DE Talk | The End of “Jobs Near Me”: Conversational AI Is Changing How Candidates Find Jobs

“Jobs near me” has been the most common job search for 20 years, and with the introduction of AI, it’s on the verge of disappearing. Candidates have stopped typing keywords. They’re asking ChatGPT, Gemini and Claude for jobs that fit their experience, their schedules and their lives. If your jobs aren’t in the handful of answers those tools give, candidates never see you. And because the AI platforms share no search data, most employers have no idea what candidates are asking. In this episode of the DE Talk Podcast, host AJ Selvey of DirectEmployers Association sits down with Ben Russell, co-founder of SonicJobs, to talk about what’s changed, the misconceptions costing employers good candidates, and how CareerGPT brings a ChatGPT-style conversation to your existing career site without replacing your website or ATS. Ben also shares early CareerGPT data showing that more than half of candidate conversations lead to an application start.

About DE Talk

The DE Talk monthly podcast features honest and open dialogue between powerhouse industry experts on a variety of HR topics ranging from OFCCP compliance advice to emerging recruitment marketing trends, equal employment opportunity initiatives, and insightful solutions that help infuse new life into your HR strategies.

Episode Interviewer

Episode Guest

Transcript

DirectEmployers Association (00:00):

Get ready. The DE Talk Podcast starts now. Insightful conversations and dialogue helping you put the human factor back in HR.

AJ Selvey (00:08):

Welcome to the DE Talk Podcast. I’m your host, AJ Selvey, DirectEmployers Vice President of Membership Development. And today we’re talking about something that sits the very beginning of the recruitment journey, how candidates actually find jobs. The way people search for jobs has changed dramatically and new technologies, including AI, are creating even more ways for candidates to discover opportunities. Joining us to explore this evolution is Ben Russell from SonicJobs. SonicJobs works in the job search space, helping connect candidates with employment opportunities and helping employers reach job seekers through today’s evolving search landscape. Let’s get started and welcome Ben.

Ben Russell (00:47):

Thanks AJ. Great to be here. Thanks so much for having me on the show.

AJ Selvey (00:51):

Absolutely. Fantastic that you’re able to join. And before we really jump into the broader conversation, how about you tell us a little bit about yourself and your role there at SonicJobs?

Ben Russell (01:01):

As you mentioned, my name is Ben Russell. I’m one of the co-founders at SonicJobs, and I spend most of my time really thinking about how we can position emerging technology to really improve the hiring experience for both candidates and employers. So a big focus for me right now is the launch of our new CareerGPT product, which brings that conversational experience people already use in ChatGPT onto an employer’s own career site without asking the employer to replace his website, its ATS or any of its existing workflows. And outside of work, I’m originally from the UK, but I now live in the Bay Area. I like to cook and travel and generally just spend time with my wife and two young daughters.

AJ Selvey (01:41):

Would you mind giving us just a little bit more of an overview of the company and what you focus on specifically in the job search ecosystem overall?

Ben Russell (01:50):

Yeah, absolutely. SonicJobs focuses on making that journey from job discovery to application, again, work better for both parties, so for candidates and employers. So historically that has meant helping candidates find relevant roles and removing friction from the actual application process. But more recently, we’re also helping employers adapt to this new age of AI-led search and discovery.

AJ Selvey (02:18):

From your vantage point, as far as where you are now, what are you seeing change most significantly about how candidates are finding jobs today? Because I know it’s dramatically different from when I recruited back in the ’90s and even when I was at a job board back in the early 2000s as well.

Ben Russell (02:36):

So AJ, I think it’s fair to say search behavior has fundamentally changed. I mean, you can see that in the meteoric rise in AI assistance like ChatGPT usage and also the corresponding drop in traditional Google search, so traditional keyword search. And so we’re seeing that move from searching in keywords to asking in complete thoughts, in complete sentences. So in our world, candidate may not know the job title they need, but they know their experience. They know the hours they can work, how far they can travel and what they want next. And so they increasingly expect to describe their situation and get a useful answer. So I think the best example of that comes in the most common job search query over the last 20 years, really since Google, in jobs near me. I’m of the opinion I’d make the prediction that this search term, jobs near me won’t exist in a few years time, because when you think about it, it’s really a symptom of the poor search experience we’ve been stuck with for the last 20 years because we know that’s not what candidates actually want.

(03:46):

And if you’re based in the center of London or New York City, and this is still where that’s the most common search, they’re not then going on to apply to every single job near them, barista, accountant, sales manager, it doesn’t matter what it is, it just needs to be near me. They search that way because they know that search isn’t good and it’s not good at understanding what they want. So the candidate’s basically saying, “Hey, just give me everything and I’ll do the work. I’ll filter. I know what I can do.” So they can scan a list. But if you think about jobs near me, and again, New York City, any big metric, you might get tens of thousands of results. It’s a lot of work. So that’s fine. We all did that. We’ve all probably done that search before at some point in our career.

