Leveraging AI in Recruitment With Chris Sund and Holly Stewart

AI is all the hype at the moment, and in recruitment, it’s obvious that it needs to be used, but how it should be used is not as clear. Today on The Staffing Show, we welcome the president and COO of Uniti Med and GQR Healthcare, Chris Sund, and the head of software solutions, technology, and trading at GQR Healthcare, Holly Stewart, to discuss how you can use AI to impact your business and how their new AI recruitment platform, Nebula, can help you do that. 

Tuning in, you’ll hear all about what Nebula does, how it differs from other large-language models, how top-performing recruiters are using AI to save time and recruit different people who may not have otherwise been noticed, and so much more. We delve into where recruiting needs human relationships and why those areas should never be automated, before discussing how agencies can get the most out of Nebula and how they are making the adoption of this tool easier for recruiters. Finally, our guests share their thoughts on the future of agencies that make use of AI tools versus those that don’t. Thanks for listening.

[0:01:13] DF: Hello, everyone. Thank you for joining us for another episode of The Staffing Show. Today, I am super excited to be joined by Holly Stewart, who’s the Head of Software Solutions, and Chris Sund, who’s the COO of Nebula. Thank you guys so much for joining me today. Very excited to have you on the show.

Today, we’re going to be talking about AI, which we have always talked about. It’s what everybody’s talking about all the time. It’s hard to avoid. We’re going to be going deep into some best practice around data and learning some new things around data enrichment. I think it’s going to be really impactful if you are looking at ways that you can use AI that will have an actual impact on your business today. Chris and Holly, really excited to have you on the show. To kick things off, could you tell us a little bit about who you are and then a little bit about what Nebula is doing in the market today?

[0:01:58] HS: Absolutely. Thanks for having us, David. We’re excited to be here. As a quick intro, I have been in agency recruitment for about 10 years all over the country: Boston, L.A., D.C., New York, before moving into SaaS and that world. I deeply understand the agency persona and what goes into that decision-making process, and that’s why I am so excited, specifically about bringing Nebula to market. Chris, you want to introduce yourself?

[0:02:28] CS: Yeah. Thank you so much for having us. I’ve been in staffing for about the last decade. All of that time has been in healthcare staffing and working across travel, as well as locums industry as well. Excited to say the least, to talk a lot about this software.

[0:02:47] DF: Thank you, guys, for the introductions. Why don’t you tell us a little bit about Nebula? What are you guys doing?

[0:02:52] HS: Nice. Nebula, at its core, is an AI-powered recruitment platform, and I can confidently say it is the biggest productivity unlock that recruitment has seen in years. It makes recruiters four times more efficient. At its core, and obviously, we’ll get into a lot of the key differentiators here, it is a human-led tech-enabled product that automates manual repetitive tasks and allows your best recruiters to focus on human-centric, high-value activities that AI we believe should not replace.

[0:03:29] DF: That’s great. That sounds really exciting. Chris, did you have anything you wanted to add to that?

[0:03:34] CS: I think Holly did a nice high-level, and we’re going to get into. It just has a lot of different capabilities. I think during this time, we’ll get into what’s that mean for automation and AI versus that human touch. That is so important in recruitment.

[0:03:51] DF: Awesome. Well, let’s go ahead and jump in then. Holly, one of the things that we’re seeing in the industry is everybody’s talking about AI nonstop, but most agencies haven’t changed how they actually source candidates. Why do you think there’s such a gap between the hype of AI and the reality?

[0:04:10] HS: Oh, okay. I love this question. I think the first thing we have to acknowledge is that AI, honestly, how often you hear it, it makes me nauseous. It’s an overused term that I genuinely think people don’t understand. Maybe a little bit of an offshoot with this comparison, but I am personally very into food provenance and understanding traceability and all that kind of stuff. Sometimes I compare AI to the organic label at the grocery store. As consumers, we know that we’re supposed to want organic. I do think that generally, all agency owners want to be involved with AI and be on the cutting edge. The problem is now everything is labeled organic, and it’s what actually makes a product different, and AI feels that way. The label is everywhere, but the substance itself varies a lot. People don’t know, like am I genuinely getting something new or different, or am I getting a wrapper on ChatGPT, just being white labeled and upsold on it?

