How to balance AI and automation with humans in hiring

There’s no escaping automation in today’s world of work. It speeds up processes, reduces the labour burden on workers and drives consistency. But as last week’s topic revealed, you can have too much of a good thing.

A connection of mine, Jan Tegze , Director of Technical Recruiting at Tricentis , was discussing this exact issue after he came across an ATS that claimed to screen out over 50% of applicants before they even make it to a real, live human being.?

Among other things, Jan wanted to know how you can guarantee you get the best possible shortlist if a machine screens out such a large volume of candidates.

Fresh off the ChatGPT article, with the runaway chatbot still living rent-free in my head, I couldn’t be blind to the obvious crossovers between the two.

Over-reliance on AI and automation often generates as many if not more problems than it solves. Therefore, if we’re going to reap the benefits of these solutions, we need to be targeted about the ways in which we put them to work.

So this week I’m going to be discussing how you get the balance right between robots and humans to ensure you hire more efficiently and avoid the pitfalls of over-automation.

I’ll be looking at the following:

• The big wins in simple automation
• The bigger wins in more complex AI and automation
• What happens when you over-automate
• What humans do better than bots
• How to strike the perfect balance

Simple automation: small changes that make a big impact

So right off the bat, I’m going to start with what you can automate without investing much time, money or effort.

Candidate communication

Candidate communication is a time-consuming but crucial element of every hire. According to CareerBuilder, “38% of workers say they’ve been frustrated with employers not acknowledging receipt of an application.” So if you get this right, you’re going a long way toward delivering a quality candidate experience.

While some ATSs might claim to screen out 50% of applicants before a human ever lays eyes on them, any of the top ones will offer you the ability to automate acknowledgment of receipt. And they will also have additional email automation and reminders to help you keep candidates updated.

Automating responses and updates in this way is an easy win when you consider data from the same CareerBuilder survey that says, “67% of job seekers have a positive impression of a company if they receive consistent updates throughout the application process.” These updates could include the following:

• Notifications for receipt of application
• Interview reminders to the candidate and your team
• Reminders to your team to provide feedback
• Interview follow-ups to the candidate

These simple automation help you deliver that crucial gold-standard candidate experience whilst simultaneously contributing towards your time and effort-saving goals.

Two figures that illustrate that point perfectly are from a Dice report that shows: 44% of recruiters say they lose up to 10 hours a week on tasks that could be automated, and 55% of recruiters say automation has helped save time on tedious manual tasks.

Introducing more advanced AI and automation to the hiring process

It’s likely I haven’t told you anything new so far. We all know the main function of automation is to deliver results faster and with less effort.

But can it deliver better results?

Improved candidate sourcing

Automation and AI can help you source candidates more efficiently by using advanced search algorithms and machine learning to identify the best-fit candidates for the job. It can pull important information from a much broader range of sources and ensure you have access to all potentially relevant data.?

This is particularly useful for global hiring as not all regions have as extensive use of Linkedin as Europe and the US.

This means you can make more accurate, data-driven decisions. It also frees up more time for you to focus on those higher-level tasks such as engaging with candidates and building relationships.

Streamlined screening and filtering

This one relates to the issue that kicked off this week’s article. Automated screening and filtering of candidate CVs or applications can save you a significant amount of time. This is particularly useful for high-volume recruitment, where you’ve got to review hundreds or thousands of applications or resumes.

Whilst we’re right to be sceptical about an ATS that filters out 50% of applicants, when recruiters manually screen CVs, they are only dedicating 6-8 seconds to decide whether a candidate is suitable or not. Therefore, it would be a stretch to suggest that this process is any more thorough or accurate than automated screening.

Enhanced data management and analysis

Automation and AI can help you manage candidate data more efficiently, reducing the administrative burden on recruiters.

These solutions can provide near-instant insights and predictions as to where the market is and how it’s moving through analysing data sets that would take years for an individual to read through.?

You can benefit from a better understanding of the industry and the role you hire for. And you can understand with a greater degree of clarity which skills are becoming more and more valuable for companies and which characteristics to look out for in potential hires.

Data analysis tools can also provide insights into recruitment metrics such as time-to-hire, candidate conversion rates, and cost-per-hire. All of which allows you to make data-driven decisions and improve your recruitment strategies.

Reduce human error and bias

Human error in the recruitment process can result in duplicate candidate entries, incorrect contact information, or missed screening questions. All of which can lead to missed opportunities and wasted time.?

The use of automation and AI can alleviate the burden on recruiters and ensure that all candidate information is accurately recorded and processed.

These solutions can also reduce bias throughout the process. For example, automated screening can use standardised criteria to evaluate all candidates. And an ATS can remove candidate names and other identifying information from resumes during the process.?

Automated assessment tools such as pre-employment tests and AI-powered video interviews can provide an objective assessment of candidate skills and abilities. And automation can use AI-powered sourcing tools to identify and attract a more diverse pool of candidates, increasing the likelihood of you selecting candidates from underrepresented groups.

What can happen when we over-automate?

If you automate too much of your recruitment cycle, you risk tipping the balance. And when that happens, benefits can quickly turn into disadvantages.