(04:30):

But now again, you can actually describe what you want and get matched with the jobs you’re actually seeking and actually can do and want to do. So it’s a really big and I think a really exciting shift. 

AJ Selvey (04:42):

I mean, that’s definitely a big shift and things have definitely changed over the years as well. So in that same sense, where are people going today to find those opportunities? Again, being in that space, I’ve seen it evolve, but from your perspective, how has that changed over the last several years?

Ben Russell (05:00):

So I think candidates are still using a lot of the same channels, so like job boards, obviously Google, social platforms, and of course employer career sites. But as we mentioned, I mean, these AI assistants are becoming a really significant discovery layer. So ChatGPT, Gemini, Claude, now boast over a billion weekly users and are growing five times faster than the early days of Google. And so if we think back, it feels like we all blinked and woke up one morning and Google was the dominant player. If you compare that to now, it’s actually happening five times quicker than those early days of Google search, which is really exciting, really interesting, and provides a couple of specific risks and of course opportunities. On the risk side, AI is influencing discovery a lot. So if you think about Google, you may start on page two or page three of results, get a smaller slice of traffic, and then you can work your way up, keep iterating, you move from third page and then you move into a second page, and ideally everyone wants to be on that first page and you’ll see your traffic, your candidates, your visits increase accordingly.

(06:09):

But in the AI world and on ChatGPT and these platforms, if you’re not the answer, if you’re not an answer in the limited answers that these platforms are designed to give, you’re essentially invisible. There isn’t the same long tail effect that there is in traditional search, which is obviously, again, a massive risk, but also if you can embrace it and really move quickly in this space, a massive opportunity. So employers need to think about both places, how AI describes them externally, and then also what happens when a candidate arrives on their own site.

AJ Selvey (06:44):

That’s still been a big part of the discussion too, as far as what happens to that job seekers experience once they get there. So definitely something that still needs to be addressed overall. But if we kind of go back to the job seekers, the candidates, you mentioned it earlier that it was keyword based before because we’re using Google, we’re throwing in keywords that we may know here or there, and now it’s more towards phrases. Is that going to continue, you think, going down that path as we go into these LLMs?

Ben Russell (07:17):

Yeah, I think so. I mean, I think the shift is more fundamental than it being another channel. We don’t think about ChatGPT, Claude Gemini, your AI assistant of choice as being another channel. Again, it’s that fundamental shift in behavior. So traditional search, whatever platform you were doing that on, makes the candidate translate their life into a job title and a location. As we mentioned, lots of the time the candidate can’t necessarily even figure out how to do that. So that’s when they’d start with something like just show me all jobs near me. Whereas conversational search lets them start with the actual problem that they want solving. So instead of your job title, location, you can have something more interactive. I’m a medical assistant. I need evening shifts. I can’t do evening shifts. I want somewhere I can progress, et cetera, et cetera. So they can really refine the answers with follow-up questions.

(08:16):

So in our early data on CareerGPT, we can see this conversation happening in real time and we can see that conversations on average are five messages long. So it tells you that people are exploring and clarifying, not just typing a query and then scanning a list, try to find the needle in a stack of needles.

AJ Selvey (08:40):

Makes perfect sense. And then really flipping the script a little bit and thinking about the employer side of things, because they’re the ones with those roles that they’re trying to get advertised and trying to get more visibility towards and finding that perfect applicant and perfect candidate that hopefully will be with them long term. What are, I guess, the misunderstood pieces that employers have right now or their misconceptions about how candidates are actually searching for jobs?

Ben Russell (09:08):

I think employers often assume the candidate knows the employer’s language and actually I guess knows the employer way better than they’re likely to do. So their job titles, department names, they all make perfect sense inside the company, but candidates don’t think, and maybe more importantly, don’t talk, communicate in that way. So again, today they can describe their skills, their constraints, their ambitions. And if the search experience only works when someone guesses the right keyword correctly, then the employer is losing a lot of good candidates before a recruiter can even see them. They also assume to know what the candidates actually want and I guess how they’re searching for their brand. In today’s game, GEO isn’t the hardest challenge conceptually if you have the data, because if you know how candidates are searching in natural language, you know what they want, then you can build content to match.