I think that leads to a ton of decision paralysis, because a buyer isn’t sure what’s actual innovation versus pure marketing. Then, on top of that, David, I think that staffing agencies are super discerning buyers. They’re sales organizations. Whatever they invest directly impacts revenue. They’re not just asking, does this technology work in a vacuum? They’re asking, will my entire team actually get on board and adopt this if I invest in it, and will it improve how we place candidates? Unless it is delivering these results at scale, they’re going to keep doing the same thing until they know it works.

[0:05:54] DF: Absolutely. Chris, I know you’ve spent a lot of time with agency owners. What’s the most common misconception that you’re hearing about AI and recruiting?

[0:06:03] CS: Great question. I think there’s a few of them. I think one of the misconceptions is, just to piggyback off what Holly was talking about, is you know you should be using it, but how should you be using it? Compared to when the Internet first came out, well, I should be using it. Does that mean I have a website? Does that mean people know how to get to my address? Am I marketing on there? Am I selling on there? AI is the same thing. AI can touch just about every faction of your business. A lot of owners are looking at all these different products that can do all these different things and trying to figure out, what do I buy? Do I buy them all? Do I buy none? Where is that?

The biggest one of all that I thought maybe by this time I wouldn’t hear as much is that AI is going to replace the recruiters, or replace the recruitment. Recruitment is such a heavy people relationship. I tell, really using Nebula for those Marvel people out there that are listening, and maybe get this reference that we are not looking to build Voltron, which in the Avengers is this robot that has its own brain, and it’s going to take over the world. We don’t need that, right? But we love Iron Man. Iron Man is a regular person with his brain, his heart, and our recruiters are that superhero. If we use the technology the right way and we give them these powers, they can do remarkable things with it. It’s just making sure that you’re getting them the right technology to do that.

[0:07:40] DF: Chris, you said that the other day. I think that’s one of my favorite. I’ve already repeated it three times. I’m like, turning recruiters into Iron Man is such a fun way to think about that, of how much more can they do. I also want to go back to Holly. You mentioned there is so much confusion in the market about what to buy, what’s actually going to have an impact, what matters, what doesn’t. You brought up, not the organic concept is fun too, because it’s like, everything is labeled AI. Everybody is buying AI. A lot of it is wrappers and looks great, but doesn’t actually have an impact. From your perspective, what is the actual variation, or difference then within AI, if it’s not – What is the organic supposed to mean for AI?

[0:08:23] HS: Great question. Again, I’ll probably relate this back to food providence, but there’s so many of these labels that are just branding to make you as a consumer feel almost negligent if you’re not buying something that has that label. But when it comes down to it and you actually do the food tracing and sourcing it, what hasn’t meant that much, well, the AI title, again, it’s slapped on a lot of products that are out there right now. Like I mentioned, a lot of these are wrappers on ChatGPT. It’s just making API calls and white-labeling something. I know, specifically agency owners and people in that business are very frustrated to be misled in what they’re purchasing.

How Nebula specifically is different and how other products are differentiated in the market is the large language model that they’re built on. Nebula specifically is its own LLM. It is not ChatGPT. It is built for recruitment and trained on over 15 years of recruitment specific data. The results are always going to be new and different in the natural language processing piece, which we’ll get more into means that those results just come better and better with time. First of all, people should understand what the underlying LLM is in any product that they purchase, and they should understand if it was built specifically for recruitment, or if it’s been recycled from some other place, because it will dramatically impact the results.

[0:09:52] DF: The entire LLM is custom-built.

[0:09:55] HS: Yes, it is. What’s specifically exciting is something that you see a lot these days as pushback to AI, or negative feedback, is that there are discriminatory behaviors, even with an AI, which is baffling to me personally.

[0:10:10] DF: Built by people that have some systemic –

[0:10:13] HS: I know. Which is wild, but it’s true. What’s cool about Nebula is that we actually have anti-bias built into the LLM. Don’t ask me too many questions, because then we’ll have to pull in engineering. But we got a patent, which people know patents, we got a patent on the first try for being somebody that implemented anti-bias in your LLM, so you know it’s good.

[0:10:38] DF: That’s really cool. Yeah, that’s amazing. That’s also such a key part of it, because it could be so problematic with how it does approach. That’s awesome.

[0:10:46] HS: Absolutely.

