The Impact of AI on Recruitment: The Technologist’s Perspective

Gavin Wilkins

Group IT Director, Harvey Nash PLC

Gavin Wilkins has worked in the IT industry for over 30 years. After training in formally provable software he worked for 13 years as a consultant for PA Consulting, specialising in real time software development for industrial applications in the manufacturing, automotive, and telephony sectors.

Nothing instills fear and confusion in recruiting leaders these days like talk of the imminent tectonic shift represented by artificial intelligence, and its resulting impact on the human workforce. Unfortunately, much of the zeitgeist around AI has been informed by media hype and not by a technologist’s worldview. We were lucky to speak with Gavin Wilkins, Group IT Director at Harvey Nash Group, one of the world’s most celebrated professional recruitment and outsourcing agencies. As a computer scientist, Wilkins cuts through the hype and misconceptions concerning the impact of rapidly proliferating technologies on recruitment as an industry, and clearly defines what agencies need to do in order to remain competitive and relevant in the next ten years.


RIX: The landscape surrounding the recruiting world is changing quickly. What shifts are you most excited about? For instance, Google introduced Google for Jobs, a product taking aim at job boards that will likely make quite an impact on the industry.

Gavin Wilkins (GW): I’m not a recruiter, so I look at it from a different perspective. I see lots of shifts in the business and the markets we operate in, but I think I’m more nervous than excited. They create opportunities for us to evolve – but they are challenges. Very little excites the business per se from a technology perspective. We see some interest around video interviewing  and deep learning but I think mostly what we’re seeing is the business urgently looking for ways to meet some of the challenges and opportunities that are coming to bear with the shifting global landscape. You mentioned Google; they’re actually a threat, but not in the way you might think. We’re seeing some the best recruiters service accounts like Google and Amazon, and then get lucrative offers to work directly for such companies and leave their third-party agencies.


RIX: Do you think recruitment agencies need to be more agile?

GW: Yes, we need to identify where we can really shift the focus to adding real value versus simply delivering a service. I think the increasing view in the market is that we’re inundated with fixed price offerings and a plethora of companies who run services internally. So one of the main challenges is try and identify what extra value we can add to the entire recruitment process to make it worthwhile for clients to use a recruitment company. That applies when dealing with clients as well as with retaining internal staff, something that is becoming more of a challenge.

What makes me particularly excited is thinking about how we can utilize new technologies to address some of these issues. For example, leveraging new technologies to reduce the impact of margin pressures by automating tasks around the volume placement process. We’ve looked at off-shoring and have had mixed success. As a technologist, I’m much more excited about using technology to revolutionize the entire recruiting process to make sure the recruiters are focusing on high-value steps and not drudge.


RIX: There is often confusion as to what encompasses AI in general, and what it means for our industry. How would you define artificial intelligence? How is that different than what many people believe to be AI?

GW: I’m a computer scientist by training so I have a pure view of what AI is. I get very irritated when I hear people talking about AI, as they’re not talking about AI, they’re talking about machine learning. Back in the good old days, we would call it fuzzy logic. That’s not to undermine the importance and capability of machine learning, especially deep machine learning. This is where I see it making a huge difference in recruiting.

I’m also very excited about using bots to transform data from one system to another. The most exciting thing is using machine learning to help in the matching of requirements. I think there’s a real future for machine learning (or AI if you prefer to call it that).

The third part of that (with bots to help move and transform data, and machine learning to take away some of the drudgery) is analytics to give a global view of data. And not just CRM data, but data from CRM systems, from HR databases, from finance databases, from timesheet databases, and more. One of the most important things that technology can deliver to recruiting in the next 2 to 3 years is that critical analytics capability so that we can give managers actionable data to identify good and bad behaviors, help them decipher what works and what doesn’t, generate a real view of their pipelines, and tell them what’s working and what’s not in those pipelines. Most importantly, technology can help us identify what activities are profitable and what are not as the two are not connected.


RIX: In your opinion, what’s the first wave of AI that will impact recruiting?

GW: It’s a huge leap. For example, the use of analytics to be able to look at the whole recruiting picture from the CRM all the way through to payment makes a difference, impacts how staff are managed and trained to become good recruiters. This goes down all the way to asking what types of business work for an agency and what’s profitable. There’s no point in having a really good consultant working on a project that isn’t profitable and even at our company, we don’t have this issue resolved yet. We’re actively combining our databases into a data hub to give ourselves a global view of the end-to-end process.


RIX: What have been the biggest obstacles to rolling out more progressive technologies like AI?

GW: The only challenge I have is resourcing. If I could double my resources, I could do what I wanted to do twice as quickly. To me, this type of activity builds on success. If I was able to go out to a managing director and show him the data as to how far he could drill down in his funnels, and look at the behavior and results of those that have a good funnel, that would draw much more support for investing in progressive technologies. You have to start small and show useful information and then encourage enthusiasm for the entire process. The worst thing you can do in this scenario, in my experience, is say “we’re going to build you an all-singing, all-dancing thing” as you’ll never get there. There are difficulties in encouraging other parts of the company to share their data and open themselves up to be examined. That level of transparency can be daunting. Finance gets very twitchy about providing access to invoice data. But, if you have a client, it doesn’t matter how many roles they’re looking for; if they’re not paying you for 90 days, somebody needs to ask the question – is this business worth our time? Because right now the salesperson is just looking at the hundred open roles and getting excited. But it’s bad business if we’re not getting paid.


RIX: Another major issue on the mind of global recruitment leaders is GDPR, local EU legislation that can have a major impact globally. How do you see global firms like Harvey Nash dealing with this type of local legislation as candidates and clients span across multiple geographies?

GW: GDPR is undoubtedly going to be a challenge. It’s going to be interesting. I think that it matches the DPA very closely when you look at it in many respects. There are some cross-benefits when dealing with American clients. We are in the process (DPO and I) of examining the apparent legislation and understanding what we need to do. We’re building a list of questions as to what needs to be answered hopefully by ICO in due course. We’re working with external companies to advise us.

The overall attitude is that we need to embrace GDPR and become experts on it. If possible, we want to be accredited and show compliance which will help us generate new business around the world – if such an accreditation is introduced.

The trouble is that there are so many GDPR experts coming up all over the place, yet no two of them actually agree with each other. There are many questions that we have and every company has that needs to be addressed by ICO to have the bottom line. For example, will this apply to historical data? This is a simple question but there are big implications. We also need to understand the legal basis of processing which I think is article 6.1c if I remember correctly. We need to understand how local legislation applies in relation to GDPR and know it and put that in our legal basis of processing. That has enormous impact on erasure – do we have a legal responsibility to refuse if someone asks us to erase their PII? The key thing is that the information commissioner’s office needs to get a lot more information about on how these articles will apply in the real world so that we can respond appropriately.

By October, we hope to have a full-fledged project running to take us through until May of next year to prepare the business and technology for GDPR. We have quite a few activities going on now: we’re doing our data searches, looking at classification and what data we have where, and the classification we have supported with software so we can oppose it. For example, if you try to send a CV, your system can tell you not to do that. We have a lot of work to do, but the activity needs to be split to things to do at a technology level and things to do at a business change level.


RIX: Let’s conclude with some final thoughts: what’s the most exciting initiative you aim to accomplish from a technology standpoint in the next 18 months?

GW: Two things that stand out for me are analytics, which I’ve talked about, and the use of bots.

We want to develop bots to fundamentally move data between databases so that we don’t have clunky processes requiring administrators to oversee them. For example, moving a contractor from placement to timesheet processing and then to a managed service and finally to offboarding.

That is allowing us to look at bots and determine how we’re going to make the technology work for us.