A decade ago, business units like marketing and HR were primarily driven by individual preference, experience and instinct. That approach to either discipline won’t work in this day and age, where most businesses are painfully aware of inherent biases that may exclude minorities and other talented candidates. Conventional recruitment has long been labelled as expensive and time-consuming, with companies often starting from scratch for each vacancy. Worse, because it targets only active candidates – those who respond to an ad or are on a headhunter’s list – it can miss ideal candidates who are not currently looking for a job.
Research from recruitment firm, Alexander Mann Solutions shows that HR professionals are ready for AI to shake up their function. 96 per cent believe that artificial intelligence has the potential to enhance talent acquisition and retention. But what’s already in the market, what works, and what are some of the limitations that HR tech will inevitably have?
HR tech: a hot new horizon
HR has the potential to become one of the most data-driven functions in business if it can challenge the fragmented and outdated traditional recruitment market. According to research from data analytics firm Alteryx, HR is the next untapped business function to reap the benefits of adopting a data-led approach. “As the benefits of big data become increasingly obvious, I believe we will start seeing greater adoption of big data in sectors such as HR that have historically proven reticent. Thanks to a broadening analytics culture and self-service analytics systems that enable everybody, regardless of technical knowledge, to understand and decipher big data, I would expect to see even the most qualitative of businesses look to embrace the revolution,” says Alteryx chair and CEO, Dean Stoecker.
For the human resources department, Stoecker explains there’s a wealth of analytical techniques to turn on, from capability analytics that measures the organisation’s talent, to competency acquisition analytics that assesses the acquisition of skills, capacity analytics that assesses the efficiency of personnel, and employee churn analytics. “Going deeper, there is corporate culture analytics, recruitment channel analytics, leadership analytics, and of course, employee performance analytics – which is where analytics might start, but needn’t stop.”
International recruitment firm, Alexander Mann Solutions has trialled several Robotic Process Automation (RPA), which aim to make the routine elements of the hiring process as efficient, timely and accurate as possible. The goal is to free up HR teams to do much more meaningful work, according to the firm’s director of technology and operations consulting, Laurie Padua. “Our robot ‘DORIS’, for example, manages the volume filing and arrangement of documents without errors, while ‘ISAAC’ can schedule thousands of interviews. As a result, some of the blue-chip brands in our client base participating in the trials have recognised the potential to scale these solutions in the long-term,” she adds.
From analysis conducted by GR8, recruiters spend between 40 per cent and 60 per cent of their day sourcing candidates, so effective automation can free-up around three to five hours per day, Padua adds. “(This enables) hirers to achieve true productivity. Particularly when you consider that 23 hours are typically spent screening CVs for each individual role.”
Alexander Mann Solutions has found that AI in recruitment is particularly useful when managing error-prone processes such as complex HR forms, compliance and on-boarding, as well as ‘swivel chair’ processes which involve jumping between tasks. “However AI has its limitations,” says Padua. “Anything involving ambiguity, context or change requires a level of emotional intelligence which only humans can provide. These tools are not intended to eliminate the role of HR and recruitment professionals, they are designed to relieve them from repetitive and low value tasks that consume valuable time which could be best spent leveraging their interpersonal skills and engaging with employees and candidates.”
According to futurists, HR is the new horizon for disruption. But the question shouldn’t be how can HR outpace marketing as a data-led business function, says Inma Martinez, a venture partner at Deep Science Ventures, Imperial College’s tech-sciences innovation accelerator where is mentors data sciences and product innovation projects. “The question should be how do recruit humans in a super digital AI world,” she says. “Machines are doing all the quant stuff so we need humans to think about the customer experience, new levels of service, new ways products can be fun and useful. What you want from people is creativity and a focus on innovation. It’s not so much about processing. Jobs are now being done by machines. Their managers are humans. That’s definitely where data and technology in recruitment and human capital is moving towards.”