[0:10:47] DF: One of the other areas, and Holly, I wanted to jump back to you on this as well, but job boards, I think they’ve trained agencies to pay for not only the same candidate over and over. Recently, one of our customers said they did an audit on all of their job board spend. I don’t know the exact amount they were spending, but 15% of their placements that they had gotten from the job boards were in their database, which is crazy. It’s a huge problem. It’s like, all right, you’ve got these people already. How do you think about breaking that cycle?

[0:11:20] HS: Totally. It’s funny that you mentioned that as an example, because I think if a lot of agency owners, they know that they’re spending the same money to get the same candidate. They understand that a lot of these traditional job boards that were actually pretty impactful a decade ago, aren’t what they used to be, but they don’t really know what the next step is. I think that they’re aware of it and they’re philosophically and financially exhausted by it. Nebula breaks that cycle by providing resume, all contact details and normalized skillset profiles for each candidate that can then be imported into your own ATS, or CRM.

It also integrates, as mentioned, with overseas CRMs and ATSs. They’re going to be stored for the duration of their careers. We also have an additional tool set that allows candidates within your database to be rediscovered and engaged. A lot of products out there only address net new candidates, or candidates that exists within your database. This tackles both. It allows you to identify net new, capture net new, and then re-engage the existing database, which is so, so important.

[0:12:35] DF: Okay. I think about this, like the different type of categories. You’ve got enrichment, matching, and then you’ve got re-discovery and then engagement. Is that the breakout?

[0:12:46] HS: Definitely. It’s intelligent engagement, so it’s multi-step sequencing that mitigates human error by running itself once someone who is specifically good for a role is identified.

[0:12:58] DF: Awesome. That sounds great. Chris, you brought this set up the other day that you’d found that 9% of total revenue from a staffing agency goes towards candidate acquisition. For the owners that are listening to this, who really don’t have an idea of what their revenue number looks like, how should they even start measuring this and maybe why is this important?

[0:13:20] CS: Yeah. For context, that is an average across a lot of industry. There’s definitely some flux there, anywhere from 5 to 15, depending on your industry and what’s getting average. Every agency, owner, and leadership has to be looking at your candidate flow and how much you’re spending on that candidate acquisition. You should be looking at all your placements in business you’ve done over a lifetime and understanding where do they come from. In the example of like, they knew 15% of our database, but they also knew in total how many they had.

Once you figure that all out, then it comes down to, okay, how much do I spend on that tool a year versus my placements? Without getting too nerdy in the financial, the best advice I want to get is, I’ve seen companies go, “Hey, this lead costs me this much per placement,” I want to know that. This lead source candidate source costs me this much and this much and this much, but they’re not equal. If I have a $500 and $1,000 and a $1,500 one, of course, I want to spend $500. Now, the question is, are all those lead sources owned by somebody else, or are they your own? Are they driving traffic to your own brand, or in there? You have to give it an additional value.

If I’m buying a lead from another website, for example, I need to put that in that bucket to go, I’m basically paying for their advertising dollars, right? But if that same bucket was if I pay that much for a referral or someone, I spent this on my Google ad, and they went on my website, but I do pay for Google ads, that’s a better bucket. If those two dollars are equal, you’d rather put all your money into your own website and own that marketing in that dollar. I think the biggest advice is you need to know the dollar, but you also need to be smart with your money. You also have to be going, am I helping my own database? Am I helping my own brand, or am I paying to enrich someone else’s brand?

[0:15:19] DF: Yeah. It’s such a funny concept. When I look at the market, I’m like, staffing agencies are funding job boards, marketing funnels, and database building. Then the job boards are simultaneously going out and trying to sell drive. It’s like, he’s really funding the competition. At some point, I don’t know, one of the major job boards, I would not be surprised if they roll out their own AI agent that screen sources and does 50% of what the agencies do. I think that is an interesting way to think about it.

Also, I talk about this a lot on the podcast, but candidate lifetime value by source is what I think the future of tracking is, to make sure that you’re spending correctly and own data, like you guys are talking about, is always significantly more valuable. I know I –

[0:16:09] CS: I want to jump on the job board side, because I don’t want to come across and be like, this – you find your own candidates. You enrich your own candidates. You get them on your website is the only answer. I do not think it is. I know it’s very clear out there. If I am searching and I either use the hotel one a lot, if I’m searching an area for hotels and I would just stick to my favorite brands, because they’re rewards, or I know their quality, that’s great, but I’m probably not doing that. I’m going to go on a Price Line [inaudible 0:16:39], Tripadvisor, Google, and then I’m going to look at all of them and make my decision of where I’m going to stay.

