Talent Advisor 2.0

A few months ago, I wrote an article called The Rise of the Talent Advisor in the Age of AI.

The argument was that Artificial Intelligence would increasingly absorb the transactional work that has occupied so much of recruiting for decades.

Sourcing. Outreach. Résumé review. Scheduling. Screening. Workflow coordination. Reporting. Analytics.

As more of that work becomes automated or AI-assisted, I argued that the recruiter has an opportunity to move upstream.

The future recruiter becomes less of a process administrator and more of a Talent Advisor.

A workforce strategist. A market analyst. An operations consultant. A business partner. An AI-enabled talent architect.

I still believe all of that.

But after spending the last several months looking more closely at AI, manufacturing, infrastructure, supply chains, labor markets and Talent Intelligence, I think I stopped the argument too early.

Because the next evolution of the Talent Advisor isn't simply becoming more strategic once hiring begins.

It's learning to see the hiring need before the requisition exists.

That is what I mean by Talent Advisor 2.0.

Talent Advisor 1.0 helps the business hire. Talent Advisor 2.0 helps the business see what's coming.

The question I've been working on since then is how to make that happen in practice. Giving a recruiter a chatbot doesn't automatically move Talent upstream. The intelligence has to connect to the way the business plans, the way recruiting works, and what the organization remembers.

The requisition is late

Most recruiting organizations still begin their work with a requisition.

The business identifies a need. Headcount gets approved. A job description gets written. A requisition opens. Then Talent gets involved.

That process made sense when recruiting was largely an execution function. But consider everything that may have happened before that requisition ever appeared. A manufacturer accumulated backlog. A company won a major government contract. A business announced a new facility. A competitor expanded into another market. A supply-chain bottleneck appeared. A company acquired a supplier for a critical manufacturing capability. A government funded a strategic industry. A manufacturer reshored production. Automation changed the skills required on the factory floor. A new technology entered the business. Energy infrastructure expanded. Capital moved. Those were all business signals.

But viewed through the right lens, they were also talent signals.

The requisition may simply be the last one.

From recruiting intelligence to business intelligence

Talent Intelligence traditionally asks useful questions:

Where does the talent exist? What does it cost? Who employs it? How scarce is it? What markets should we recruit from? What competitors are hiring?

Those questions still matter.

Talent Advisor 2.0 adds another layer.

What is happening in the business?

What capabilities will those changes require? When will they be required? How confident are we that the demand will actually materialize? Do those capabilities already exist internally? Can they be developed? Should they be acquired?

And if the organization eventually needs to hire them, how difficult will those people be to find?

That changes the operating model. Talent Intelligence becomes the bridge between a business signal and a workforce decision. It connects what the business may need with what the labor market and the existing workforce can realistically provide.

Instead of waiting for demand and then reacting to it, Talent begins looking for the signals that create demand.

Follow the signal

Consider the kinds of signals a Talent Advisor can follow. A semiconductor company announces billions of dollars in capital investment. That's a talent signal.

An aerospace company acquires a specialized supplier because critical manufacturing capacity is constrained.

That's a talent signal.

A company opens a second manufacturing operation in another country to reduce supply-chain concentration.

That's a talent signal. An energy company plans additional power capacity to support growing data-center demand. That's a talent signal. A manufacturer installs more robotics and automation. That's a talent signal too.

The obvious headline in each case may be about capital, infrastructure, technology, supply chains or operations.

But underneath the headline is a simple question:

Who will know how to do the work?

That is where Talent Intelligence becomes much more interesting.

The Talent Supply Chain

Businesses have spent decades developing sophisticated systems for understanding physical supply chains.

Where do critical materials come from? Which suppliers are concentrated? What is the lead time? Where are the bottlenecks? What happens if supply disappears? How quickly can capacity be added? How resilient is the system?

We should be asking remarkably similar questions about skills.

Where does critical expertise live? How concentrated is it? How transferable is it? How long does it take to develop? How much of it exists internally? How much exists in the external market? What happens if demand suddenly increases?

People and skills have supply chains too. And workforce risk can become operational risk very quickly. A company may have the capital to build another plant. It may have the equipment. It may have the real estate. It may even have the customer demand.

But if the specialized expertise required to operate the facility is scarce, then the workforce becomes part of the constraint.

That is the Talent Supply Chain.

Not every signal deserves action

There is a problem with looking upstream.

Once you start watching the business environment this way, there are signals everywhere.

