Your Reference Check May Be the Candidate Themselves

James Lord:

I've heard so many stories of recently as well, where people are giving job offers out to a candidate and it's this different candidate teams up to every stage of the interview. And then a laptop gets sent to somewhere, like, that doesn't even exist. And then someone comes and picks up the laptop and then before you know it, you've just given them the keys to the kingdom.

Yes, it's very important on the application level. But don't be fooled that just because they've passed that and you start interviewing these people that you actually know they are who they say they are.

Steve Smith:

Hey everyone, welcome back to another episode of Work Tech Weekly. I'm Steve Smith, Managing Director of Growth at Rep Cap.

Hiring has a trust problem. Candidates use AI to fabricate credentials, and companies use AI to screen them. In the space of a year or two, fraud went from barely registering to top of mind for talent leaders.

One of the last checkpoints before an offer is the reference check. It's also the part of the process people take least seriously.

My guest is James Lord, who leads expansion in the Americas for Refapp. Refapp is a Swedish platform that runs evidence-based, role-specific reference checks right inside the applicant tracking system.

James has spent more than a decade in reference checking. In that time, he's watched the share of suspicious references nearly triple.

In this conversation, we talk about what a reference check tells you that a background check never will. We get into candidates posing as their own references, and hires who make it through every interview as someone else entirely.

We also talk about where AI belongs in the process, and where a human still has to make the call.

If reference checks are still a formality in your hiring process, this one is worth your time.

Let's get into it.

James Lord, welcome to the podcast.

James Lord:

Thank you. Thank you for having me.

Steve Smith:

What do you think people are getting wrong in the conversation right now? There's no shortage of people with opinions, but where do you think that they're missing the mark?

James Lord:

I think everyone's trying to engineer everything just for the sake of engineering it, you know, chucking AI at everything and anything. Just because it sounds cool and you can kind of gloat to your other friends in talent acquisition, but I think, you know, there's that age old saying, right, if you can eliminate something, eliminate it, and then if you have to keep it, automate it.

But you shouldn't just be automating things, you know, just for the purposes of automating them. You should always be looking, okay, what are we actually trying to solve here? Let's just start from a first principles perspective. What's the biggest problem that we're actually trying to solve?

Rather than just having this kind of shiny object syndrome, which everyone's got right now, because there's so many HR tech vendors that are out there that are waving these kind of amazing little pieces of technology. And yeah, quite honestly I think a lot of talent leaders are distracted by that.

They're having their head turned consistently and then before they know it, they're in various subscription plans and all the rest of it. I think, look, recruiting and talent acquisition isn't actually that difficult, is it? Let's be real. We're not putting people on Mars here.

We're not doing like a SpaceX job. I think the reality is we've made it harder than it needs to be. And I think a lot with new technologies that have come out specifically, you know, AI. I think people are just trying to make themselves look smart.

Steve Smith:

Yeah, totally agree. Well, you know, you're an interesting part of the talent acquisition pipeline, reference checking. So you've been doing this for over a decade, which is a long time for anyone to be evangelizing the part of the hiring process that people find pretty perfunctory.

What keeps you going in it?

James Lord:

Yeah, good question. So, I had obviously to address the accent that follows me into North America. I had the opportunity to move to Canada a few years ago with a reference checking business. And I think, you know, that opened a lot of doors for me personally.

Right. You know, that was a lot of opportunity. And then I have been very fortunate to join a business who I believe are the best in the industry. RefApp, you know, that those guys come from an evidence-based and role-specific foundation. So rather than just kind of like guessing the best 10 questions to ask, everything is evidence-based.

And the journey that those guys are on and the business that they've built over in the Nordics is super impressive. So it made complete sense for me to join them and to help them expand their business here in North America. So, yeah, I've been doing it a long time, but I know the space pretty well.

I know the vendors well and of course every now and then you do come across some horror stories of people, clients would share with me about what's been said on a reference and the bullets that they've dodged, so to speak. So, yeah, look, it's fun. I think reference checks are an important but oftentimes overlooked part of the process.

You know, I do surveys on LinkedIn. I look at a lot of data consistently, and it seems like around about six or seven companies, out of every 10, or 60, 70% of all organizations still require some form of reference checks. But when you start to get under the hood and you look at how they're doing it, it's either being done with a background checking company, and that's not their forte, of course, that's not what they're specialists in.