Well, guess what? Every hotel company, every agency is going, “Oh, why’d you have to go on that company’s website? Because I have to pay them a fee and they’re taking my margin.” But guess what? The traffic is there. I just tell people, you got to be smart. You need to be building your own brand. You need to be making sure your reputation lives out there and it’s important. Also, when you do look at the total traffic and bringing, once again, new awareness, I also say there is a value of bringing that new awareness to your brand as well. It’s just being careful if you’re only relying on that and you’re not building your brand at the same time, it’s probably not a smart recipe.

[0:17:25] DF: Yeah. I couldn’t agree more. I think the multi-channel strategy is where we’re at, or omni-channel. You’re going to have to source for multiple channels always, and they serve a major purpose, which is why they do it. All right, Holly, question for you. If you look at top-performing recruiters versus average ones, what are the best people doing differently with technology right now, and why does that matter?

[0:17:47] HS: Good one. Average recruiters are still spending time on manual search methods. Traditional ones that they were taught and have worked for decades, but just aren’t going to put them at the top anymore. Specifically, spending hours building Boolean searches in places like LinkedIn for sourcing databases. Top performers have shifted to natural language search and AI ranking. That means that they can describe the profile they want in plain English, and the technology actually surfaces the most relevant people and ranks them. This drastically reduced the time that they’re spending surfing and this time building relationships.

Now, something important to understand and identify here is not just that technology is doing this quicker. There is an inherent difference between Boolean and natural language processing. Boolean, because it’s keyword matching, it will exclude individuals who don’t have specific words on their profile, versus natural language processing. Essentially, what it does is it looks at contextual clues and says, even though this person does not have the specific word on their profile, I holistically understand that there is very high likelihood that they have the skill set. What’s really impactful is not only that saving you all this time and stack ranking individuals, but it’s also surfacing newer, different people that you never would have engaged with before. Mind you, while all these average recruiters are doing Boolean searches, they’re inundating the same people with messaging.

Whereas, the people that necessarily don’t have these keywords on, they’re not being contacted as much. It surfaces new and different people and then drastically increases a response rate as well.

[0:19:39] DF: I love that.

[0:19:40] CS: I want to jump on top of that. One of the things when we are showing Nebula for the first time to different buyers that you see some excitement on is when you’re looking for recruiting, and what Holly was just saying, you’re using the same tools as competitors, the same major platforms, the headhunt out of, and you put in your searches, you get the same results on the first few pages. We keep going to those same, because they’re based on that. That Boolean search, very basic search terms. What that means is you start pulling up candidates that aren’t getting the same outreach as everyone else. You get these unique scenarios based on how you put it in there and that’s really where some of that magic starts. That’s only the start of the magic of what it does. But right off the bat, you are not getting the exact same people everyone else has seen on page one.

[0:20:35] DF: Holly, there’s a real tension between speed and personalization and outreach. I’ve had a lot of agencies talk to me about how they wish the recruiters would spend a little bit more time making every message unique and they send the same message over and over and over again, and we know that that just doesn’t work, especially if you actually hit the same candidate with the same message a couple of times. Candidates really do care about having to feel like it’s human, feeling like they’re unique. How do you think about this trade off?

[0:21:00] HS: I’m going to disagree with you on the question a bit. I actually think it’s a misconception that there has to be a trade off with personalization and speed and that they have to be at odds. In reality, I think the tension exists, because recruiters are trying to reach out to too many people, who aren’t actually necessarily a good fit. What we statistically see with top recruiters when they use technology like Nebula is that they’re able to narrow the market down to relevant candidates at first. When you’re trying to reach out to the right 20 people, instead of the wrong 200 people, personalization actually becomes a lot easier. Couple that with how our campaign style technology functions, which is to say that the AI will build you a multi-step outreach campaign that is a mix of email and text message, and there are a lot of different personalization fields that you can insert into the message to make it feel very custom.

Once that flow is set, you are really able to set it and forget it and it sends it at the exact data that you want to. I do think with a set up like that, there doesn’t have to exist tension between speed and personalization. You can address both.

