Talent Decides Where Capital Goes

Why the AI boom is really a recruiting problem, and what we built to solve it

By Brano Vargic, Founder and CEO, Talent Square

The boom nobody can staff

The AI boom will not be won by the best algorithm. It will be won by whoever can find a welder.

Underneath the software is a construction boom. Data centers, power grids, chip plants. Tens of thousands of electricians, pipefitters, and HVAC techs to build them. The chips are not the bottleneck. The people are. Deloitte and The Manufacturing Institute put 2.1 million manufacturing jobs unfilled by 2030. It now takes longer to fill an HVAC role than a software developer role, a reversal almost no one saw coming.

The workers exist. Almost no one knows how to reach them. I know, because two years ago I was one of the people who could not.

Capital has noticed. BlackRock, Ford, Google, and Carhartt just launched the Alliance for America’s Skilled Trades, pledging hundreds of millions to train the workforce. Good. But training builds supply over years, and the demand is now. You cannot train your way out of a shorta   ge this quarter. The question no one funds is simpler. How do you reach the workers who already exist, fast, at reasonable cost, and without a single outside specialist?

Every side of the table

As recently as 2024 I ran a boutique manufacturing company here in the States. I could not find a welder to save my life. That was not my first seat. In twenty-five years I have been a chief human resources officer, a chief operating officer running digital transformations, and a manufacturing CEO. Now I build software.

My career never looked like a ladder. It looked like a jungle gym. That is the only reason I saw this gap. You understand a job best once you have sat on the opposite side of the fence. I have run the HR department. I have run the factory floor. Not many people in my field can say both.

Why no one built the fix

In five years of quarterly budget meetings as a CHRO, I never once saw an HR request crack the top ten IT priorities. Not once. So when software got cheap and everyone piled into HR tech, they automated the visible parts of recruiting. The applicants. The profiles. The databases. They left the hardest part untouched. Reaching the people who were never in the system at all.

The workforce that went undercover

That is what happened to the frontline. As recruiting optimized for the desk, the skilled workforce moved into 1099 contracts and one-person shops. Roughly 70 percent of them never apply on a job board and have no LinkedIn profile. Every scraper and AI sourcing tool fights over the same visible minority. The majority stays invisible. Recruiting stopped being a search problem. It became a distribution problem.

Ask yourself. If you run a plant or a fleet, when did your best welder or driver last find you through a job board? If you place workers for a living, how much of your week goes to chasing people who were never going to raise their hand? I have sat in both chairs. The answer is the same. We are all fishing where the fish used to be.

I did not reason my way there. I measured it. My team analyzed more than 3,000 Meta ads recruiting truck drivers. Nearly all shouted the same thing. Cents per mile. So I asked drivers what actually mattered. Not the money. Everyone offered the same money. What moved them was concrete. Getting paid on time. Full loads both ways instead of driving home empty. More nights in their own bed. Two layers decide everything. Reach gets in front of people who are not looking. Message wakes them up and moves them to act.

Why is this suddenly solvable? Three things moved at once. The frontline workforce scattered into 1099 work and stopped using the front door. The channels where they actually spend attention, Facebook, Instagram, soon TikTok and the models, matured into precise, cheap, self-serve advertising. And the AI to run a campaign well, without a specialist, finally got good enough for anyone to use. None of this was true five years ago. All of it is true now. The shortage did not create the opening. The collision of those three did. The window belongs to whoever moves while everyone else is still arguing about whether AI is ruining hiring.

The marketing department talent acquisition never had

So we built it. Talent Square is the marketing department that talent acquisition never had. Recruiting became marketing years ago. Most of the field never noticed. If reaching people is a marketing problem now, why does almost no one run it like a marketer would?

Our software does the two things a marketing department does. Job Vacancy campaigns fill the role in front of you. Employer Branding campaigns build demand with the people you will need next, because people move on the seventh time they see you, not the first.

Simple enough to run yourself

I am obsessed with simplicity. The platform is self-service by design. You upload a role, set a budget, and in about ten minutes it launches, runs, and optimizes a campaign across Facebook and Instagram. No agency. No media specialist. No marketing team. The AI writes the creative for a few dollars. The campaign tunes itself. An AI layer screens the ads for hate speech. Compliance runs in the background. Every candidate lands in your own account, so you own the pool. And we measure it like e-commerce. Impressions, clicks, conversions, cost per candidate. Recruiting becomes a funnel you can finally see.

Simplicity has a second payoff most people miss. If one campaign takes ten minutes, an agency can run a hundred. What used to need a media specialist per client now runs across a whole client book from one account. Simple is not just for the owner between deliveries. It is leverage for the agency placing for fifty of them.

What it replaces

The difference lands best in a story I still tell with a grin. A manufacturing CEO showed me a proposal from a staffing agency and asked, half-joking, if Talent Square could beat it. I laughed. They wanted three weeks and over 3,000 dollars up front. That included a paid workshop to define the “persona” of a warehouse worker, and 1,400 dollars for a company video. Then a monthly retainer plus media.

