Employing a combination of AI for hourly hiring, human interaction, and an ATS that facilitates both is a powerful combination to win the hiring game
2023 has quickly become the year of artificial intelligence (AI), especially in the world of work. Professionals from software engineers to medical researchers are using ChatGPT and similar generative AI platforms to write code and discover new drug treatments. The recruitment space is no less impacted by this new wave of intelligence.
Aptitude Research found that 63% of companies planned to invest in recruitment automation but were unsure of what to automate or how they currently automated. With the recent explosion of conversational AI technology like ChatGPT, there’s bound to be an even greater gap between awareness of AI and possessing the confidence to use it effectively in hiring procedures.
Fountain examined text interactions between employers and applicants to understand the dynamics of messaging behavior and how employers can bridge communication gaps using AI when it’s needed most.
Our findings highlight the following:
- Hourly employers afraid of losing the human touch cannot rely solely on their hiring managers to take on all applicant communication.
- AI for hourly hiring is most beneficial in identifying what we term as “actionable” messages, where real interaction can make the most impact, perfectly balancing human and artificial touchpoints.
- There’s a definite emotional cost to applicants when they’re reliant on human communication during the hiring process, which has a downstream impact on your hiring outcomes and implications for the emotional well-being of your hiring managers.
Hourly hiring in a mobile-first world
The hourly hiring market operates very differently compared to desk-based roles. Hourly applicants follow a path of least resistance when it comes to accepting a job offer. Speed matters most, and these applicants are applying to multiple roles at once and are most likely to accept the job that makes the fastest offer.
Mobile phone applications have the highest prevalence in gig and hourly work, with 86% of gig applications, like delivery drivers, performed on mobile phones followed by transportation at 76% and warehousing and logistics at 74%. This falls in line with the emergence of Gen Z in the workforce, now making up nearly 20% of the working population in the United States, with 37% of the population comprising the hourly-driven leisure and hospitality industry.
Gen Zers coming into the workforce are seeking multiple jobs at once, looking for shift flexibility, and wanting to apply directly on their phones. For employers, this means ensuring application processes are mobile-friendly, but also rethinking applicant communication for hourly roles. While many applicant tracking systems (ATS) offer texting capabilities, employers in the hourly hiring market will need to ensure they’re set up for high volumes of text while also thinking about optimization overall.
A key challenge in the hourly hiring market is not just keeping up with applicants’ technological preferences, but also considering the demands on individuals doing the hiring. In the hourly space, hiring managers are not usually desk-based recruiters but busy store managers juggling many responsibilities with limited computer time. Even mobile-friendly hiring processes may not be truly “friendly” if these managers are required to respond to every applicant question, manage scheduling, and follow up on incomplete applications.
Additionally, an examination of the most common rejection reasons within Fountain showed unresponsiveness as the number one reason applicants are removed from the hiring process.
For context, Fountain customers last year engaged in 200 million texts, with 97% of those coming from the employer side of the conversation. The busiest locations saw over 1,000 texts in a single month that received manual responses from the employer. At 20 seconds per text, those responses made up an entire work shift. Busy summer and holiday periods only exacerbate this problem, and expecting hiring managers to engage with an applicant who has already ghosted the process worsens the time expenditure issue.
While employers know faster responses mean a greater likelihood of landing a hire, telling busy hiring managers to “go faster” should not be part of a strategic solution.
Rethinking applicant conversation
The use of AI provides one avenue for employers seeking applicant engagement while reducing human labor demands. AI “bots” in hiring workflows are already widely used to screen and filter applicants. Conversational AI technology, commonly referred to as a “chatbot”, is also embedded via browser, text, and instant messaging software to handle applicant queries and collect information around the clock. This chat-based technology relies on natural language processing (NLP) techniques to mimic human communication, for better or worse.
Harvard Business School, in partnership with Accenture, found 27 million “hidden workers” in the United States alone, many of whom had been pushed out of employment by AI trained to filter for specific keywords and skip over applicants with easily-explained employment gaps. Additionally, the cost of a negative chatbot interaction can lose companies as much as 30% of their site visitors, along with poor reviews that live online in perpetuity.
Employers who use only chatbots to handle their application processes face high risk, especially if the NLP model they use was developed even more than a year ago as it would already be considered outdated. The attraction of speed and automation, in this case, is much more likely to sour applicant relationships in situations of high complexity, technical hurdles, or simply when applicants prefer to speak to a real live person.