Or they're making telephone calls, which is just like, well, it's 2026, so there's no validity. There's no verification. There's no authenticity, there's no audit trail. There's a he said, she said, and then you get into defamation issues and all that good stuff. So it is a minefield, so you may as well just use a company that just makes it as a button in your applicant tracking system.

Steve Smith:

Well, let's make sure that we're clarifying terms for our audience. You know, what is a reference check and what is it not, and how is it different than a background check? Right?

James Lord:

Yeah, no, that's a good question. So look, the way that I categorize that is that a background check is factual information, right? It's pass fail, true false, or yes, no. I either have a criminal record or I do not. Simple as that. References is really where the story comes to light. So, okay. James is not a criminal, but that doesn't mean to say that James is gonna be a good employee.

Steve Smith:

Right.

James Lord:

James like under this particular circumstance? How does he deal under pressure? What's his communication style like? Are there any counterproductive work behaviors that we might not know about? You know, James himself, as the candidate, would never, of course, highlight that in an interview. And similarly, it's never gonna come up on a resume.

But the people who know the candidates better than sometimes the candidates know themselves, they can shed light. Again, if you lead them, if you ask pertinent and role specific questions, you can lead them and you can open up that conversation and you can learn a ton. Now, that is not necessarily going to change the hiring outcome.

I've got hundreds of examples of where references do change the hiring outcome, i.e., a job offer is rescinded as a result of what's been learned. But more often than not, it's an opportunity to improve an onboarding process for a candidate. You know, you don't know, right? A candidate could be on the autism spectrum. They might deal with things, you know, they might have a different interpretation of something versus another candidate.

And if we don't ask these questions that are really specific to talk about observed behaviors, how are we as the new employer ever gonna know how to onboard them the best way possible? And then beyond onboarding I often like to think about reference checks almost as a SWOT analysis, you know, strengths, weaknesses, opportunities, and threats. So that leads very nicely into L and D, learning and development.

Right, because James could be good at this category, but not so great at this stuff, and we really kind of need to level him up so we can then start to plan programs and training and that's really valuable because, you know, it's like most things, Steve, you know, whether you buy a house, you buy a car, you know, you start dating someone new. You don't really know what you're getting until you're three, six months down the street, do you?

Right. And this references is kind of almost like the benefit of hindsight, someone else's lived experience. So, personally, I actually don't like the term references or reference checks because I think it's a very outdated term. And I think it also has a lot of negative connotations of how people have, you know, their own lived experience of doing them and not really receiving much.

But what I do think it is it gives us a great opportunity to look into a person and, you know, all of the facets. If you think about a person who's like a diamond, right? All of the different facets, the different sides of the person. That's really valuable. And where else in the process have we got that from? From independent third parties that have no stake in whether you hire this person or.

Steve Smith:

Well, I think that maybe it would be useful then to also talk a little bit about RefApp, because you know, as we said, reference check, this reputation for being performative. It's a checkbox. Everyone says nice things, but nothing useful actually surfaces, obviously. It's just like a company coming out of the Nordics.

Design is very important and everything needs to have a purpose. What is RefApp doing differently to make this a useful part of the talent acquisition process?

James Lord:

We live where people spend most of their time, which is the applicant tracking system or the HRIS. I think that's such an important thing. You know, the tab graveyard is very real. People, especially talent leaders right now, they've got tools coming out at them from all over the place and, you know, everyone's forgotten their password, so you end up having to reset and all that stuff.

So when you've got a fantastic integration into your Ashby, Greenhouse, Teamtailor, Workday, Workable, you know, pick your favorite ATS, we're pretty much integrated with all of the major ones. RefApp sits as a tab or a stage change inside that ATS. So while our own platform, Refapp.com, is of course very pretty, it's user friendly. You know, you can trigger a reference request on a candidate within 10 seconds. And 89% of our references come back within 24 hours, which is a stat that I'm very proud of, of course.