His company had no HR department. So he ran Talent Square himself. Live in forty-eight hours. AI visuals for a few dollars. A company video he shot on his own iPhone. At ten dollars a day, about 300 for the month, the campaign surfaced 25 warehouse workers in three weeks.

The pattern holds. Another manufacturer spent about 300 dollars and hired two welders from eight qualified applicants. A traditional placement fee would have run 6,000. The enemy is not agencies. It is the old, expensive, slow way. And it is on borrowed time.

Why agencies win, not lose

Here is the part I want every agency owner to hear. The math looks like it should scare you. It should not. A few hundred dollars can do what a six-thousand-dollar placement used to. Does that make agencies obsolete? It does the opposite.

Talent Square is agency-first and white-label. One account runs your entire client book. You mark up the media the way you always have. Your client sees leads, not your cost. The candidates are yours. The margin is yours. A new revenue line, not a new competitor. The agencies that win the next few years will not be the ones with the biggest rolodex. They will be the ones who run fifty client campaigns from a laptop and keep

the spread on every one.

We earned that model the hard way. More than 700 frontline campaigns by hand over five years, before we wrote a line of the product.

The message, and the machine reading it

Reach is only half of a marketing department. The other half is the message, and whether anyone can still read it. That is where I spend my own study time now. I call it AI discoverability. I benchmarked entire banking and telecom sectors and found employers saying two things at once. A polished, marketing-built career site. A stack of dry, HR-written job posts that do not match. It matters more every month, because candidates now ask AI what a company is like to work for. The machine reads both faces and surfaces the contradiction. In my benchmarks, almost none of the employers were even legible to the AI summarizing them. Ask it the same question three times and you get three answers. The employer controls none of them.

Here is what that looks like in practice. A regional employer spends real money on a career site that promises growth, autonomy, and a great team. Its actual job posts, written by HR in another building, read like legal disclaimers. A person might not notice. The model notices instantly. It reads both faces, cannot reconcile them, and hedges. So when a candidate asks what this company is really like to work for, the answer comes back vague, generic, or quietly negative, and the employer never sees it happen. Reach puts you in front of the right person. Discoverability decides what the machine says about you once they go looking. Most employers are losing that second race without knowing it started.

Where I think this goes

So we build ahead of the channel. TikTok joins Facebook and Instagram within weeks. And we are building the infrastructure to place our ads inside the large language models themselves, the next place a candidate will ask who is hiring near me.

I am not guessing about that. I teach marketing at Grand Valley State University, and my students are my testing ground. More than half of them told me they now start a job search in ChatGPT, not Google or Indeed. That is not a forecast. That is this year’s graduating class. We are not in the AI game. We are in the placement game. Same engine, new surface.

Access for everyone

The through-line is democratization. Reach like this used to belong to big brands with big budgets. We hand it to the small manufacturer, the regional agency, the operator who could never afford a media team. When a hyperscaler picks where to build next, what does it weigh first now? The tax break, or whether it can staff the build? Workforce availability has become a top-tier factor in that decision, and it is climbing fast. Whoever can field and keep a frontline workforce wins the projects, and the capital that follows.

Which brings me back to the welder. In 2024 I was the one who could not find him, staring at an open job and an empty inbox. We built the tool I wish I had had. The welders were out there the whole time. So were the drivers, the electricians, the nurses. I just had no way to reach them. Neither did anyone else. That was never a talent shortage. It was an access problem wearing a shortage’s clothes. Talent is everywhere. Access is not.

If my mom can run it

Simplicity is not a slogan. It is a test. I run it on the person who most needs the product to be simple. My mother. She is 78, a retired English teacher, not remotely technical. If she can run a campaign, anyone can. If she cannot, we go back to the drawing board and make it simpler, until she can. That is the whole mission in miniature. Access for everyone, not for a selected few.

Meet the leader behind the success of Talent Square

Brano Vargic, Founder and CEO

Brano Vargic has spent twenty-five years on every side of business. He has served as a chief human resources officer, run operations and digital transformations as a chief operating officer across banking, consumer finance, and insurance in Europe and Southeast Asia, and led a US manufacturing company as CEO. He has built companies and sold them. Today he is Founder and CEO of Talent Square, the Grand Rapids, Michigan platform reaching the frontline workforce the last two decades of recruiting technology left behind, and a visiting professor of marketing at Grand Valley State University. He studied at Cornell University’s ILR School, Erasmus University Rotterdam, Netherlands, and IESE Business School in Barcelona. Slovak, married to an American, and a father of three, he does his clearest thinking on long morning walks with Charlie, his English Shepherd and most patient listener, and still measures every product decision by one test: could his 78-year-old mother use it?

“Talent is everywhere. Access is not. That is the gap we exist to close.”