Hiring Beyond 2025: Breaking the Algorithm, Beating the Bots, and Building Better Teams

Mihaela Draghici, AJ King at UXDX EMEA. Video: https://youtu.be/0j5L6H4-YDY

Readable transcript: edited from the recording's captions for readability (fillers and false starts removed, punctuation and section headings added). Wording is the speaker's own. Timestamps are positions in the video. Names marked [?] could not be verified against the audio.

AI screening and taking the human out of hiring

[00:00:10] Host: Hello, good morning, good morning. How you doing? Grab a seat, grab a seat. Good morning. How was breakfast? Amazing, amazing. Yeah. The weather is nice, everyone. So, don't say that. We're in a dark room inside. The weather is terrible outside. Here is where you want to be, where you want to be all day. Incredible.

[00:00:31] Host: Okay. We talked a lot about AI yesterday. We're going to focus a little bit on hiring and AI as the starting point for this discussion. Of course, in the audience, if you think of questions as you go, the Q&A should be open throughout. We won't have it popping up in the background, but pop any questions that you'd like in there as we go, and we'll have slightly longer than the usual five minutes for questions from the audience at the end.

[00:00:59] Host: I think the easiest way to start and frame this is: we now live in a world where everyone is experimenting with AI tooling to try and make their lives more efficient. And so we're at a stage now where AI screens a lot of our candidates before a human ever gets to see the person behind the CV. So, I guess to start us off, what are your thoughts on that? What are your thoughts on taking the human out of the beginning of that hiring process?

[00:01:30] AJ: To begin with, I will say, first of all, anything I share today is my personal views and not those of my employer. Just full disclaimer, before I get in trouble. I also mentor a lot of researchers outside of the company. What I hear is that in the market today, especially where there are more candidates than open roles, it's starting to feel a lot more transactional. That results in impersonal rejection emails, or even worse, complete ghosting. So what should be a human-centric experience is feeling a lot more transactional, more focused on the employer that's trying to hire and emphasizing what they're looking for, and not as balanced in terms of the candidate also having the opportunity to evaluate and assess if it's the right fit for them.

[00:02:28] Host: Yeah, absolutely.

[00:02:31] Mihaela: I personally have mixed feelings, actually connecting to this. A lot of the time, candidates feel this compelling need to fit the job description. I've noticed over time that instead of putting out their skills and what they're great at, sometimes they just want to match the job description. We were discussing this yesterday as well. I find it a bit ironic, and I've been thinking about it: candidates looking for a job take the job description text, put it in, I don't know, ChatGPT with a prompt, and then get a revamped CV that perfectly matches it. And then that CV gets sent and gets reviewed by another AI, and then the question is, will it be filtered out or passed through to the next stages?

[00:03:29] Mihaela: So there are these dilemmas and these things to think about, and personally I don't have a clear answer for it, or a suggestion or a recommendation of what we should do. For me it's just a matter of looking at how things are happening right now, trying to understand, and then figuring out how to make the most of it. But again, I think I'm going to become a bit unpopular, I'm going to bring it up again: it still comes back to the people, and the humans being authentic, being genuine, and playing on that a lot more. Standing out, being more human in a process that is more AI.

[00:04:11] Host: Yeah, you said you'd be unpopular. I think that's quite a good message. I think that's quite a positive message to have in this sort of space, talking about AI. I love it.

What early-career candidates bring, and what we look for

[00:04:20] Host: So, AJ, let's talk from a researcher's perspective. What are you seeing that early-career researchers are doing differently, that you think, as researchers who are hiring, you should be paying more attention to?

[00:04:39] AJ: I think it's super refreshing to see the energy they bring to the workforce. There's a lot of curiosity to try new things and do them differently, and challenge legacy processes and methods. They're a lot more open to collaborating. So I would say just that energy and curiosity and trying new things. It's very refreshing.

[00:05:10] Host: Nice. The energy of youth.

[00:05:13] AJ: Yes.

[00:05:13] Host: Yeah. I don't remember it, it's been a long time.

[00:05:15] Mihaela: Neither do I.

[00:05:18] Host: It's a good balance. Amazing. And Mihaela, from an engineering leadership perspective, what are your thoughts on that? Or maybe beyond that, how do you evaluate that mindset, the potential that you can see in someone early career-wise? How do you manage that from an engineering standpoint?

[00:05:37] Mihaela: Just a disclaimer as well: I'm responsible for cross-functional teams, and actually I have more experience in hiring PMs and designers, but also engineers. There are some things that are valid across the three roles. Beyond the expertise and the knowledge in the specific field, one, we're looking for people that are really great team players. And you might roll your eyes and think, oh my God, such a cliche. But nowadays it is essential. In the setup that we have, with people working in pairs and in cross-functional teams, we really need people to collaborate and work well and glue well together.

