather talkative!
We love sharing our thoughts on social media. Several company forums are overflowing with comments. Thousands of suggestions reach us in surveys.
I have already mentioned it in the beginning: It is commonly accepted that 80% of business-relevant information is unstructured. It is made up of text, mainly. This unstructured content is growing much faster than structured data.
Natural Language has a huge potential as a source of valuable insights. But, until recently, it was rarely analyzed or used in decision-making. This was because the process is too time-consuming. Sometimes, it is even impossible to read and analyze thousands of lines of text. Natural Language Processing technologies automatically process and analyzes textual content. They provide valuable insights and transform this "raw" data into structured, and valuable information.
Reason number two: Chatbots are cool
I’ll assume most of you have heard about Amazon Echo, Google Assistant, Apple Siri, or Windows Cortana. These are all chatbots. A chatbot is a computer program designed to simulate a conversation with human users through Artificial Intelligence.
Chatbots are proving to be extremely popular in the HR world too. According to a recent Forrester survey, roughly "85% of customer interactions within an enterprise will be with software robots in five years' time". And "87% of CEOs are looking to expand their Artificial Intelligence workforce" using AI bots.
Most organizations are making an effort to improve labor efficiencies, reduce costs, and deliver better employee experiences. They are quickly introducing AI, machine learning, and natural language processing in their strategy. In recent years, chatbots are beginning to be part of the digital transformation agenda. This is having an impact on important HR areas, such as recruiting, onboarding, training, career path development, and benefits.
Reason Number three. More and more Open-ended questions
If you've ever been in a medium-sized or large company, it's very likely that you often had to fill out surveys. Most of them came from HR Departments.
These surveys are made up of open- and closed questions, which differ significantly and impact the way we interpret their answers.
Closed questions supply the participants with a specific line of possibilities. They ask them to choose one of the available answers.
An example of a closed question is a question for the Employee Net Promoter Score, which is: How likely is it you recommend this company as a place to work?
You then have a closed range of numeral choices, from zero to ten. These questions are easy to analyze, as it gives us a straight numerical answer and can thus be analyzed more easily.
In contrast to closed questions, open questions give participants the freedom to respond and comment on whatever comes to mind in a survey.
Such an open question can also be found in the Employee-Net Promoter Score.
Open questions are a company's most valuable source. The feedback that employees provide is extremely beneficial. They give us invaluable information about our strengths and weaknesses, the whys of it all. As a result, we’re able to see the areas where the organization can be improved.
During my professional experience in People Analytics, however, I have found a number of specialized labor climate consultants that only process closed questions. And, when they have to analyze thousands of employee suggestions through open questions, well, they sort of give up.
Some would manually analyze a 1% sample of the total responses and write their report based on that. They feel perfectly content with it too. I believe that this is not a good practice if you really want to listen to the people at your organization.
NLP technology is the solution. It automatically processes and analyzes text content and supplies valuable information, transforming that "raw" data into structured, manageable information.
Because let me remind you again of our NLP definition mentioned before :
Natural Language Processing is the "ability of machines to understand and interpret human language the way we human beings do."
We, human beings, are very good at understanding natural language. Arguably, we were born with that ability. We have been practicing all our lives too. So it seems simple and natural to understand and generate messages, detect ironies, ambiguity and.
The only problem with our species is that we don’t scale as well as machines do. If I read a book a day, I will need 60 thousand years to read the books of the Library of Congress of the United States.
For properly trained machines, on the contrary, it is a piece of cake to ingest those millions of books, analyze them and return potentially valuable data about the topics they talk about, the cities, regions or countries they mention.
This raw, machine power, gives us the ability to analyze the open-ended questions of surveys and derive meaningful insights, without having to scour through thousands and thousands of pages of text ourselves.
And... last 3 reasons come in the next post...
In the previous post, I gathered the first 3 reasons to embrace NLP in Human Resources.
Reason number four: Listening to the voice of the employee is a must
Have you heard about the Voice of the Employee? And, if so, are you already listening to the Voice of the Employee and leveraging its value?
For HR professionals, it is essential to listen to what employees say. Voice of the Employee programs systematically collect, manage, and act on employee feedback. The voice of the Employee, or VoE, collects the needs, wishes, hopes, and preferences of the employees of a given company. VoE considers specific needs, such as salaries, career, health, and retirement. It considers implicit requirements too. You will want to understand employees better.
Text-based data sources are a key factor for your organization to understand the "whys" and act on them to make improvements. For example, a survey like the Employee Net Promoter Score may tell you that the degree of satisfaction of an employee is 6 out of ten. The open-text question can reveal that she is not happy with the extra-hours, or with the fact that professional training is not sufficient.
Apart from the open-ended questions we mentioned in Reason number three above, there are other effective ways to gather employee opinions, suggestions, and innovative ideas.
Think about the value you can gather from in-company web forums, or from public social networks, employee panels, or even emails, whenever it is possible.
Reason Number Five. Exit interviews point out to necessary improvements.
An exit interview is defined as: "A survey conducted with an individual who is separating from an organization".Most often, this interview occurs between an employee and an HP professional.
The HP professional either records or simply writes down the answers.
The interview is very likely to include questions, such as:" "What are your main reasons for leaving?" "What did you like most, or least, about the organization?" "What, if improved, would have caused you to stay at the organization?" "Would you recommend the organization to others as a good place to work, study, or join?"
An organization can use the information distilled from an exit interview to assess what should be improved, changed, or, what should remain intact. Sometimes, no change is necessary.
More so, an organization can use the results from exit interviews to reduce employee turnover. Or, to increase productivity and engagement. Some examples of the values that can be derived from the analysis of exit interviews include: -shortening the recruiting and hiring process, -reducing absenteeism,-improving innovation, -sustaining performance-And, also, reducing possible litigation.
Reason number six: Automation of resume reading saves a lot of work.
A nifty application of NLP is about the extraction of relevant information from resumes. It definitely saves a lot of work as a recruiter. However, resumes can have many different formats and sometimes it may not be technically possible to fully extract all the relevant information from all these different types of formats.
But one can get started with simple steps and at least extract whatever is possible from some of the known formats. Even this will save a lot of work.
The approach we regularly take with resumes combines Machine Learning and linguistic engineering approaches.
Any AI solution of people analytics should serve to support decision making in human resources, especially in those areas where there is uncertainty or incomplete information. You could build and test your own AI prototype for Human Resources in 3 weeks by leveraging the AI expertise of our team. Test, before you leap!
Sign up for this program!
Find out how Byteflow's expertise can help you and your company.
Byteflows AI is a boutique NYC based data consultancy with a global reach. Our small team of dedicated AI & data science consultants and experts strives to deliver business value with AI.
[Healthcare, MLTC, Insurance, FinTech, NLP, Hedge Funds, Pharma, Political campaigns & more]
Contact us: email@example.com