According to Martinez, marketing wasn’t always seen as a purely data-led business function. “I remember in the old days, no one from the marketing department made it on the board, just like how in many companies today, no one from HR makes it on the board. In marketing, everything is about data; are you hitting your targets, is your campaign successful, what are the metrics for that success, and so on. It’s extremely quantifiable, but in the old days, marketing wasn’t seen that way. People used to think marketing was an abstract concept no one really understood,” she says. That changed, especially when boards began to realise how marketing can impact the bottom line.
“HR needs to be seen the same way. Boards only react to market circumstances. If in the next six months, boards heard how investing in HR technology leads to better profits, they’d all want to implement it. It’s a lemming market, unfortunately. Once one company says ‘we’ve cracked the code. This is what makes us competitive,’ everyone will sit up and listen.”
Start-ups riding the AI bandwagon
The productivity and time management argument has lead innovators to develop AI-heavy HR technology. The second biggest argument is how AI can counter inherent human biases. But is the technology truly ready?
According to Martinez, software that just scrapes LinkedIn for data won’t hold any real value for businesses with a specific ask. Technology can select candidates based on criteria companies have already outlined, but how does that help level the playing field for candidates who don’t fit those traditional definitions, but may still have the skills to do well in that role? Martinez calls this tacit data, and something that is exclusive to human assessment. “Technology can’t access bio data like ‘does she look alert, does she seem passionate about the role, will she fit in with the culture.’ This can only be assessed when you meet the candidate. This kind of tacit knowledge can never be taught to robots. It can’t be put into words, explained to someone, or programmed into AI software.”
Martinez warns that many programmes may enter the market to help assess feasibility of candidates, but there’s still a need for peripheral data on how these people will fit into the company. “About 56 per cent of mergers fails due to human discord. They don’t get along, the boards are fighting, no one agrees on who the CEO should be, and eventually the people destroy the deal. It’s important not to forget that human element,” she adds.
One start-up in the HR tech space is Snap.hr, which combines AI and machine learning technology with a specialist in-house team to match companies with skilled tech talent. To achieve this, Snap.hr is vastly improving matching, transparency and communication between businesses and jobseekers for better hiring, and higher retention rates.
The start-up already works with over 1,600 companies, including Skyscanner, Transferwise, Tesco, and Moo, and over 12,000 tech specialists, all screened and vetted for suitability. According to Snap.hr founder and CEO, Raoul Tawadey, 95 per cent of candidates will respond within 48 hours of being contacted by a business. The average time to hire is 12 days, which is half the industry standard.
Similarly, within the first week of being on the platform, jobseekers will receive an average of five interview requests and over 95 per cent of candidates who have been placed via Snap.hr are still at the same company after six months.
“We’re building the future of recruitment starting with the most in-demand technical talent. General platforms like LinkedIn require a huge time commitment and traditional recruiters can lack transparency and tend to push their own agenda. Snap.hr combines cutting-edge AI with a focused team of recruitment experts, to offer companies the opportunity to interview and hire Google-level tech talent quickly and easily,” says Tawadey.
Another start-up behind AI-powered recruiting software is Beamery, which works with the likes of Facebook, Dropbox, House of Fraser and a host of others. The recruiting software firm recently raised $5 million in funding in a round led by Index Ventures to double its staff and open a new office in the Bay Area.
What makes this AI fuelled software start-up different is that it allows companies to treat candidates like valued customers and turn recruiting into strategic sales and marketing, it says. Beamery’s customers report a 39 per cent reduction in costs and 31 per cent acceleration in time-to-hire.
“Talent and HR used to be cost centres, necessary evils rather than unique advantages. In the modern economy, companies are starting to recognise that creating a world class talent function is one of most important differentiators to their success. Beamery transforms how businesses interact with talent, and is a primary system of record for this relationship,” explains Beamery CEO and co-founder Abakar Saidov.
Both Snap.hr and Beamery tout the power of AI as a way to cut down cumbersome processes while ensuring that their software can continually improve. Tawadey’s vision for Snap.hr is that by processing data such as skills, requirements, qualifications, interests and salary expectations, the platform then lets companies contact desirable candidates directly (and vice versa) creating a faster, open and transparent mode of communication. Features like the shared diary means that companies can book in an interview immediately as they see a candidate is available, eliminating the need for a third party recruiter to make arrangements.