[00:06:30] Mihaela: We are looking for people that are open to discussions, and we're looking for good communicators: people that are able to articulate their opinions and their ideas, and at the same time listen to their colleagues' or other people's opinions and ideas, and be willing to have discussions, give feedback and find common ground. So it's more about finding the common ground and reaching agreements and alignment, rather than one-upping each other and fighting over who's right.

[00:07:05] Mihaela: And we live in a world where we need to solve really complex problems. So another very important thing is we're looking for problem solvers and critical thinkers, and people that are able to showcase how they think. We have exercises in the interview stages that we actually use to evaluate for that. We give people challenges and see how they solve problems, how they think through that, and how they explain their thought process as well.

[00:07:40] Mihaela: And then another thing I'm looking at a lot, thinking of potential in people at early stages in their career, is the willingness to learn, the willingness to grow, the willingness to develop, because everything is always changing. You don't want people that are happy with where they are. We want people that are willing to grow, to move forward, learn new things and experiment. So mostly these areas.

Signals and exercises in the interview

[00:08:13] Host: I love that. And it's interesting. I wanted to tease out the more practical way that you can assess that. Obviously interviews are very, very short. We have a very small chunk of time with a person who might end up working with us for a very long period of time. You mentioned some of the critical thinking exercises you do. I guess a question to both of you: what kind of signals are you looking for? What kind of exercises do you do in your interviews to work out that someone is that growth-mindset team player, that they are going to be a good fit for the team moving forward?

[00:08:49] Mihaela: I can go first, and I can actually roll on from what I was mentioning earlier. We have different exercises throughout the interview stages, whether for designers or PMs, but also for developers, where yes, we give them problems to solve, we give them challenges. One idea is to look at: okay, how do they understand the problem? Do they ask further questions? Do they try to get more information about the context, so that they can formulate some ideas afterwards, or do they jump straight to solutions?

[00:09:29] Mihaela: And actually, an interesting topic I was discussing with Tali earlier, which is quite relevant nowadays: are they using AI to solve the assignments or not? How do they use it? Are they able to use AI as leverage to help them out? And then are they able to explain clearly how they used it in their assignment, and to their advantage? Because it is important nowadays to look for that as well.

[00:09:58] Mihaela: And then, talking about working in a team and collaboration, we also assess a lot for that, for example for engineers, through pair programming sessions. We look at how they work together, how they explain things while solving a situation, how they get feedback, how they listen to feedback, how they take suggestions into account throughout the process. So these are a couple of things.

[00:10:23] Host: Useful, really good.

[00:10:25] AJ: Yeah, I think along similar lines, for the collaboration, it's focusing on the behavioral questions and understanding how they act in tension. How are they solving problems? How are they giving and receiving feedback? How are they collaborating across disciplines? Situational, specific examples help with that, because they can craft a scenario: this is what happened and this is what I did. It's super helpful to understand how they navigate those things. And then for critical thinking, I think it's: are they asking thoughtful questions to frame the problem and really understand what they are trying to solve for, before they jump into the solution? If they're able to articulate their thought process, rather than this perfect plan, it's always helpful to get a glimpse of how they solve problems.

[00:11:25] Host: It's interesting you say rather than that perfect plan. Do you get skeptical if someone comes in with what's almost too good of an answer? Is it AI generated? Literally an earpiece during the interview?

[00:11:36] Mihaela: I think, again connected to the discussion I had earlier this morning, it's difficult for us, and we're still learning how to spot it. But with experience and over time you get it, when you have the experience in a specific field, like UX research and overall product design or product management. In a way, you notice whether it's a genuine response or not, and in the conversation you start figuring things out. And if you have a doubt, as an interviewer for example, you start probing, and once you do that, you realize whether the answers stand or not.

Making sure the tools don't overlook strong candidates

[00:12:23] Host: That's fair. Let's move on a little bit into some of the early stages of that process. I appreciate you are both hiring managers. You're not really involved in the decision of what the practice or the process is for hiring within your company at that fundamental level. But how are you ensuring that the AI or the hiring tools you're using aren't overlooking strong candidates, aren't biased against certain things? How do you factor that into your hiring process?

[00:12:59] AJ: I can take that quickly. Honestly, we use third-party tools, so we don't really have that much visibility into what happens in the third-party tool. But I think something that candidates can do to stand out in the application is doing that personal and proactive outreach to the hiring manager or recruiter, especially in the initial phase, when the role just gets posted and you're in the first wave. And then trying to learn and genuinely understand what the role is about, what's the expectation, what it takes to be successful. Taking that extra step could help beat the bot, essentially.