(10:14):

But without CareerGPT, which I’ll hopefully have some time to talk about a little later in terms of the data aspect, without that, no one has that data because the big AI labs don’t share anything compared to Google and Google Search Console where you see all the keyword searches, the volume, you really see how people are using because for Google, they just want you to bid on these keywords. But ChatGPT, et cetera, they don’t give any of that data out. So you’re just playing the game in the dark. You don’t know how people are searching, so you don’t know what content in order to build. You don’t know if it’s really working or not. You are really playing that in the dark. And they’re the two biggest challenges in this new AI led GEO world that we’re now in.

AJ Selvey (10:58):

And really, let’s go to the data, right? So according to your data that you have there inside of jobs, how are job seekers typically searching for them? You’re going to know better because like you said, nobody has access to that information because the LLMs keep it tight.

Ben Russell (11:12):

The CareerGPT is a really interesting product for a couple of different reasons, but just on the data side, what we can see with the data from CareerGPT, which again is just sitting on your existing career site, is how conversational and practical these interactions are. So people we can see, again, and we report back to the employers we’re working with, how and what people are asking about roles, location, shift, qualifications, benefits about fit. You can see, it’s quite interesting in the data, a lot of people start the interaction, I guess, thinking or maybe fearing that it’s another not intelligent chatbot. And so they kind of treat it that way. They almost put that keyword in. But once you see the richness of the response that we’re able to provide, people then start shifting how they’re using it to be more like ChatGPT where you can see people then instead of going from keyword and location, they can see those results and then they start thinking, actually, let’s do it the other way around and actually start putting in the constraints.

(12:13):

So I don’t have this degree or that degree, or I have this experience, but I don’t like that aspect of the job. Again, we’re seeing those really deep interactions with the product. So as I mentioned, we’re seeing on average five messages from candidates searching, refining, tweaking. And then from that, within CareerGPT, we’re seeing 96% of those conversations produce a relevant job recommendation. And maybe more exciting, certainly on our side and for the employees we’re working with, 57% are leading to an application start. So we can really show top of funnel, you can provide that really knowledgeable, intimate, recruiter-like experience, but for all of your jobs instead of just the one or two hardest to fill. So again, we’re super excited what early data is showing and where this product is going.

AJ Selvey (13:03):

Those are fantastic numbers. I mean, anything above 50% is kind of unheard of at this point.

Ben Russell (13:10):

Absolutely.

AJ Selvey (13:11):

And so if we get into AI and the evolution of search overall, because it’s way different than what it was 20, 25 years ago, but we’re seeing how people are really searching across the web to look for something. So something that always kind of pops into my head, and you see this stuff as you go through any of your social feeds is stop using AI like a browser, don’t use it like a Google search, but you kind of see people continue to do that. What does, I guess, the evolution look like? And you’ve kind of alluded to it a little bit in terms of the job search piece of it versus what has been done in the past. And you want to kind of expand on that a little bit.

Ben Russell (13:54):

As we mentioned, job search is definitely moving from a page of links, which essentially is what traditional search does today. You put some keywords in, you then get endless numbers of links for you to figure out this one is probably the best. And then you go to another website and then you almost start again, where is the answer I’m looking for in this data one step removed? So it’s moving from that page of links to an answer, but also a recommendation and guidance. That’s the conversational nature of it. So instead of making a candidate open 10 job descriptions and pair them manually, AI can do that for you. AI can understand what the person is trying to do, what they’re trying to achieve, and then ask follow-up questions to refine, again, to guide and surface the roles that fit. So it’s really important, therefore, that the answers are accurate.

(14:53):

So it’s really important, therefore, that the data and the context is tightly controlled. So again, with our product with CareerGPT, bringing that Career GPT-like experience onto the career site, it’s grounded in the employer’s data, the live jobs, approved careers content, so it can give the candidate a better answer whilst giving the employer that visibility into what questions are actually being asked. That feels like the biggest shift. You don’t need to hit it on the first strike where, again, you put your job title in location, it’s like that’s as good as this search is going to get. Now you do the work. Now you can really go deep, spend more time on the discovery layer, and then hopefully just find the right roles that you can then go on to apply. So it’s a lot more efficient to spend that little bit more time upfront, again, really easily.

(15:49):

Natural language is the best method of doing this. Spend a little bit more time on the search, on the discovery piece, and then at the end going, these are the two, three, ideally one, this is the perfect job for you, and then to go on to apply versus needing to do the other way around of going, okay, I’m not sure this is the right one, so I’ll apply to it anyway rather, and applying to hundreds, thousands of jobs and that kind of scattergun approach.

AJ Selvey (16:16):

The way people are doing it or had traditionally done it, yeah, I mean, it’s not as refined as it could be through the natural language conversation that you would have through an LLM, because we talked about keywords and that natural language search too. Can you explain the difference in terms of the results that we would get to where if we’re using an example like, I don’t know, marketing jobs in Boston versus what marketing jobs would be a good fit with my experience? And I’m assuming at that point by asking a question like that, that I’m providing maybe my skillset, my resume to see what would be available. But again, you’re the expert in that. Let’s ask you, find out what you’re seeing and what people are doing.

