How is AI going to transform recruitment in the customer management industry?
It’s both an exciting and challenging time for the recruitment industry. As the consequences of COVID-19 start to be felt, video will play an increasingly important role which we won’t turn back from. Additionally, smarter decision making across the recruitment process is going to be essential to allow larger scale recruitment initiatives to be undertaken at reduced cost. AI is going to be central to effective and efficient recruitment from now on.
As Humanotics launches our People Solutions team, I thought it was apt to provide a quick round up of the ways AI is taking hold and what technology is available.
If you want to do things quicker, use automation. If you want to do things smarter, you need AI
Market search
First, it’s important to understand the difference between automation and AI. If you want to get more of the same, you use automation. If you want to make smarter decisions faster, you use AI. If you are feeding the process with poor candidates, you will just be more efficient at finding poor candidates, period. Replacing mass searches driven by keyword search with AI powered candidate engagement tools is the first step. AI allows a much more sophisticated approach to matching job descriptions with potential candidates. Platforms will also learn from recruiter feedback to continually improve their search results.
Forget old fashioned job boards. Targeting the right advertising platforms based on the role and type of candidate required can also be done by AI. Recruitz.io finds candidates based on their online behaviour, and ‘hyper-targets’ them with the right job ads on social media, Google and elsewhere. Impressive.
Video analytics provides physical, emotional and behavioural insights
Initial screening
Initial screening needs to be fast and ensure the best candidates are prioritised at the next step. Increasingly, at this stage we are seeing less and less live interaction. Video recorded automated question and answer interviews are common and AI is enabling some innovative analysis to be conducted. Analysis of physical, emotional, and behavioural data, combined with conversation, sentiment and topic analysis is providing richer insights on individuals going through screening. It is easy to see outliers using this data which can pinpoint exceptional as well as unsuitable candidates more quickly. We ourselves are developing this capability with partners right now — talk to us if you want to know more.
Role playing
Role playing is important for contact centres to test a candidate’s ability to handle more challenging customer conversations.
AI powered tools such as CRIS can ask open questions and process the answers on-the-fly, scoring them against model answers and changing the direction of the conversation based on the response. These solutions are even allowing a certain degree of role play to be carried out at this early stage. While this approach cannot replace a human interviewer, this flexibility certainly makes the experience more personalised and creates a better impression on the candidates.
Role playing with avatars is coming
Avatars, such as those created by Uneeq, will soon be used to drive a more human like interaction during the screening and role play sessions. Companies such as Feedback Avatars are already showing that putting a face to an automated customer survey can increase response rates. Doing the same with interviews could put candidates more at ease and elicit better data.
Maintaining candidate interaction
Once a shortlist of candidates has been identified, maintaining a positive and continual conversation with them is important if they are to remain engaged in what can sometimes be a long process. Additionally, the contact centre could also be juggling with the recruitment of 10, 20 or many more people at one time.
Chatbots, powered by AI, are supporting recruitment teams dealing with ongoing candidate questions. These can range from simple things like ‘what’s my holiday entitlement?’ to ‘what’s the status of my application?’. The latter may seem straightforward but providing an answer that satisfies the candidate so she doesn’t feel like she needs to pick up the phone can be challenging. AI working at the backend can look at the data you hold about a candidate, see the status of the application and use Natural Language Generation to construct personalised responses that give a truly individual, human-like response.
Narrative Science and Wordsmith are software tools that generate narratives from data.
Continually improving the recruitment process
Coming back to my initial point about automation — if you keep doing the same thing, that’s what you will always get. AI does have its limitations and candidate bias is one that is well documented — get into some research.
Train the AI to deliver the right outcomes
Simply, you need to continually train the AI to help you make better predictions each time. This requires an end to end approach over a period of time, since you want to know — did you hire and retain a great employee in the right role? A few tools will be able to do this work for you, suggesting what are the common factors in each role and across the candidates that resulted in the best outcomes. This is quite sophisticated analysis and requires solid data, but applying Business Intelligence tools and a strong data science skillset will deliver useful insight.
Getting started — prioritising your goals
There are lots of technologies and options to think about. There isn’t one platform that will deliver a perfect solution to support your contact centre recruitment. Think about the outcomes that are most important to you to guide your first steps. Where do you need improvements?
- Too many candidates? Use AI not just automation to screen and filter potential candidates faster.
- Short lead times? AI reduces the workload with each candidate and ensures you have to go back to square one less often.
- Not reaching the right candidates? You can achieve a higher proportion of qualified applicants because you are targeting your advertising with AI.
- Poor retention? There are many reasons for this of course but AI can give you the best start since you are continually learning what makes a great candidate for each role.
The human touch
The point I mention frequently in these articles is that AI works best when you empower your people with it, rather than try to replace them. It frees up their time to focus on the more valuable activities as well as improving the performance of the recruitment process over time, not overnight.
Augmenting not replacing experts
Of course, there is still a role for outsourcing of recruitment to specialists who are already ahead in terms of making use of these tools with pre-qualified candidated and experts who can source niche candidates in those harder to fill roles. Paradoxically, finding candidates to fill many of the AI related roles is one area where the human touch is essential, as there isn’t enough volume to power effective learning models as yet.
Want to know more?
There is a vast landscape of technology options in the marketplace. A number of tools claim to provide many of the functions I have talked about. Olivia by Paradox, Arya or Skillate are examples. You can read more about these and other platforms in this article.
COVID-19 means that times are a changing in recruitment and contact centres. We need to look forward to embrace technology, while not losing sight of the fact that people, not technology, must remain at the heart of our talent search strategies.
If you have seen other interesting technology or think differently, please comment!
Dr David Naylor
david.naylor@humanotics.co.uk