[0:22:16] DF: Yeah. I feel like, the examples I hear about that are like, oh, I’m taking and copying and pasting this for my Excel spreadsheet, or I’m using a tiered, just taking, doing the same message to everybody every week. It sounds like, yours has the flows and some personalization tokens built in throughout it, so it does feel a little bit more human, but allows the speed.

[0:22:37] HS: Absolutely. We understand that people have different pieces in their tool set and they leverage Nebula in different ways. Something that’s beautiful about the product is it is truly end to end. You lose a lot of this statistical chance of human error when you don’t have to move between a lot of these products and it’s another way that we increase efficiency and productivity. You go from this smart search and sourcing tool set to this super personalized campaign, multi-step outreach, and then you can actually jump directly into our third feature set which is ATS-like tools, where you can move candidates through a custom pipeline. It just really makes the process super intuitive and again, mitigates human error.

[0:23:23] DF: Awesome. Sounds great. This is a question that I’ve had this conversation probably 50 times in the last two years. I had it yesterday even. I think that the reason we’re all having it so much is that we’re seeing AI evolve so rapidly that the mark for it keeps changing. What parts of recruiting do you think should never be automated, or maybe will never be automated and which parts should stay connected to that human relationship and what your guys’ perspective on that?

[0:23:54] HS: I think there’s a lot of products that have tried to automate every single piece. We actually disagree with those. There’s a lot of different products on the market now that are trying to do the initial pre-screen automated intake call themselves and then pass that data along. We absolutely do not think that that should be automated. I think the parts of recruitment that should never be automated are moments that require trust and nuance, and there’s so much of that. Obviously, I think those agencies can become numb to it, because we have these conversations every single day. But whether they’re leaving their current job, or relocating, or taking a career risk, but that is not a transactional decision. It is a human one. They need to have a human that they’re talking to who understands the company, their specific career goals and can give honest advice.

Technology can and should surface the right candidates. It should streamline outreach. But anything with those pivotal conversations where somebody is making a life decision is where great recruiters will always matter.

[0:25:01] DF: That’s great. To talk like, any of the transactional stuff, I think that’s what we all talk about getting automated. The screening, the collecting information, this is a fast way to do it. But a big life decision relocating, changing jobs, some of those bigger elements. Any other specific areas that you guys feel like, just are going to stay, make it three years, five years from now, anything else that has had this conversation so much, I’m trying to uncover what does it look like in the future, because it is, I feel like with AI voice in the last three months, just see the evolution of it is changing quite a bit.

[0:25:34] CS: I guess, I would think five years from now, 10 years from now, would you want to go in for an interview and have a robot on the other side of the table interviewing you and deciding if you’re going to spend your time, your career? We spend so many thousands of hours with our career and our time is our most valuable asset we have. Holly said it, that’s such a major life decision that those type of things, you want to person. I’m not saying it’s going to get better, but right now, I don’t know, over the holidays, I had to make a phone call to customer service. It was a large company, and so they have some good money invested. I would say, it’s one of the better voice AIs that I’ve probably dealt with, but it was still horrendous. It was not as fast as a human. As you’re talking, you know what you’re going to say.

I don’t know about you, but any time I use ChatGPT, it rarely even gets the answer right on the first time, right? I usually have to go back multiple times, where a normal person would have heard that and said, “Oh, that’s what you need.” We’re talking about a job here. Yes, there’s functions. You want to collect my data and put it over here, you want to take this file over here, you want to do that. That’s great. But when it comes down to whether or not where I’m going to work, who I’m going to work for, those are some of those human elements that just, it makes the most sense to keep it that way.

[0:26:57] DF: Great answers on both fronts. Holly, I want to go back to the data side of this. When it comes to building a competitive advantage with your data and combining your data with this enrichment layer and Nebula, something along those lines, what are some things that the audience should know when they’re thinking about how to get more out of the database and their data strategy?

[0:27:20] HS: Great question. Again, going back to my oversaturation with the term AI, I think there’s a couple terms that we should define for the audience, so that they know what to look for. All of these pieces exist within Nebula. First, predictive. Predictive means that the product is making intelligent recommendations and it has high accuracy matching. We have that in our search and sourcing component that’s surfacing these net new candidates through natural language processing. Another term that you see thrown around a lot, when we talk about the data science behind AI, is agentive, and agentive essentially just means intelligent automation. There are a lot of different agentive pieces within Nebula that will automatically take someone from identification to the project management to automatic outreach.