Ben Russell (16:57):

The first query kind of gives you everything that matches in those keywords. So marketing jobs in Boston, in some ways, the candidate knows what they’re going to receive from that and then you got to do the work because I’m sure there are a lot, I would speculate, tens of thousands of marketing jobs in Boston. Whereas that second example, what would be a good fit? It’s asking for judgment. So to answer it well, the system needs to understand the person’s experience, as you mentioned, AJ, and I guess what a good fit means and which requirements matter. And so what we’re seeing and what we’ve built with CareerGPT is the need for follow-up questions. And that’s really, I guess, the difference and what really changes the experience from just a retrieval mechanism to guidance. And again, that’s just exactly how we would interact, AJ. If I was a recruiter in this scenario and you were the job seeker, you might start with something like, I’m looking for a marketing job in Boston, but I wouldn’t go, great, here’s all 10,000, off you go, that’s all you want from me.You’d immediately ask a follow-up question like what level of experience do you have?

(18:08):

Any industries that you’re looking for working patterns? That’s why recruiters is obviously just a massive and established industry. Again, I think we can bring that experience now with AI, and that’s exactly what we’re trying to build with CareerGPT. And a slightly bigger point, I think this is a really encouraging and really optimistic use of AI. I mean, if you’re a CEO in the US, I bet job search isn’t that hard because you’re bombarded with recruiters who are trying to find you that next role, really working hard because they’re obviously going to get paid on the backend, so you probably don’t have to apply to maybe any jobs manually going through an ATS. You don’t need to do all this search, reading and refinement. You’ve got literally someone working for you to do that. But what about hourly workers? What about people looking for part-time work?

(19:03):

There isn’t a model today which works for that. There isn’t, to my knowledge, recruiters working for someone part-time two evenings a week at that coffee shop because economically it doesn’t work out for either party, for the employer to pay for it or to the other side, the recruiter to do that work. But again, I think that’s able to change now where I think everyone can have that personal recruiter experience irrelevant of the job that you’re applying for. And so I think in this world of AI moving in a million directions all simultaneously and by writing lots of negative stories, I think we can build a really positive one here with CareerGPT. And again, really help candidates and help employers irrespective of the individual role. It shouldn’t just be CEOs, CFOs that get this kind of beautiful concierge experience.

AJ Selvey (19:56):

Yeah, that’s a great example because I was actually wondering about the 70% in that service industry, right? So that’s where majority of the jobs sit. Absolutely. So those who are in the C-suite, they typically do have a network of people that they know, okay, maybe somebody’s going to need me here, somebody’s going to need me there, or my skillset may be a fit there or here. But it also leads into, we talked about job seekers and what they’re doing these days and their behaviors changing, but employers need to meet that behavior now. And so they’re typically used to more of the keyword-based search versus what I believe to be more of the skills-based discovery, meeting those natural language models. Can you talk about what you’ve seen and what you’ve advised even some employers to look at?

Ben Russell (20:39):

They need to make those opportunities understandable in the language that candidates actually use, not that kind of internal jargon we talked about earlier. So the job feed, the careers content, it must be current, it must be up-to-date, and it must be clear enough for an AI system to understand the questions about skills, schedules, location, pay, progression, all of that, and really be able to answer, which again is a shift that we didn’t need to consider even as much as a couple of years ago. So now really thinking about how to build content so it’s really clear and is proactively answering questions because that’s now how candidates communicate. And then the other side of that, which we touched upon is they also need to know what these questions are because as we mentioned, external AI platform, ChatGPTs of the world, they don’t hand any of that data.

(21:32):

Again, just to reiterate, I just know they don’t give any prompt data back in terms of what’s being searched, the volumes, how you’re showing up for that. So the employers without it, the employers never really know how they’re doing and whether they’re optimizing the right thing. So it kind of goes both ways. And again, that’s how we’re thinking about this challenge of you need to build content in all forms in job listings and in the website to be really clear, but you also need to have the right content. We think about how you solve it as a growth loop because it’s not obviously going to be a one and done. How people are searching today is going to keep changing as more and more candidates enter that AI-led search and discovery.

AJ Selvey (22:14):

And the keyword-based piece of it from still current, still people are still using it that way, but I would say even maybe some of the past, you could find that information out if you’re looking for SEM and SEO type of keyword type of information, and how much you’re willing to bid to show up in various pages for some sponsored ads. Well, like you said earlier, it’s a little bit of an unknown without understanding how the LLMs are going to be reacting to that information. You’ve gone through this process several times. What are some of the biggest friction points that you see in today’s job search and the application experience and are employers really making it harder for applicants to apply than what they realize is happening? Because there’s a lot of things that are packed in there and we could probably spend a good 30 minutes just talking on those pieces, but what are you seeing?