[00:13:45] Host: More of the personal touch up front, right?

[00:13:47] AJ: Yes, exactly.

[00:13:50] Mihaela: Again, personally I don't have that much experience with it, but I'm just thinking about what I would consider. It makes total sense to know the models clearly, and the training data, and have transparency over what training data it's being fed, and have clarity over that. And also be able to have the AI check against itself, against biases, and look at, I don't know, diversity analytics. So take the rejected candidates and compare them with the accepted candidates, and see if you have failures in patterns, or bias on one side or the other. And I think that's the most important. Over time it's easy now with this to spot bias, and it's important to look at that. But I think that diversity analytics side of things is essential.

[00:14:51] Host: Yeah, it's really important to look at everything that comes in and then almost do a data review on a fairly regular basis, to make sure that the things that are meant to be driving efficiency are driving efficiency, rather than weeding out the wrong people or ending in a difficult situation. I think it's a really tricky balance. Obviously the personal touch, I think, is really important. The benefits of systems that make it easy to apply for jobs are obviously really good from a candidate perspective, in a market where you're applying for lots of jobs.

[00:15:23] Host: From your personal opinions: things like Easy Apply on LinkedIn. You can click a button and apply for a job in 30 seconds; you can apply for ten jobs in ten minutes. That very much flies against the suggestion of the personal touch that you've put out. But for a candidate who is trying to find a job in a saturated market, what are your thoughts on Easy Apply?

[00:15:49] Mihaela: I think volume. When I think of Easy Apply, it's just numbers, and I'm not sure that necessarily means quality. And then it's about having criteria in place and ways to filter for the quality.

[00:16:14] Host: Nice.

[00:16:16] AJ: Yeah, I think similar. It's just a lot of noise that comes through with Easy Apply, and the systems might not necessarily be designed to handle that volume and nuance. So that underlines the importance of personal touch and networking, and staying in touch with people in the companies that you hope to work for one day, and building those relationships over time. Not at the last minute, when you need it, sending a message like, can you refer me to the role, when you haven't worked together in a very long time. So just nurturing those relationships over time, so when the opportunity does come up, you have that established relationship.

[00:17:01] Host: Yeah, that makes sense, and it's important. Although I'm a people pleaser, so if you want to be referred for a job with me, I will, even if we've met over coffee and only spoken for five minutes.

Helping recruiters understand what you're looking for

[00:17:10] Host: Let's chat a little bit about further into the process. Obviously it hits a point where recruiters in our company are also doing a sift through what the AI has probably chucked out for us. We're hiring for very specialist roles an awful lot of the time. How do the pair of you help the recruiters, who might not be embedded in design functions and product functions and engineering functions, understand what it is you're looking for, so you don't lose people at that stage?

[00:17:41] Mihaela: I can take that first. I have an easy one, and I spoke about pairing a lot yesterday. We actually pair with recruiters as well, throughout the different stages in the interview process. Just to give an example, in recruiting designers: through the portfolio reviews, through the first interview stages, through the assignment, the practical exercises. That helps the recruiters, over time, and I would say a rather short period of time, understand what we're looking for and the kind of questions we're asking, better than them getting a brief from us. Having a list of skills required and criteria and questions we ask, that's useful, that's helpful. But I feel that having this interaction, being part of those processes and pairing together with the expert responsible for hiring for a specific role, adds a lot more value and helps recruiters get a better understanding.

[00:18:54] AJ: What worked for me, I think, first and foremost, is steering away from using discipline jargon with recruiters. Like, strong mixed methods experience. Well, what does that mean for the recruiter? So just simplifying that language and saying, hey, the candidate should be comfortable with using different research methods, and being able to influence the direction of a design or product with the insights. Really simplifying that language, so they're not having to learn your craft to be able to help recruit. And partnering with them for maybe the first few candidate reviews, so you're giving feedback and pointing out specific things that you're looking for, and having that clear job description and criteria. Just helping coach them, essentially.

[00:19:46] Host: Coach them, lose the jargon. Mixed methods definitely sounds like something a cocktail maker does.

[00:19:52] Mihaela: I would like to add a reply to this one, actually, regarding the jargon, because it's something that I have very strong personal opinions on. Yes, to some extent. However, I still believe that it is important for recruiters to understand part of the jargon, and the meaning of certain frameworks or approaches and so on, because they come up in discussions. When they're filtering candidates, screening candidates in screening calls, they need to be able to understand to some extent what the candidate is talking about. I do believe that it is still important to grasp that context. Yes, they will never be experts in UX research, but it is good for them to understand the context of that world.