Missed opportunities for quality candidates

An over-reliance on automation can result in you missing quality candidates who may not fit the exact criteria programmed into the system. This can happen if you only rely on keywords and screening questions to filter out candidates.

Think once again about that ATS that screens out 50% of applicants. How is it doing this? If it relies on keywords, it’s relying on applicants using those keywords in their CVs or applications. This tends to favour applicants that have optimised their resumes for ATSs or resume parsers, rather than those that are genuinely the best candidates.

Employers and recruiters know that this is happening. The Harvard Business School paper, Hidden Workers: Untapped Talent shows that 88% of them say that qualified high-skill candidates are vetted out because they do not match the exact criteria in the job description. This number jumps to 94% in the case of middle-skill candidates.

These numbers seem to correlate with data from CareerMinds that shows 75% of resumes are never read by a human. Whether or not this means recruiters are currently over-optimising their screening process, I’ll leave it up to you.

Increased administrative burden

Instead of reducing the mundane tasks required of recruiters, over-automation can actually result in an increased administrative burden. This can happen if the system is too complex, if there are too many rules and conditions programmed into the system, or if insufficient time is allocated for onboarding recruiters.

Problems like these can result in pushback from recruiters and result in hiring teams abandoning solutions altogether, even though they’ve been tied into lengthy and expensive contracts.

Difficulty in adapting to changing recruitment needs

An over-automated system can be inflexible and difficult to adapt to changing recruitment needs. If you need to pivot quickly to address changing market conditions or job requirements, an overly automated system may be unable to keep up. This can result in missed opportunities and an overall loss of business.

What humans do better than bots

So, where to avoid using automation? Well, we’ve been talking a lot about the importance of using automation and AI to help deliver a great candidate experience. However, if you rely on it too much, it can lead to a lack of personalisation in the recruitment process, which can negatively impact the candidate.

Too much automation can leave candidates feeling like they are interacting with a robot instead of a person, and they may become frustrated with the lack of human connection.?

This is evident in data from the aforementioned CareerBuilder survey, which shows that “71% of workers agreed that automation in the hiring process makes it impossible to tell if the company’s culture is the right fit for them.”

As a result, the conversations through the hiring process remain a decisive factor in the best candidates signing on the dotted line and why automation should be considered as a method to prioritise the important conversations.

Every minute you spend speaking with a candidate is an opportunity to build their confidence that the role and company are a great fit for them. That said, your main challenge as a recruiter is to make sure your time is spent having impactful conversations with the right candidates.

Overall, automation should be seen as a supportive tool to free up recruiters’ time and enable them to deliver the best possible candidate experience.

And let’s not forget how crucial delivering the best possible candidate experience is now. With the shortage of talent at a 17-year high – according to ManpowerGroup’s Talent Shortage survey – and four out of five employers globally reporting difficulty in finding the talent they need, competition in getting to the top talent first and retaining it through the process has never been more fierce.?

How to strike that perfect balance between bot and human

Balancing between AI, automation, and the human touch will determine the success of your recruitment process. But how do you achieve that balance?

Focus on the candidate experience

Mapping out your ideal candidate journey from application through to day 1 at the company is the first step. You can then identify where automation might enhance their experience, and where it might jeopardise it. This will help you identify the quick automation wins, the automation that you want to work towards, and the areas you want to leave (for now at least).

Set your team up for success

Your recruitment team remains crucial in attracting the best talent, they are the crucial point of contact for helping a candidate understand your culture and ambition. Listening to their frustrations and pain points can help direct where improvements can be made and where AI and machine learning can contribute.

Be pragmatic with new technology

It takes time to build trust in new technology, particularly in more advanced automation, AI and machine learning. The first step will always be to test the technology alongside the existing process so you can understand the benefits and the limits. You can then build confidence in what AI and machine learning can contribute to your hiring process, without having unrealistic expectations of what they can deliver.?

A best practice approach, regardless of the tech

This approach is required regardless of the technology we are adopting. Whether it’s a new ATS or generative AI.

Last week, I joined a conversation with Irina Shamaeva ???? , Partner, Brain Gain Recruiting, Greg Hawkes , Principal Sourcing Leader M&T Bank, and Enrico Heidelberg , a freelance Senior Recruiter on Recruiting Brainfood , where the discussion was centred around prompt engineering with GAI.

This concept might seem new to many of us but its aim is being targeted with GAI in order to yield the best possible results from it.

Rather than just employing this new technology without first understanding how to use it to its full potential, recruiters have been exploring and experimenting with where it can be used most effectively. And early signs are that it can be particularly valuable in sourcing.

The key is to treat it not as software like Google or LinkedIn, but as an iterative, two-way conversation where you gradually steer it toward the results that you need. Again, as with automation, it requires the guidance of an informed human.

As Irina said during the conversation, “If we give it poor prompts, it spits back something watery or nonsensical.” Therefore sourcers, or prompt engineers, need to “push it and dig deep like gold diggers,” in order to get the best results.

And that’s the case with AI and automation more broadly. It’s not a magic bullet. However, by following the guidance outlined in this article, you can begin to build a strategy for your business that will see humans and AI working together to help you hire more efficiently and successfully.