Ben Russell (23:04):

Hey Jed, I really like that word friction. That’s exactly the word that we use. And the first friction point is often even before the application. So we talked about a lot, but weak search, poor matching, huge list of responses. It does really feel like a garbage in, garbage out start to this process. When we think about friction, you immediately go to the application flow, which I’ll hopefully touch upon. But just before that, now we can see almost two parallel worlds, one with that kind of search discovery on a ChatGPT like experience with that specific platform, and then the sort of traditional search. What we hear from candidates and employers is you can really see that stark difference. So I guess the first bit of friction is just real top of funnel, bad search, you’re going to get bad results, you’re probably going to get bad apply starts.

(23:53):

So just improving search and matching feels like you’re 80% of the way there. And then the second is the application flow. A lot of these can be really tough, especially on mobile. Lots of redirection, lots of account creation, lots of typing and retyping, personal data. Everyone’s favorite question, please upload your resume and then spend the next three pages typing out your resume. Each step creates another chance and another opportunity to leave. And candidates are taking this opportunity, that data of apply, start to apply complete hasn’t changed massively in the last 10 years. It’s still around 5%. And put another way, 95% are not completing that application. Maybe a chunk of that is to just do with it’s a poor result for them and then the rest is because it’s just sometimes too hard to apply. We tend to treat discovery and actually making the application as separate systems, but for the candidate, it’s one and the same and it’s one journey.

AJ Selvey (24:54):

I remember companies trying to remove obstacles over the years and still remember a story that an employer was worried about when they changed their application process. They were trying to figure out why the applicants kind of fell off the cliff. They couldn’t find them anymore. I went through the application process with them once and I said, “So you make them watch a 15-minute video before they can actually go through the process. And so there’s your problem. Take the video out.” And things change dramatically for them.

Ben Russell (25:24):

I’m sure

AJ Selvey (25:24):

But it’s just little things like that. They think they’re making some improvements, but they’re putting in unintended additional hurdles or friction points that prevent the success from occurring. So going back to the employer side and really staying there, what do they need to do differently to make it easier for them to find those jobs? What do I need to do if I’m an employer to say, “Okay, I’ve optimized my jobs. We’ve already done the SEO piece. Maybe we’ve done some SEM. I’ve got a great career site. I’ve got an ATS on the backend. What else do I need to do?”

Ben Russell (25:54):

Just touching on your point, we’ve seen a new example of the video. I mean, we’ve seen some fairly, what feel like crazy design choices when it comes to the application. And I think employers see the application as a workflow. Obviously design obviously put a lot of effort into them while the candidates are the one who are feeling the accumulated effort. One extra click on a wireframe design for how we’re going to apply. It’s minor on a process map, but on a mobile at the end of a long day can just kind of feel too much. We agree with you. It’s a delicate balance and something honestly that we’ve though about for a number of years is I guess this balance of friction and there’s good friction and bad friction in the job search process. Good friction is are you real? Which is again, not necessarily something we had worry about maybe 10 years ago, but are you real?

(26:44):

Are you relevant? And do you want this job? Is this the right job for you? We should use technology to almost add as much of that good friction as possible. That’s really the time that you should spend in an application, really thinking about if it’s the right thing for you. But there needs to be a counterbalance. Otherwise, if you’re just adding more friction, we think about it as add more good friction, but just remove as much of that bad friction as humanly possible. The bad friction being, as you mentioned, redirection, lots of redirection, and just lots and lots of repetition. We do think it’s delicate balance. We don’t sit here with this is the answer. You can see application flows live today, which are one page applied. Just give me your email and thanks a lot, all the way to, again, watch that video and do some training.

(27:29):

It’s just about trying to optimize on what could be optimized, remove some of that bad friction. And again, the good friction is actually what recruitment is. It’s, “Hey, what are you looking for? What are we looking for?” And that kind of fitting. So just on that point. And then on the employer side, I guess what can they do to make improvements? What can they do to make things easier? And where should it start? It should start on your own career site. I mean, really just ask someone who doesn’t know your business, can they still explain what they want and can they find a role? Ask someone to try and find, give them a kind of persona, “Hey, you’re looking for this sort of role or just ask them to invent one and just watch them search on your website is just an amazingly impactful, sometimes heartbreaking and interesting kind of way to start.” You don’t have to wait for a full website redesign, replacing infrastructure.