The good news is that employers do not need to go “all in” on conversational AI. Instead, they should evaluate not how AI can communicate with applicants, but how AI can be used as a strategic partner to target situations where human interaction matters most.
Fountain sampled 2.5 million text messages sent and received between recruiters and applicants in 2022. NLP techniques that are common to the world of chatbots were used to categorize message intent from applicants into two buckets: “actionable” versus “non-actionable”. Actionable messages included any applicant text where help was requested, the applicant faced technical hurdles, asked a question, or provided an application update. We consider these messages to be those where human interaction is necessary. All other messages were classified as non-actionable, meaning no specific request or need was indicated.
Of the nearly 500,000 applicant texts, NLP identified that 30% fell into the actionable bucket. This establishes two things:
- Conversational AI could be employed to respond to 70% of applicant messages, but hiring managers are still needed for the remaining 30%.
- NLP can be used, not just to interact with applicants, but also to pinpoint situations where human intervention is most valuable.
Most interesting, however, is the relationship between time spent in the application process and the type of message sent. Fountain’s NLP analysis found a negative correlation between the number of actionable text messages sent and time spent in the hiring process. In short, the greater the number of actionable messages sent, the faster the applicant moved through that portion of the hiring stage.
This highlights that the volume of texts does not automatically mean longer processing times. AI technology that can flag these actionable messages for recruiters and hiring managers will create a communication process where truly engaged applicants move through the funnel at a faster pace, which means a greater likelihood of landing a hire, and an ultimately satisfied applicant. This also reduces applicant exposure to AI that may miss the mark and create a dissatisfactory applicant experience.
The impact of emotional contagion
How applicant emotion plays into the efficiency of the hiring process is also important to consider. Employers focused on creating positive applicant experiences can reap the benefits of employee referrals from new hires and generate positive brand perception across the board. But what happens when an applicant has a negative experience within your hiring process?
Fountain’s text message analysis also used NLP to label applicant sentiment into three buckets: positive, negative, and neutral. Most applicant text messages were neutral in content, with 25% skewing negatively and 5% skewing positively. Segmenting these messages by the eventual hiring decision of each applicant (either hired or rejected) showed that rejected applicants were two times more likely to send a negative text message during the hiring process.
This is a bit of a chicken-and-egg scenario: Were these applicants rejected because they were negative during the hiring process, or did they face difficulties during their application and expressed more negativity as a result?
While our examination did not arrive at a definitive answer to the question, we expect that many of these applicants may have been hired if they had access to better communication.
Additionally, it’s important to recognize the downstream emotional impact of compounding negativity on a hiring manager expected to handle all applicant communication. This is specifically an area where conversational AI, through sentiment tagging, can identify a struggling applicant early in the process and either resolve any questions and issues faster, or refer the applicant to a hiring manager right away to prevent further negativity.
Some of these applicants may not be a good fit regardless, but utilizing AI for hourly hiring as a strategic partner also benefits the employer brand by leaving all applicants feeling seen and properly evaluated in the hiring process.
Own the conversation
Communication with applicants, especially in the hourly worker space, does not have a one-note solution. Only implementing conversational AI in the process leaves employers prone to the limitations of a codebank. However, expecting recruiters–who are more likely to be busy store managers–to answer all applicant questions, resolve technical issues, and perform tasks like scheduling will leave applicants ghosted and likely to find employment elsewhere.
There must be a middle ground. Our text message analysis using NLP showed two areas where AI for hourly hiring can serve as a powerful strategic partner:
- A robust conversational AI can identify messages where a human touchpoint is needed to handle high-complexity situations and questions, leaving hiring managers responding only to messages where their input is most valuable, and
- Identifying applicants at risk of a negative application experience early in the process means salvaging a potentially lost hire and also negative brand perception
Employing a combination of “intelligent” AI, human interaction, and an ATS that facilitates both is a powerful combination to win the hourly hiring game. Fountain’s recent launch of Candidate AI Agent for hourly hiring has already seen hiring times drop by 87%, with an average time to hire of just 2.6 hours. Those are results applicants and employers can benefit from, and all without losing the human touch.
Learn more by reading the full white paper, AI for Applicant Engagement.