Another agentive piece is that we do sourcing for you once you spin up a role. An agency owner, a recruiter spins up this position, and they set the parameters, they pull a couple of people that they like for the search. How we make them more efficient is by allowing them to spin up multiple searches concurrently. The agentive automation, what it does is it will source, identify, and start to message people in a search that you’ve already spun up for you. You can spend one day, spinning up two or three different searches and it will continue that outreach for you, so that you can take on more of a workload there. There’s a lot of very important impactful agentive pieces within the product.

Then the last term I’d go out to is generative. This essentially just means personalization at scale. It comes back to the campaign and multi-step messaging that we’re talking to. It’s about being able to mass message, while having that messaging be very targeted and feel very personalized. Some just quick data points on the product itself is from a quantity standpoint. We have the entire US and global labor market map, so there’s over 200 million profiles within the US and 1 billion globally. Huge numbers.

From a quality perspective, each data profile is rich in aggregated data, including data that’s being pulled from public and private pools and then structured in a very usable way. The third piece that I’d highlight about Nebula is the recency and the readiness. These are not stagnant profiles that have been existing on job boards since the beginning of time. We use intent signals, and we flush these data profiles every 90 days. All of the profiles that you see in there are continuing to be refreshed with a degree of normalcy.

[0:30:06] DF: That’s great. I’ve over the years used so many different data tools. One of the issues historically has been the open rates just drop, the spam rates drop. It sounds like you guys are actually taking care of that component of it with the AI as well, which is awesome. One of the areas that –

[0:30:23] HS: That I’ll add to. For agency owners, one of the biggest pieces is being able to monitor the impact of these tools as well. There’s a lot of data points that exist within the tool, so you can see into those who roll up into you in terms of open rates, land rates, number of folks contacted. Not only can those individuals A-B test what kind of messaging is landing, but the agency owner can understand how effective the tool is as well.

[0:30:49] DF: That’s cool. That’s cool. One of the things that I know we’ve talked about is the enriching, updating candidate profiles. For somebody who doesn’t have a tool like yours, what’s a practical way to make sure that they’re keeping their database from going stale?

[0:31:01] HS: This is tough, because practical and easy are not the same thing. Even small agencies can have significant databases. Practically, recruiters need to continue to update candidate profiles with updated contact data, salary information, skill set information. They need to categorize those individuals into tear sheets, so they can re-engage them over time and engage them in whatever cadence is going to be most impactful, to just ensure brand recognition. While there are those practical steps that they can take, they are super fallible and super monotonous, and this is not where people should be allocating their time. Agencies that want to stay on the cutting edge and that want to keep the best possible recruiters and leverage them at their highest utility, they need a tool that can do those pieces for them.

[0:31:52] DF: The idea of having a recruiter call everybody is like, good luck, and update the record correctly. Speaking of that, so behavior change is something that is difficult always for all of us, I think, especially when you’re training teams. Holly, when it comes to getting recruiters to change how they work, that can be one of the harder things to do, especially if there’s a technology adoption curve. What have you learned about change management that’s surprised you?

[0:32:16] HS: Fantastic question. Very close to home, they have been trying to switch me off of teams onto Slack for a full quarter before I did it, because as an individual, you know how you are most efficient and recruiters don’t adopt technology, because leadership says they should, if that’s right or wrong. They adopt it, because it makes their job easier immediately. If the tool saves them two hours in the first week, they’re going to use it forever. If it requires a ton of training, a ton of behavioral change, extra or repetitive data entry before they even see value, adoption is going to drop off very quickly.

The real key is making the product not only valuable, but super intuitive to the end user. We address this twofold by having a very usable end-to-end tool set within a super accessible user interface.

[0:33:16] DF: Is there a training and learning curve, or is this something you can hand to a recruiter and they’re like, “Oh, I get it and I can use it”? Do you think it’s at that stage?