(28:20):

There could be a feeling in the industry where it’s always like, “Oh, we need to change something really fundamental. We need a brand new website before we can make candidates happy. We need a new ATS,” whatever it might be. The way that we think about it and really how we position CareerGPT is it’s adding that conversational layer to the site, but it’s still using all the existing infrastructure. So again, it sits on your existing career site, using and grounded in your existing content, using your live jobs, proof pages, and it can give the candidate that modern experience that now honestly they demand really, really quickly. We’re talking in days, certainly not quarters or years like some of the bigger providers. And it gives the employer that first party insight. It almost provides a little bit of auditability at the same time providing that beautiful candidate experience because you can now search and interact with your website in that new way.

(29:13):

So now you can ask CareerGBT, “Hey, what are the benefits that are offered here?” You can see the answer and it’s the same type of answer that a ChatGPT is going to give. If you’re not happy with the answer, then you can very easily fix it. It’s a little bit of a kind of start local, solve global mindset. I know again, we’ve seen employers be able to make some really quick, easy optimizations from the first days of usage, which is great because things are moving fast and speed is the name of the game.

AJ Selvey (29:42):

Would it be, I guess, a good assumption on my part to derive from the information that you just gave us that the biggest misconception or something that employers need to consider is they don’t have to throw the baby out with the bath water to really start this process. They can just refine what’s already there. Is that really where you want to typically start that conversation?

Ben Russell (30:04):

Yeah, absolutely. That’s exactly right, AJ. We’ve designed this tool, CareerGPT, again, to just sit and really reflect. It’s like putting an LLM, your own personal LLM on top of your own data set. You can dog through this product from minute one, asking these types of searches, really doing actually what candidates are doing today and seeing the results. One feature of CareerGPT that we’re really proud of is you control the content that it’s using, that it’s looking for. So career site is kind of a given, but you can also allow it to index and look at other websites. So this could be your careers page on Glassdoor so it can pull in your positive reviews and what candidates are saying about you there in Indeed, whomever it might be. You can also just upload PDFs of data. The website’s probably the tip of the iceberg in terms of the information, the data associated with working there.

(30:57):

So we built it so you can upload PDFs of knowledge. You don’t need to even wait for, “Oh, I don’t like this answer, but it takes me some time in order to update my website.” We even try to shorten that step where now you can just upload your benefits 40 page document that you give to candidates on day one. But constantly try to find ways just to add that speed and I guess just remove those sorts of steps of this is going to take a lot of time, this is going to cost a lot of money, this is going to need a lot of cross team. We built this so really it can go live in weeks. And then you’re in this modern world and then you can optimize really quickly from there.

AJ Selvey (31:38):

That’s phenomenal. It sounds like you are really very consultative with those employers that you’re working with to help them adapt and modify and change as time goes on. Can you walk us through a little bit more about what SonicJobs actually can help employers do? It’s not just about uploading PDFs and those type of things and consulting. I know there’s quite a few things that you do.

Ben Russell (32:00):

We’ve built CareerGPT in this space to give the employer its own conversational careers experience. I think we can all agree that candid behaviors fundamentally change and this product allows you to cater for it really quickly and really sophisticated. It provides that same experience as a kind of ChatGPT. And just to reiterate, it’s sitting on all of your existing infrastructure. It sits on the existing website, it’s using the live jobs, the approved content, which I mentioned can take a couple of different forms. Obviously it could be websites, could just be what is traditionally offline data that stuff that you have, you can keep feeding into it, making it smarter and smarter. And it’s keeping the existing ATS and all the existing workflows in place. And no data is shared with any of the LLMs, any of these AI companies or any other company at all. It’s entirely private, secure, and fully auditable.

(32:55):

And by audible, I mean we reflect all of this data. You can see all the individual conversations, the searches, prompts that were made, the conversation depth, and we show you all the answers as well. It kind of works both ways. It’s one thing being able to see, hey, candidates are coming and they’re searching for this specific role type and whether there’s this shift schedule, that’s great. You have that insight now, but you can also see what answer’s been given to the candidates and whether you’re happy with that. And thinking about how that scales to the global challenge of getting that visibility in the actual AI assistance, it’s the exact same process. We think about you can solve today and your prize for solving is the metrics we talked about earlier, AJ. Almost 100% of candidates are finding a relevant job by using this tool and 60% are style and applies.

(33:47):

You get this massive conversion uplift on day one whilst you’re also building for the future and in this GEO optimized world that we’re already in, but the start of that we’re only going to be moving more into for the next 20 years.

AJ Selvey (34:02):

Yeah, it’s just crazy how quickly things have changed in a relatively short amount of time, right?

Ben Russell (34:08):

Yeah, yeah, 100%.