[00:20:43] Host: I suppose: take out some of the jargon from how we talk to them about things, but prepare them for the fact that a candidate is going to use that terminology within the space.

[00:20:52] Mihaela: Yeah.

[00:20:54] Host: Nice, I like that. That's a nice balance to it. I thought I was going to have to move my chair between you. No fighting, no fighting. Okay, cool, that's good. Hopefully there are a lot of questions, so the Q&A doesn't just become a punch-up.

Onboarding and supporting junior hires

[00:21:06] Host: Fantastic. I'm really keen to get audience questions on this, so although we've technically still got a little bit of time on our discussion, I'd like to move on to Q&A quickly. So, very briefly, in 20 or 30 seconds: once you've hired someone. Let's talk about the junior level, because I think currently, with the rise of AI doing lots more of the smaller or easier efficiency jobs, breaking in at a junior level is more difficult. What are the pair of you doing to mentor and to coach and to help junior people when they dive into crazy high-paced, high-delivery environments, to get them up to the sort of skill level you're at?

[00:21:48] AJ: I would say at that entry level, researchers need a little bit more guidance and structure. So I like to prepare a very detailed onboarding document that includes their 30, 60, 90-day milestones. What do they need to accomplish, when do they need to deliver their first research report, or if there are specific stakeholders they need to meet. It's outlined, so it helps them get on the right track right off the start. Another thing that's worked is pairing them with an onboarding buddy or mentor, maybe more of a senior researcher or lead researcher, who can be their support system. Because they might not necessarily feel comfortable asking you specific questions as their new manager, but they would feel more comfortable having those discussions with their peers. So those are a couple of things that have worked.

[00:22:49] Host: I love that, and I love the important distinction of having a strong senior mentor that is not your manager. I think it really helps, especially in those early days. I subscribe to that. Plus one. See, back on track. Anything else, Mihaela, that you do differently?

[00:23:08] Mihaela: Similarly, having an onboarding structure with clear milestones. And for us it's essential to have a pair, a more senior member of the team that also has the willingness and openness to teach and mentor and work alongside. And the other thing that I thought about, and I saw in different contexts, was giving the new joiners, especially if they're juniors, access to people around them. Not only part of their immediate team, but outside their team, people they could learn from or reach out to for help. So making them comfortable in knowing that they could go to other people outside of their immediate bubble to ask for help and to learn more, and get more comfortable in the role, and more comfortable with what they already know and what they are willing to learn.

[00:24:16] Host: Amazing. So I guess, as a full summary, what I've heard from you both today is: at that very early stage, candidates, anyone in the room, anyone online who is looking in this current market, AI sifting is going to happen to you. So try to put in that more personal touch. It can be awkward and weird to reach out to hiring managers and recruiters, but it's worth it. It is worth that extra personal touch off the bat. Hiring managers in the room: if you're not pairing with your recruiters and getting them upskilled and coached in the terminology and what you're looking for, then that is going to be a break in your process.

[00:24:56] Host: And then finally, making sure that at the onboarding stage there's a really strong culture of support, which I think you're suggesting, around new hires. Actually, at whatever level someone comes into a business, you can be an expert in your craft, but you are a new starter and usually a first-timer in a new company. Every company is slightly different. So, wonderful, thank you so much for the conversation.

Q&A

[00:25:23] Host: I think we'll now move on to some Q&A from the audience. Let's see. Oh no, I don't have to press this. I won't press this, I'll put this back down. Amazing. Oh, I love this. Lots of upvotes for what I can already see, even though I've only read the first bit, might be quite a spicy question. "I believe interview exercises are unfair to candidates typically. Free work for the company. People are under enough pressure without the hassle of spending time on these things." So I'll dissect some of the spiciness from that. I'm going to assume that you're not just using interview candidates to do your job. But the exercises themselves: how do you ensure that the exercises that you present in interviews do show off the skills you're looking for? What are your thoughts on this as an unfair approach?

[00:26:13] Mihaela: I can answer very quickly. I know from discussions with other friends that work in large companies, enterprise-level companies, that there are a lot of confidentiality issues in giving people who are not within the company access to company code and company data. So in a way, that's already out of the question. So it's about having exercises or challenges, small challenges that replicate work scenarios, but are not actually real life, not working on real products and coming up with real solutions or real code. They're just replicating work scenarios. And finding a balance: giving an assignment, for example, that wouldn't take a month of a person's time or a week of a person's time. Trying to find a balance of, okay, you can invest some time in this to be able to showcase your skills, your knowledge, your approach to solving a problem. It's about that, really.