[0:33:23] HS: I think it’s super adoptable. I think that there are ways in which our team ensures that it is adopted, which is that we do onboarding sessions that are driven by the individuals and help them actually spin up specific searches, so that we know that they are seeing an immediate impact. It is a very intuitive interface that we have mirrored off of the best different products that have historically existed. It has to look intuitive and usable and recognizable to historic products that just don’t have very good results, but we understand the flow. We do that by having a great customer success and customer support workflow that continually exists during trial, during onboarding and into perpetuity.

[0:34:10] DF: Awesome. Sticking on this topic, Holly. For the agency owner who’s tried to roll out new tech, watch their team ignore it, what did you do differently, or what have you guys done differently? It sounds like the UI/UX of it is a key component. The success side of it is key. Any other specific things that you think are making it easier to adopt?

[0:34:26] HS: Yes. I think the mistake that I see most often is trying to change everybody at once. You need to start with your top producer. If your best recruiter adopts something and suddenly, they can manage so many more searches than others, are closing more deals, the rest of the team is going to notice and adopt that. Recruiters, more than listening to what management says they need to do, they trust pure success more than management. We have to get the product into the hands of the top producers via trials, via pilots, which we’ve done and they become ambassadors and champions for the product.

We also separately have an amazing ambassador program in the space that Chris and I have launched. We start with high-impact individuals that the market trusts and their candid feedback. If they like the product, they’ll –

[0:35:19] DF: Bring it in.

[0:35:20] HS: Get to follow.

[0:35:20] DF: That’s great. Jumping over to you, Chris. Somebody who’s listening to this podcast, I know that you contribute a lot of podcasts and information as well. What AI sourcing materials, content if you’re not looking to go buy AI tools right now, but you’re looking to learn. Any recommended spots that you would suggest people go to? I know for myself, I’m currently obsessed with it and subscribed to and paying for Nate Jones’s Substack and podcast, and I think he has some of the most impactful boots on the ground, upfront AI content. But curious if you have any sources, or areas that you’re digging in and learning.

[0:35:57] CS: Yeah. I mean, I’m going to put a plug in for probably the biggest AI podcast out there, and as far as downloads and listener and viewers, and that’s Super Data Science Podcast, which is hosted by John Krohn, who happened to be one of the founders of Nebula. John is quite incredible in what he knows and does. His podcast is sponsored by Dell for a reason. People have seen his, more than likely seen his videos. But if you haven’t, every time I’ve ever turned somebody over to his podcast that loves AI, I always comes back and goes, “Holy, wow. It is a great listen.” He has some unbelievable guests on there, to say the least. Yeah, if you’re looking to learn, I mean, go to the one that right now is number one after.

[0:36:45] DF: I will say, I teed you up on that, because you’ve recommended it to me. I went and listened to it and immediately was like, this is incredible. If you like nerding out on data science stuff, it’s good. It’s a good one to check. Chris, what’s the biggest waste of money that you see agencies making on technology right now? I mean, that’s the last, and we’re still in the tech stack race. Everybody’s trying to go figure out, what’s the right tech stack? How do I get them to technology? It can be so impactful when done right, but where do you see people wasting money?

[0:37:15] CS: That is a great question. I don’t act like I have all the answers at all. You got to evaluate your technology and all your tools and vendors. I think it’s not just technology in itself. You see it sometimes. I look at a lot of products that have really come out with some great ways to improve certain parts of the experience, or the process. Depending on your strategy, if it is part of your key strategy for the year, or your road map, I’m not saying you don’t do that and make the investment, even if it doesn’t get you a big return on it. I don’t want to tell people, don’t do that. You also have to look at and go, what is giving me my biggest ROI? Two of the biggest things in staffing is very simple. You need candidate flow, and you need job flow. You need to have plenty of jobs to fill, and you need to get the candidates to them.

There’s a lot of tools that help you in between spaces, and once again, it’s worth looking at. What’s going to make some of your biggest impacts is just making sure that you’re increasing both of those and connecting the dots along the way. A tool that helps one person in your company save some time, versus a tool that is helping, let’s say, all your recruiters save X time, you’re going to get better ROI on the one that’s making the largest impact to the company.

[0:38:33] DF: Absolutely. Let’s zoom out one more time here and the last question I got for you, Chris, is like, two years from now, what does the agency that’s figured this all out look like, compared to the one that hasn’t?

[0:38:46] CS: That’s a great question. We all wish we had that crystal ball right now.

[0:38:50] DF: Yeah, crystal ball. What exactly?