AJ Selvey (34:11):

And for employers that are kind of listening to this now and kind of reflecting on what they have in place where they may say something like, “Hey, our recruiting strategy really hasn’t changed. I’ve been here for 20 years, 15 years, 25 years, and we’ve kind of gone from paper applications to online stuff and now we have an ATS and a career site.” Really from your side of things, where should they start and how can you initially help them get over that hump to understand, “Hey, this is where we need to go, get with the times.”

Ben Russell (34:39):

We’re obsessed about this, AJ. So one thing I guess I’d say from the SonicJobs perspective is we love talking about this challenge because again, we remain really optimistic at a time when maybe there isn’t as much optimism, and again, probably quite rightly, about how we can use this emerging technology to really improve outcomes for both parties. So I guess the first thing I just would say is we love talking about this stuff. We think about it deeply, how to optimize for this new GEO world, how to optimize my brand, my jobs, my company to show up in ChatGPT, Claude, Gemini. So the first thing I guess I’ll say is we’d love to hear from any employer who’s thinking about this challenge, which honestly AJ, from our conversations is every single employer. I can’t imagine a single company, a single Fortune 500 company who hasn’t had this exact conversation and this challenge probably passed down from leadership of what are we doing with the likes of a ChatGPT?

(35:35):

How are we getting our content, our jobs, our brand out there? So first thing is we’d love to talk, to share stories and some of the insights that we have. Again, we’re based in the city in San Francisco, a couple of miles from all of these players, so we’re an open book on that respect. The other thing we can offer is speed. We can put CareerGPT on your existing career site and dates, not quarters like a lot of other providers, and we certainly don’t have the same costs and overhead associated with them. We’re trying to position this product, CareerGPT, to be honestly a no-brainer starting point to move into this new world. It doesn’t need all of your teams to work together and for you to change and do a massive integration implementation. We’re bringing a personal, private, secure ChatGPT, Gemini, and placing it on your career site.

(36:23):

And so we can do that really quickly and the insights you get are from minute one. The benefits are twofold. The biggest one is you’re just providing a way better experience immediately and therefore converting a lot more candidates into applications. That’s the game. That’s why we’re doing all of this is to help candidates find the right job and apply and get hired. That’s what employers want and that’s obviously what candidates want. So you get all of that value, but in the meantime, you’re also collecting all of that interesting and insightful data, how candidates search, turning a light on in this game that everyone’s kind of playing in the dark, you can drive really interesting proprietary insights and then you can feed that into the LLMs and start driving increased acquisition as well. I would say, AJ, there is historical precedence for this. If you think about the early days of Indeed, they mirrored Google almost exactly the search, the website, the experience, everything.

(37:22):

I think for a while they even referred themselves as Google for jobs. Obviously now that’s kind of funny that there is actually a product for that. But the idea was that if you look and act exactly like this new platform, then candidates will use you like they use these platforms. It’s why Indeed figured out jobs they made quicker than any other platform. They collected data like Google and then they were able to optimize and then ultimately dominate this new channel. We think that acting like ChatGPT is a really good way to optimize for ChatGPT and all of these AI assistants, again, because you get the data. I think it’s just really interesting, a little bit look at the past on how things are kind of similar and different with Google. We think there’s a lot of similarities about how providers who have really embraced this new paradigm of search have gone on to win.

(38:15):

And I guess maybe it’s no surprise, Indeed’s try to do it again. I know for a lot of employees we speak with, that’s not what they want. They don’t want to get stuck behind these intermediaries for another 10, 20 years. We’re offering speed so that you can really strike now whilst the game is just beginning.

AJ Selvey (38:33):

You kind of hit the nail on the head there with how there’s the traditional old methodology, if you will, versus the new and the old still kind of hanging on and trying to adapt and align with it to keep their strangleholder on the market where employers are still trying to break out of that and kind of be seen on their own instead of standing on their own pedestal versus paying somebody who’s already there.

Ben Russell (38:53):

Yeah, I agree.

AJ Selvey (38:54):

So if we do really look ahead three, five years, whatever that may be, a couple different questions come to mind. What do you think the job search process looks like in terms of if I’m looking for a job later on, what does that look like for me? And at the same time in the recruitment process, is that going to look completely different too?

Ben Russell (39:11):

I think we can all agree, AJ. The genies is out of the bottle now with this new way of search and natural language is a way better vehicle for this. It’s why we’re communicating in that way now. We call it natural language. It’s really just how people talk with each other. Going back to the example I gave earlier, if you imagine speaking with a recruiter, you wouldn’t just repeatedly yell the job title and location that you were looking for. You certainly wouldn’t say jobs near me. Jobs near me might be an aspect of what you’re looking for, but you talk about it in terms of why that’s significant. Is it about your access to transportation? Is it about your home life that you’re trying to cater for? You just talk about yourself, right? You talk about yourself, what you’re looking for, what you’d like, and you’d expect that recruiter to talk back and ask you questions and kind of refine.