Factory announcements. Investment commitments. Projected job numbers. Technology announcements. Government incentives. Expansion plans. New markets. Acquisitions.

Not all of them happen.

And not all of them create meaningful hiring demand.

That means Talent Intelligence cannot simply become a larger collection of news and data.

It has to become better at judging probability.

That's why I have also been thinking about the idea of a Talent Intelligence Confidence Score.

A useful workforce signal should probably be evaluated across several dimensions:

Probability
How likely is the underlying business event to actually happen?

Timing
When would the workforce capability be required?

Workforce impact
How much demand could the event realistically create?

Talent scarcity
How difficult will the capability be to obtain?

Those questions turn a press release into something more useful.

I see the Confidence Score as a working framework, not a finished formula or a promise that we can predict the future. A number is only useful if we can explain the evidence behind it, what remains uncertain, and what would change our assessment.

Probability and urgency also need to stay separate. A highly likely expansion may have little impact on a broadly available skill. A less certain project could deserve early attention if it depends on expertise that takes years to develop. The score should help us choose whether to watch, investigate, prepare or act. It should not make that decision for us.

A new facility announcement might be interesting.

A fully funded, permitted facility already under construction in a market with limited skilled labor is much more actionable.

The goal isn't to predict everything. You can't.

The goal is to improve the probability that the Talent organization is looking in the right direction.

The Intelligence Feedback Loop

A Confidence Score gives us a starting point. Then we have to test it against what we learn in the field. An expansion announcement may suggest demand for a particular skill. Market research tells us where that skill exists. Candidate conversations tell us whether those people are available, interested and willing to make a move. Each step can change the assessment.

That's the Intelligence Feedback Loop. Business signals shape the questions we ask. Recruiting brings back evidence that can strengthen the signal, weaken it or change the workforce response. If candidates consistently tell us that the compensation won't work or the location is a barrier, we need to bring that information back to the business. The original score shouldn't remain untouched just because it looked convincing on a dashboard.

The loop works in the other direction too. A pattern across candidate conversations may expose a capability shortage before an expansion is announced. The Talent Advisor connects those observations, checks whether they hold up and updates what the organization knows. That is how upstream intelligence and day-to-day recruiting become part of the same work.

The conversation is part of the intelligence too

Candidate conversations are where we test a lot of those assumptions.

For decades, the ATS has primarily been a System of Record.

Candidate. Job. Stage. Activity. That matters. But the actual recruiting relationship often lives in conversations. Compensation changed. Relocation became a concern. A competing interview appeared. Candidate interest increased after meeting the hiring manager. A counteroffer became likely. One concern may determine whether the candidate accepts. That's why I think recruiting technology is evolving through another sequence:

System of Record to System of Engagement to System of Intelligence

The objective isn't simply to document that a conversation happened.

It is to understand what changed because of the conversation.

When candidate context becomes institutional knowledge instead of remaining scattered across phones, inboxes and recruiter memory, the organization becomes better at making decisions.

That is also Talent Advisor 2.0.

I think of this as Field Intelligence. Candidate conversations, sourcing observations and external market signals should become reusable institutional intelligence. A recruiter may learn that compensation expectations are moving, relocation is becoming harder, a competitor is recruiting aggressively, or candidates with a particular capability are harder to engage than the database suggests.

Candidate motivations matter here too. People may be willing to move for different work, a better manager, more stability or a clearer path to advancement. Those conversations can help explain workforce trends long before they show up in a report.

One conversation is an observation. Repeated observations across searches may indicate a pattern. The useful record preserves the date, the market, the source and the context, so the next person can distinguish a current signal from an old assumption. Recruiters need a practical way to capture that context, check an AI summary and make the relevant learning available to the next search.

That doesn't require circulating everyone's private story. We can retain the recruiting context appropriately and share the market learning without turning candidate trust into a data collection exercise. The point is to stop losing useful knowledge every time a conversation ends.

Institutional hiring memory

Systems that learn from historical hiring patterns and recruiter behavior offer another possibility.

Prior locations. Industries. Experience levels. Candidate profiles. Recruiter preferences. Previous hiring behavior.

That creates the possibility of something recruiting organizations have rarely had in a structured way:

institutional hiring memory.

That could be extraordinarily valuable.

Persistent organizational knowledge means the learning survives the individual search. The next recruiter should be able to understand why a sourcing market was chosen, which requirements were relaxed, where candidates dropped out, and what changed before a hire was made. A list of successful hires alone misses most of that story.