[00:27:25] AJ: I would say more recently I've eliminated that from my hiring process.

[00:27:35] Host: The exercises? Or the unfairness in your exercises?

[00:27:36] AJ: The exercises. But even when I've done them, I try to make them super irrelevant to the company. It's like a random pet shop, totally unrelated to the company that I was at at the time. Because the importance is more how they're thinking and planning the research. How are they collaborating? Are they asking questions as they work through the exercise? And being able to manage that perception that we're getting free labor, because that's not the goal. So those are the two approaches.

[00:28:08] Host: No, that's really good. In the last hiring process that I did, I got a bit of pushback, because I wanted to change the process we were going to do. I didn't think the process that was laid out as standard was actually going to tell us what we needed to know to differentiate candidates. In the end, it ended up being a heavier time cost for me as the hiring manager, but I did an on-the-fly exercise. Candidates came in, I gave an exercise and said, we're now going to do, essentially, a simulated research scoping session. Here is the challenge. And I deliberately left out all the important details that you have to probe for, and then assessed how well they probed through. But again, it was an irrelevant experience. So on that, how do you feel about in-the-moment exercises? What are the benefits of those? What are the downfalls of in-the-moment exercises?

[00:29:06] Mihaela: I think they allow space for creativity. I think they work if you are a more experienced hiring manager, or an experienced person in your field, because then you're able to gauge certain things on the spot. But again, from a company perspective, I don't know whether that would ensure that you have covered all the criteria, all the areas that you need to check for in order to do a fair evaluation. Because you're doing that exercise with one candidate, versus doing something different with the other candidates. So then how do you ensure that balance of fairness in the evaluation?

[00:29:53] Host: Yeah, the starting point's the same, but it becomes a very different conversation depending on where people take it. That's fair. Any thoughts?

[00:30:02] AJ: I think I... no, that's okay.

[00:30:04] Host: Yeah, that's good. From my side, as you can all probably tell in the room, I'm fairly happy not having a plan and winging it, and so for me those in-the-moment exercises don't stress me. I can imagine, for people who like to be more methodical and be slightly more prepared for things, and I actually hadn't taken that into consideration in my last hiring process, the stress of not knowing what you're going to have to do in the moment might be a little overwhelming.

[00:30:28] AJ: Yeah, I think just reviewing their portfolio, and giving the right instructions to prepare the content that's relevant for the role, is more fair. They had the time to prepare, and in reality they would have the time to plan the project. So just speaking to that, and articulating the impact they made through their portfolio, is more effective than some random exercise.

[00:30:50] Host: Well, I've hijacked this for a question I was interested in. Sorry, everyone. We'll go back to the Q&A. I'm not very good at my job. I love this one. I don't think your name is fair, "Lazy Designer." I think work-smarter-not-harder designer. Okay, and I'm going to try and deliver this in the way it was meant: "Do cover letters really matter?" Really? Do they really matter? I think the question answers itself.

[00:31:13] Mihaela: What's a cover letter? No, I'm just kidding.

[00:31:19] Host: Fair. There you go, they don't really matter. Actually, when was the last time you guys applied for a job? Did you send a cover letter in?

[00:31:28] AJ: I don't think I did.

[00:31:28] Mihaela: No. I don't remember the last time I applied for a job. But I was trying to think whether I'm checking cover letters or not. Sometimes I am, but they're not a necessary requirement. If they're not there, they're not missed.

[00:31:47] AJ: Yeah. My advice would be to just incorporate it into the resume. If there are additional things you'd like to highlight, a skill or passion or enthusiasm, just include it in the resume.

[00:31:58] Host: Amazing. And there are fewer documents to check for the AI.

[00:32:02] Mihaela: Yeah, that is fair.

[00:32:05] Host: I believe my timer has just clicked onto a new two minutes, which I believe is my time to awkwardly stand and ask for feedback. But before I do, I do quite like the top question, so in one sentence: AI hiring processes being sprung on people. This person says, "I would like to know up front if a human sees my application or a bot." Do you think we should be putting the sort of Spotify "this music was made by AI" disclaimer on our hiring processes? Yes or no? That's an interesting question, and I only want a yes or no from you. Should we be clarifying up front?

[00:32:36] AJ: Yes.

[00:32:36] Mihaela: Yes. Yeah, I think that'd be fair.

[00:32:39] Host: Wonderful. Agreement.

[00:32:42] Mihaela: We live in transparency, right?

[00:32:43] Host: Yes, I love it. Thank you so much for the discussion. I hope this was entertaining. We sprung this on you very last minute, so thank you so much for all of the insight you gave. Thank you, everyone.