But most people don't do it inside Refapp.com, most people live inside, you know, Fountain, Greenhouse, Ashby, Lever, whatever. That's where they live and we should give them the complete functionality of our tool inside that native environment where they actually live, which is exactly what we do. So that includes notifications, things like, you know, fraud alerts, because we know that statistically we are looking at almost 17% right now of all references that go through the RefApp platform are unusual in nature.

So that could mean that the candidate is sat on, you know, the same device, the same browser, geolocation, same wifi. The candidate is the reference, but they're sat in their own basement or they're in their backyard on a Gmail address. That's steve@gmail.com. I then go onto LinkedIn, I look at Steve Smith and I think, oh, that must be Steve. But actually it's the candidate, you know, acting as themselves. So, that happens a lot. And we have all of those notifications inside the app.

Steve Smith:

You referenced sort of the fraud component. And it seems like that really has moved to the forefront of the talent acquisition conversation because with a world where we have AI generated pretty much everything it feels like fraud went from really not on anybody's radar to be completely top of mind. How has your job changed over the past year or two as this segment of the market, or just concern around fraud, has just exploded?

James Lord:

I spend a lot more of my time explaining to my clients that, you know, data doesn't lie. Data is binary, right? It's black and white. It's true false, yes, no. The whole internet is built on zeros and ones, you either are or you're not. I think, you know, I spend a lot of my time explaining to people that if there are these anomalies through this digital behavior tracking that we are monitoring from candidates and their respective references, the chances are that the candidate is trying to mislead you.

Or they're trying to exaggerate or embellish their skills, experience, qualifications, and I think that is sometimes, especially because people don't necessarily think that's going to happen to them when references are done. Sure it happens at top of funnel on an application level, but if you think you've gone through the stages, might have had one or two interviews. You might have even offered a role, right?

The job offer is typically subject to satisfactory references and a background check. That's the point in which RefApp comes in. Usually. Now we are used at all different stages of the funnel, but we're in the final hundred meters, right? And when we find these anomalies, I have to almost kind of reassure my clients that the technology doesn't lie, and it's maybe the candidate is just from a desperation perspective.

It could be the candidate may have been unemployed for quite, you know, some time, but this is the last, you know, references are, you don't want to start any relationship based on mistrust. Right. And if I'm going to mislead you and say, Hey, I've got all these skills, Steve, yes, I've got this skills, I've got this experience, these are these qualifications. I'm amazing at X, Y, and Z. And then I turn up on day one and I can't do the job. I mean, that's not serving anyone.

So I think it's really important that we do that thorough due diligence and even when we think we've got an amazing candidate, because we've interviewed them a couple of times, or a couple of our hiring managers have, we can still be misled. And that is happening more and more. Like I say, that stat has gone from when I started in this space, it was around about 6%, five or 6% was the typical number of all references that were suspicious. I don't like to use the word fraudulent 'cause there are stories.

Sometimes you and I could bump into each other in a Starbucks, we're on the same wifi, whatever. Rare. But I have heard of that. But that's now almost tripled. We're now looking at like 17% of all of our references, and to put that into context, we're doing hundreds of thousands of reference checks every month and 17% is the consistent number of people that are, you know, on the same IP. The digital behavior doesn't lie.

Steve Smith:

Well, you know, we talked a little bit about, you know, the AI elephant in the room, and one of the big impacts on hiring process, which is really the trust issue. You know, AI is, you know, been weaponized on both sides of the table. You know, candidates are using it to fabricate credentials and all manner of things, but companies are also using it to screen. Sometimes

James Lord:

Yeah,

Steve Smith:

that are just fraudulent companies, you know, where does a reference check fit into this world? And how can it be a trust builder and a credibility builder?

James Lord:

Human beings. Right now, quote me on this, 'cause in five years time it could be very different. Right now human beings hire other human beings and we all value other people's opinions from their lived experience that observed behavior that they've had. Right. So I think there is so much validity in that, I think I've almost seen it kind of gone full circle because there's so many screens and there's so much kind of automation out there.

What I really just want to get back to is shaking hands, having coffee, meeting people as we now abbreviate IRL, with just meeting people in real life and getting to know them and references enable us to do that, connect with other human beings and just have, whether you want to have a structured conversation or an off the record chat, knowing that you've actually validated that person is who they say they are.