[0:38:54] CS: It’s like you said before, who knows for sure? Maybe there’s this room that we go into and the robots are hiring everybody and we’re going like, “I can’t wait to work for this robot brand company.”

[0:39:05] DF: We send our robots to the interview, right? You send the robot.

[0:39:10] CS: Well, who knows? Maybe that’s where it’s at. But I really don’t think so. Once again, I’m going to circle it all back to, we got to make our recruiters better, and faster, and smarter at what they do. We just need to make sure we’re enabling them to make the best experience. We also, I don’t want to just say it’s all on them, where it’s at. The ones that figure out that total experience. In a great executive round table and listening to a lot of different industries and the different pain points they have, and we’re trying to help this person learn how to get this degree, or this next certification. In my head, I’m going, is that on your website? Did you take the question away from them? Are you taking the time to make the experience that if somebody is on your website at night, they found your brand, you did all the work to get them there, can they figure out everything they need to without talking to you?

Don’t want to say, you don’t – adopt the technology, look at it, but make sure it’s truly driving that experience. Because if your experience on the candidate side is the best it possibly can be, and you’re taking that same experience and you’re giving the same experience to your clients and then your internal employees, those companies are going to win. That’s where you need it to. Whether you’re using technology, or using the people inside your four walls, I think as long as you’re focusing on that, you’re going to be looking pretty good two years from now.

[0:40:41] DF: That’s great advice. With that, we’re going to go ahead and wrap up. Any closing comments from either of you for the audience?

[0:40:48] HS: I would say that to echo a lot of the pieces that we touched on, AI can be a very overwhelming, convoluted space, but it doesn’t have to be. It is super important to understand it, understand – trace it back to the LLM it’s built off, really understand the feature set and decide intelligently what you’re going to implement into your stack. I think connecting to what Chris said, I don’t think AI is going to undermine the need for agencies in a few years from now, but I genuinely do think that agencies that know how to leverage AI will always be agencies that don’t. I think there will be a distinct shift in the market and it’s already happening.

A tool like Nebula, that does make individuals four times more efficient is going to separate different players in the market. I can recommend personally and professionally from being 10 years in the industry that it is super important to get a tool of this type integrated into the stack, and I highly recommend that it is a human-led, tech-enabled product that does automate these manual tasks, but still keeps human elements when people stands.

[0:42:09] DF: Love it. How about you, Chris?

[0:42:11] CS: I think Holly nailed it right.

[0:42:13] DF: Nailed it?

[0:42:14] CS: Yeah. I think the one thing I’ll add is, I get extremely excited talking about Nebula. The reason is recruiters have changed. You get to sit back and watch how recruitment is changing. One of the ways was it really was a lot more head hunting before. Due to technology and advancements and disruption, now a lot of the leads are coming to you on a silver platter. Now you’re paying premium dollars for that, but they’re coming to you. While recruiters go like, “I want those just to come to me and I’m going to call them. They really want this job.” That’s great.

What I love about Nebula is your recruiters want that. They don’t want to adopt something and find out they wasted all this time and they didn’t feel like they made any money off of it. They didn’t get wins. Here you have this tool that is human trained. You get to tell it what you wanted to find for you. You get to play around with it. You get to tell it how you wanted to talk to them. Then once you do it, it goes out and it keeps looking for people and adding more and more and more, and then it automatically outreaches for you based on what you told it to do. It is like training a whole bunch of assistants. In the amount of time I’ve had employees come up to me and say, “I wish I had an assistant where I could run this much of my data.” Well, guess what? We got the assistant. It is incredible and it’s better than, in that case or what it does, it’s better than what a human can do. You can’t do that in what it’s doing.

It’s quite incredible and it tees you up these amazing leads, and then you get to do your best thing, the human side of it and get them and tell why they should come work for us. It is pretty incredible. I’m excited to say the least when I get to show it off.

[0:43:59] DF: Thank you guys so much for being on the show. Some really great insights and diving deep into the AI and what the future might look like. Appreciate the conversation.

[0:44:07] HS: Yeah, David. Thank you for having us. If anyone listening is going to SIA, Chris and myself will be there, and would love to show people the product in person, or chat more. Feel free to reach out.

[0:44:19] DF: Awesome.

[0:44:20] CS: Thanks, David.

[0:44:21] DF: Thank you, guys.