Recruiter behavior is part of the record too. Which profiles did we pursue? Which did we overlook? Where did a manager's preferences narrow the search? What did we assume about location or experience? A system can surface those patterns, but someone still has to ask whether they reflect success, habit or a blind spot.

A Talent function that remembers what worked, what failed, what candidates cared about, where hires came from and what patterns preceded successful hiring could become progressively smarter over time.

But it also raises an important question.

If a system learns from who an organization has historically hired, does it improve future decision-making?

Or does it simply become extremely efficient at repeating historical preferences and biases?

Talent Advisor 2.0 cannot outsource judgment to the system.

The advisor has to understand what the system has learned and decide whether that historical pattern still deserves to guide the future.

AI changes the role again

This is where AI becomes especially useful. It can help organize capital investment, earnings-call observations, backlog changes, facility plans, competitor activity and labor-market information alongside candidate conversations, hiring history and internal skills. No recruiter should have to process all of that manually.

But I think we're still underestimating the opportunity when we use AI mainly to write outreach, polish job descriptions or summarize a meeting. Those are useful tasks. They don't, on their own, change when Talent enters the business conversation.

The bigger opportunity is an AI-enabled Talent Acquisition operating model that connects Talent Intelligence, sourcing, recruiting workflows, automation, persistent organizational knowledge and human judgment. What we learn in one part of the process should improve what happens in the next.

Two of the biggest time killers are sourcing and scheduling. Automate the work that keeps recruiters from doing the work only they can do. In sourcing, that can mean building research lists, finding profiles against agreed criteria, organizing evidence and keeping the pipeline current. In scheduling, it can mean matching availability, coordinating interview panels, sending confirmations and reminders, and handling routine changes. The recruiter still sets the strategy, reviews whom we pursue and steps in when a candidate needs a person or an exception needs judgment. The time we get back should go into better conversations, stronger relationships and earlier advice to the business. If automation only helps us run the same reactive process faster, we've missed the bigger opportunity.

An external signal can prompt market research. That research can challenge a workforce assumption. Candidate conversations can refine the market picture. The results of a search can update the organization's memory. Automation can move the reviewed information into the workflow where somebody can use it. That is a different ambition from having a collection of disconnected AI chats.

Practical multi-agent work fits here. Different AI workflows can handle different assignments: monitoring business signals, researching talent markets, analyzing job requirements, organizing candidate evidence or updating approved knowledge. They can pass structured findings between them, with clear boundaries around what they may do and where a person reviews the work.

The useful part is the specialization and coordination. A Talent Advisor needs to know what question each workflow is answering, what evidence supports its output, and when to stop it or challenge it. Several agents repeating the same weak assumption won't make that assumption better.

Sourcing is a good example. Job requirements should lead to a target-market strategy: which capabilities matter, where people use them today, which companies or adjacent industries make sense, and which requirements may unnecessarily shrink the pool. AI can support that research, then organize candidate evidence against agreed criteria.

From there, a specialized workflow can assist with fit and risk assessment, identify gaps in the evidence, and help assemble a prioritized slate. The recruiter reviews the research and the proposed priorities before outreach begins or the slate goes into a hiring discussion. A profile cannot establish someone's motivation, willingness to relocate or ability to thrive with a particular manager. Those still require a conversation.

The workflow also needs to learn from the response. If the strongest prospects consistently decline because the compensation or location doesn't work, that is new market intelligence. It should reach the Talent Advisor and the business while there is still time to adjust.

Turning intelligence into a business decision

That brings me to the translation problem. We can collect more information than ever and still leave a business leader wondering what to do on Monday morning. The Talent Advisor has to translate AI and intelligence into business workflows and operating decisions.

For example, a signal about scarce technical expertise might become a recommendation to test an adjacent talent market, adjust the compensation plan, start an internal development program or change the timing of a facility ramp. The recommendation needs an owner, a next step and a point at which we revisit the evidence. Otherwise, we've produced an interesting report and left the operating model alone.

Here's a hypothetical example. A manufacturer plans a $100 million investment in an automated production line, with a proposed start in 18 months. The Talent Advisor asks Operations which capabilities the line will require, then tests the plan against internal skills and the external market. Suppose controls expertise is scarce, the internal development path takes two years, and the investment still depends on final approval. The recommendation could be to identify employees for development now, validate the compensation needed for an experienced controls hire and revisit the plan when the investment is approved. Capital investment becomes a workforce decision, with a timing problem leadership can address before the requisitions arrive.