This enables you to do that. I mean, we use AI in our processes, Steve, obviously, you know, we have an R and D team who are creating these questions that are best practice, but we also have elements of AI helping us create questions. So if you are in, for example, the Teamtailor applicant tracking system, you can generate a questionnaire using RefApp's data, and our AI to generate the best 10, 15, 20 questions to ask for that particular job role. So that's awesome.

You can also summarize reports, so when you get this information back, you could have three different reference providers who have all completed, you know, 150 to 500 words each on this candidate. I just want to summarize that. So we would use AI there, but you are a hundred percent right. It's been weaponized on both sides, but I think kind of goes back to the beginning point that I made, I think, what's the problem that you're trying to solve? Let's just start with that. References are connecting human beings. And that's the most important thing.

Humans are human.

Steve Smith:

And, you know, when you think about, you know, RefApp, you know, using, you know, not only AI to structure the conversation, but then analyze the responses, where's the line between AI's helping us make sense of the data and AI is replacing human judgment that references are supposed to provide?

James Lord:

Yeah, really good question. I think it is an option that we have and we've done it through the market. The market's asked us to say like, Hey, we want the option to summarize these results. I think we as a society right now have to be very careful that we are not almost becoming redundant in our thought. You know, why we are not using our brains? I speak to people all the time and they don't really want to think anymore or try and even remember things.

They just simply ask ChatGPT or Grok or Claude or whatever, you know, pick your favorite AI. And I think that's a dangerous place to be specifically in hiring when we're making judgments on other people, other human beings. They always say there's an art and there's a science in talent or in talent acquisition, and references, or this technology, is not the art. We are the science and we're not trying to replace the art, you know, we are gonna give you all of the information.

We can lead and, based on examples, say, Hey, this is what we've seen in the past, benchmark data, all that good stuff and insights. But ultimately it should be a human being that's still making that decision to move forward based on the available information. And I think that's a really important distinction to make art and science.

We are the science, we're not the art. We're not replacing a human being. And I don't think anyone ever should. And I think there's some interesting cases out there right now where some organizations are kind of leading people to that, like the thumbs up or the thumbs down. And I think that's a dangerous place to be.

'Cause very quickly a recruiter could be replaced by an agent.

Steve Smith:

Well, I mean, it's sort of like in the conversation around using AI in the recruiting process. It's just like there's a pretty solid consensus of people. Like you cannot have your AI be a black box. You know, it needs to be auditable. You need to understand what's going into that process.

But you look at reference checks, you know, companies that aren't automating it, it's just like, that's exactly what that is. It's a black box, and as you

James Lord:

You've got no idea.

Steve Smith:

like,

James Lord:

Yeah,

Steve Smith:

the hell knows what's going on in there?

James Lord:

Well look, this is the thing, right? Candidates, you know, if we think about data, you know, the language of GDPR, which is like really the gold standard, global gold standard of data protection, but very closely followed by CCPA, California Consumer Privacy Act. Right. If we look at data in that instance of data controller and data processor, the candidate has the right to request access to their information at any time, right? Because that data is about them.

You know what is really important is you can back up any claim. So if you decide to not move forward with a candidate, or you do decide to move forward with a candidate, but you pick that individual over another, well, where's your logic behind that? Right. And I think it's really important that we've got that audit trail there, because if you're just like, oh yeah, I just feel I'm just getting a good vibe, you know, good vibes like that can be dangerous, you know? Yes, it's important to trust our gut and to go with the art side, but we need to be able to back this up with data and fact and validity.

Because I've heard so many stories of recently as well, where people are giving job offers out to a candidate and it's this different candidate teams up to every stage of the interview. And then a laptop gets sent to somewhere, like, that doesn't even exist. Some address that's actually like a PO box and then someone comes and picks up the laptop and then before you know it, you've just given them the keys to the kingdom 'cause they've got access to our systems.

And you know, like how you've got to check, you know, do these checks and balances throughout the whole process. Yes, it's very important on the application level. But don't be fooled that just because they've passed that and you start interviewing these people that you actually know they are who they say they are.

And that's why I think, you know, for us that the validity piece has become even more important. You know, misrepresentation. And there's desperation in there as well. People will always lie, but that's something different. They can still be the actual right person, but they just might misrepresent their skills, experience, qualifications, 'cause they're just so desperate to get a job.