Candidates also have the power to make informed career choices through the platform. They can check their current salary against market rates and are also given the option of a personal talent manager in order to provide them with a long-term career plan. Tawadey believes that transparency is key to successful hires in the job market. Businesses can assess their competition and view exactly where their chosen candidates are interviewing. As well as this, Snap.hr’s technology can be integrated with other software services, such as Slack and Workable, to make hiring and finding a job a part of other daily functions.
Cutting through marketing speak, can an AI recruiter ever truly replace human intuition? More importantly, should it?
The diversity dilemma
Traditionally, recruiters have relied on CVs and gut instinct, which too often lead to the recruitment of a workforce that’s not diverse and who struggle to provide an alternative view point. And according to talent management company, The Chemistry Group, the hidden costs of a wrong hire can amount to anywhere between four to 20 times the salary. The company uses technology to find the right people for the right roles, ranging from psychometric tests to a Facebook app created for SAP, to prevent hiring bias.
Roger Phillby, CEO and Founder of The Chemistry Group, believes the use of technology, such as AI or big data, can lead to a more objective filtering of candidates, particularly for big companies. “You still need human chemistry at the end, but subjective screening, such as phone interviews, at the beginning of a recruitment process is time consuming, bias and financially draining – it’s bad for both the company and the candidate,” he says. To find the candidate for your company the use of technology can be easily incorporated into the process, he advises that businesses should define what ‘great’ looks like for their company and seek to find those people regardless of anything else. “Your selection process should be completely neutral, and the use of data and algorithms helps to achieve this by using computers to eliminate any discrimination of any kind,” he adds.
Another AI-led HR analytics start-up, Saberr, looks specifically at getting that culture fit just right. Saberr works with Microsoft, Seedcamp, Coca-Cola and Virgin, using an algorithmic approach to help businesses hire new employees and manage existing teams. CEO Tom Marsden believes taking a data-led approach can help encourage diversity in hiring. “Algorithms can help avoid bias because they’re blind to pre-existing prejudices and the demographics of the existing team allowing managers to go beyond gut feel to systematic data driven assessments,” he tells GrowthBusiness. “Algorithms have the key benefit of being peer reviewed. If there is a hidden bias in an algorithm we have the ability to break apart the process and scientifically analyse what is happening. We can’t do this if the hiring process is based solely on the gut feel of individuals. Unlike human bias, we can apply scrutiny and change the internal logic of the system to increase fairness.
Replacing human intelligence
According to Snap.hr’s Tawadey, in five years from now, when most people look for a job, it will be via a platform with minimal human intervention and with the bulk of the matching done by methods like AI. “If you ask employers and candidates their problem with traditional recruiters, or platforms like LinkedIn, it’s pretty much around matching. Both parties get annoyed when a recruiter mismatches them, because they’ve failed to understand their skills or what they’re looking for. What people want to do and what they’re good at is all just data that can be put through a statistical model and used to create recommendations. Luckily, we’ve found that there’s enough stickiness on the demand side to collect large data sets, so that we can continually test and learn from recommendations,” he adds.
“Our AI is broken down into two aspects. Firstly the data-mining part, which involves classifying and learning who a candidate is; secondly the machine learning part, which involves learning what a good match for a company looks like.”
When most people look at a profile or a CV, they see words and pictures. When Snap.hr’s system looks at a profile, it learn about the trends in their work history and use that to figure out what a candidate might like to do next. The platform then matches these candidates with companies and use data around the entire hiring funnel to either correct or refute our assumption. For example, if a matched candidate passed the technical assessment but fails the final round interview, the system would learn that the candidate is technically capable but failing culturally, and would start recommending candidates who could be a better cultural fit.
“That ‘human intelligence’ that we’re artificially emulating is that of a recruitment agent who would spend their time creating recommendations for companies. Interestingly, in the early days of Snap.hr, we hired a recruiter so that we could learn everything about how they think and how they manually do matching – so that we would know the steps required to automate it,” he adds.
Ultimately, AI is the must-have technology of our time. AI is the power to make better decisions, and businesses live and die by the decisions they make, as long as it’s tempered by human expertise.