(40:01):

It’s why recruitment is, as an industry, is whatever it is, 30 times bigger than traditional online recruitment marketing as an industry. It’s a lot more valuable to both parties, and we talked about the kind of scalability of that. It’s fair to assume that candidates will continue to begin with that conversation. On a ChatGPT example, Google is obviously well-placed to still be market leader in this space because obviously Google is Gemini. And so you can see AI search becoming a more prominent part of their discovery layers. They’re all going to move to this new way of starting. And we expect that the conversation will just continue. So it’ll continue across all the steps in the hiring funnel from these platforms to career sites and then onto actually applying, interviewing, onboarding, et cetera. So ultimately employer career sites will need to be able to answer, not just publish data.

AJ Selvey (40:57):

Very true. Very, very true. So as we think about today’s recruitment process overall, what do you think will look different, will it look any different in the future?

Ben Russell (41:07):

SonicJobs’ position and my opinion is that career sites will actually look completely different because they’ll need to serve two users. And again, this is just not something that we really needed to consider even a few years ago. They’ll need to serve real candidates and AI agents because both will be extremely prevalent in the market. The way that we think about that is the career site will become like a knowledge database with an interactive layer because AI agents don’t mind doing the extra work, the extra click, the extra paragraph, that’s what they’re designed to do. And so you can have a massive website with loads of data structured really cleanly, optimized for these LLMs, but the candidates almost going the other way, whereas the AI agent will work even harder, spend more time reading and understanding. Candidates are kind of moving the other way. They’re expecting results fast, they’re expecting answers.

(42:07):

So the candidate facing layer will be more conversational. So actually I think it’ll move in two very interesting directions from what a career site looks like today.

AJ Selvey (42:17):

Yeah, it’s going to be interesting to see how that evolves over time because everybody still has that traditional thought of what a career site looks like today and thinks that we’re there or they optimize it every six months, every year.

Ben Russell (42:29):

Exactly.

AJ Selvey (42:31):

Throw not just a career site change, but throw a branding change in there in that mix. That just throws everything out the window too. With that in mind, with employers that are listening, what’s one thing they should reconsider about how they approach the job search today?

Ben Russell (42:44):

The simplest way that I would put it is that candidate behavior has fundamentally changed. So there’s a real opportunity to embrace this shift without completely upending your existing business. With CareerGPT, we’re letting an employer meet that expectation on the career site it already has using the information it controls while learning again what candidates actually want. We think this is a really meaningful change that employers can make now without waiting before their competitors and really before these other platforms, these other job boards, again, dominate this next channel as well. And so we’d really encourage employers who are thinking about it, knowing they need to act in this area, obviously to reach out, but really to take some action because again, I think you’d be surprised how quickly you see positive results.

AJ Selvey (43:39):

And you’re really trying to help the employers take ownership of that entire process now.

Ben Russell (43:43):

That’s the goal, AJ. Absolutely.

AJ Selvey (43:46):

As we’ve heard today, job search is really changing and employers need to change with it. Candidates are discovering opportunities in new places, using new techniques and really increasingly expecting a job search experience that’s simple, that’s relevant, and align with the way they search for information for the rest of their lives. For employers, that means thinking beyond simply posting a job and waiting for applications, the old post and pray. It’s about understanding where candidates are searching, how they’re discovering opportunities, how job information is presented, and what happens once they decide to learn more. Ben, I really appreciate you and thank you for joining us today and sharing your perspective on where job search is headed and what employers can do to keep up. We appreciate you helping our audience think differently about how they show up for candidates. And if folks want to get in touch with you or your team, what’s the best way to do so?

Ben Russell (44:44):

Firstly, thanks AJ. It’s been a real pleasure. I appreciate you inviting me. And yeah, you can reach out to me on LinkedIn, Ben Russell, SonicJobs, or go to our website, Sonicjobs.com. There you can reach out, you can schedule a demo to learn more about our new CareerGBT product and some of the other initiatives we’re working on.

AJ Selvey (45:02):

And finally, a special thanks to DEAMcon underwriter, SonicJobs for being part of today’s conversation and for helping employers navigate this evolving job search landscape. Thank you for joining us for another episode of the DE Talk Podcast. Until next time, thanks for listening.

DirectEmployers Association (45:18):

Thank you for tuning in for another episode of the DE Talk Podcast. Stay connected with DirectEmployers on Facebook, Twitter, and Linkedin, or subscribe to receive updates straight to your inbox by visiting DirectEmployers.org/Subscribe, or you’ll receive notifications of new episodes, webinars, events, and more.