Steve Smith:

So let's think about the hiring process in 2030. If it's more automated, if it's more AI screened, if it's more data driven, the case for making it more human sourced, more relationship based signal. Like a reference check. What's the case for that still mattering?

James Lord:

I mean, I have to ask you the clarifying question there. Are we being managed by humans or are we being managed by AI? Tesla robots.

Steve Smith:

That's an excellent question. I mean, so, okay, let's go there. So, let's look at if the workforce of 2030 is a mix of humans and agents, how does the reference check change? Are we checking references on agents?

James Lord:

We could be. That's an interesting kind of thought exercise, right? And we're already starting to see that. We're already starting to see a mixture of employees and agents working together, you know, and they're reporting up into one other person and or agent. I think that it's only actually going to become more and more important. That's the way I see it, because humans that are actually still left employed, the ones that have embraced AI and have, excuse me, continue to remain relevant ultimately in an organization and can justify their W2 paycheck, right?

Those individuals, that's even more important that we hire the right and we keep the right ones. So how do we do that? Well, we ask pertinent role specific questions and understand what their motivations are. You know, going back to the SWOT analysis, the strengths, weaknesses, opportunities and threats of this person. Honestly, I don't know how we would reference check an agent or we haven't done it yet, but I'm sure that will happen at some point.

But I think certainly from the human perspective, it becomes ever more important because we don't wanna hire a dud. 'Cause what's the point in hiring someone when I can just get three agents and they can do twice the output of this human being, but if this human being is client facing, they're gonna be the face or the voice of my brand. That's really important for me to do my due diligence there. So that's, I think, the best case scenario for references and that's where I believe that they're going and I think they're becoming more and more important.

I think there is a world where we start to move towards a more kind of almost like a resume or a bank of references that follow us around almost like a blockchain type scenario. I've seen, you know, we are working on some intelligent stuff in that space as well. But where you kind of just unlock all your references and everything just comes down, everything's on a ledger, on an audit trail that's already there. The challenge, of course, is that same argument has been made about the resume for how many years? I wanna say 50 at this point, where someone's trying to standardize the resume and there's no such thing as a standardized resume.

We all thought when LinkedIn came out, it was gonna be the death of recruitment agencies. Well, staffing agencies, it hasn't, right? It's just another avenue, it's another facet. And I think it becomes more important. It's just another channel. You know, people still watch cable TV, even though we've got Netflix and whatever other streaming services out there, Amazon, all the rest of it. We've just got more optionality now and I think that's going to only continue.

So any sources where we can really tap into who that person really is, what they've done. What's their track record? Is this guy a dick? Is he gonna be wholly inappropriate at the holiday party? Like, tell me what I need to know before I pull the hiring trigger. And I hire this guy, right?

Steve Smith:

Well, this has been an interesting conversation. It's certainly not where I thought a conversation about reference checking would be going. And I think that you make a really compelling case that, you know, checking references is in a lot of ways more critical than ever, and there's a lot of ways to do it better than the conventional wisdom dictates. So James, thank you so much for sharing your time and perspective with us, and hopefully we'll get a chance to do it again sometime.

James Lord:

Love to. Thank you so much.

Steve Smith:

As hiring gets more automated, the human parts of the process are getting more valuable.

AI can fabricate a candidate's credentials. It can even help someone pass every interview as a person they're not. When that happens, the opinion of someone who actually worked with the candidate carries more weight. People still hire people. And the people who know a candidate best rarely get a chance to weigh in.

A reference is a signal that only counts if you can trust it. An off-the-record phone call feels personal. It’s also completely invisible in the hiring record..

A good gut feeling about a candidate is worth listening to. But it's hard to defend when someone challenges a hiring decision.

The same logic applies to the tools. Summaries and benchmarks can help you make sense of what references say. The decision still belongs to a person. Hand it over to a thumbs up or a thumbs down, and the recruiter becomes the next thing to automate.

If the final step before an offer is the least verified step in your process, that's the step to fix first.

If you enjoyed this episode, make sure to subscribe to Work Tech Weekly on Apple Podcasts, Spotify, or YouTube. And I'll see you next time.