Human judgment matters because information isn't the same as insight. The advisor has to understand the business context, know which signal matters, recognize a misleading pattern and decide when historical behavior should not be repeated.

In the original Talent Advisor 1.0 article, I wrote that future recruiters would not compete against AI. They would quarterback it.

I still mean that. AI doesn't replace the Talent Advisor in this model. The advisor becomes the orchestrator of specialized intelligence and automation, responsible for asking the right questions, checking the evidence and helping people make decisions. Quarterbacking means directing the work, checking the evidence and taking responsibility for the recommendation.

Talent Advisor 2.0 is really about what the quarterback should be looking at, and how to turn that view into action.

From requisition fulfillment to capability strategy

This may be the biggest shift.

Talent Acquisition has historically been measured by how effectively it fulfills hiring demand.

Time to fill. Cost per hire. Acceptance rate. Source of hire. Quality of hire.

All useful metrics. But imagine the value of a Talent organization that can tell leadership: We don't need to hire this capability today.

But the business is moving in a direction that suggests we may need it 18 months from now.

The external talent pool is limited. The skill takes two years to build internally. Two competitors are expanding into the same market. And the probability that this demand materializes is increasing. That conversation is fundamentally different from: “How quickly can you fill these 50 requisitions?” That's workforce strategy. That's Talent Intelligence. That's advisory work.

The next step is a workforce action. Leadership might choose to develop the skill internally, pursue an external hire, use a specialist partner, redesign the work or prepare a pipeline while the demand becomes clearer. The Talent Advisor helps compare the lead times and tradeoffs with the people accountable for the business plan.

If hiring is the answer, the requisition becomes the result of that work. It should arrive with a clearer capability requirement, a realistic view of the market, a sourcing strategy and an understanding of what could make the search fail.

In the example above, waiting for approved headcount could leave us eighteen months to solve a capability problem that takes two years to develop. Seeing the signal early gives the business time to choose a different response. It doesn't mean opening speculative requisitions every time an expansion makes the news.

Talent Advisor 1.0 vs. Talent Advisor 2.0

I don't see these as replacements for one another.

Talent Advisor 1.0 was about moving beyond transactional recruiting.

Understanding the business. Consulting with leaders. Bringing labor-market insight. Improving hiring decisions. Using AI and technology more effectively.

Talent Advisor 2.0 takes that same philosophy farther upstream.

Talent Advisor 1.0 helps the business hire.

Talent Advisor 2.0 helps the business see what's coming.

That means the role begins earlier. Before the requisition. Before the sourcing strategy. Sometimes before the workforce plan itself has fully formed. The advisor reads the business environment and asks: What capability will this eventually require? Where does that capability exist? How difficult will it be to obtain? When should we start preparing? And what happens if we wait?

Brett's Perspective

I don't think the future of Talent Acquisition is simply faster recruiting.

I think it is earlier intelligence.

The best Talent organizations will still need excellent recruiters.

They will still need people who know how to source, assess, close and build relationships.

But the highest-value work may increasingly happen before any of that begins.

The opportunity is to understand where the business is going and what workforce capability it will require to get there.

Then help leadership prepare before the need becomes urgent.

I've spent a lot of time watching demos, experimenting with AI tools and trying to connect the pieces around actual recruiting problems. I'm designing and prototyping these workflows; I am not claiming that a complete enterprise operating model is already solved.

What keeps pulling me back is how much recruiting learns and how little of that learning makes it into the next business decision. A conversation changes our understanding of the market. A difficult search exposes a workforce assumption. A business announcement hints at a capability we may need. Too often, those pieces remain separate.

I want a Talent function that can connect them, remember what it learned and bring a useful recommendation to the business. The technology helps make that possible. The Talent Advisor still has to earn the trust to act on it.

This is Part 2 of the Talent Advisor series. My focus here has been on how Talent Acquisition can move beyond reacting to hiring needs and begin anticipating them through better intelligence, stronger business partnerships and more informed workforce decisions.

In Part 3, I'll explore the technology behind that vision, including what I'm learning as I build and experiment with Jarvina, BlackForge and other AI-driven recruiting workflows.

But technology isn't the destination. It's what we do with that intelligence that matters.

That is what makes Talent a strategic function. Not the title. Not the technology. Not the dashboard.

The ability to recognize a workforce problem before everyone else realizes they have one.

The requisition may be the last talent signal you see.

The next generation of Talent Advisors should learn to read the ones that come before it.

And that, to me, is Talent Advisor 2.0.