For most of the last century, companies solved talent problems the same way. Someone left or a new need emerged, so you posted a job, interviewed candidates, and hired the best one. Simple. Repeatable. And for a long time, it worked.
It does not work as cleanly anymore.
The problem is not that hiring is broken. The problem is that the nature of the skills gap has changed. Skills now depreciate on a timeline that human capital planning was never designed for.
For most of the last century, companies solved talent problems the same way. Someone left or a new need emerged, so you posted a job, interviewed candidates, and hired the best one. Simple. Repeatable. And for a long time, it worked.
It does not work as cleanly anymore.
The problem is not that hiring is broken. The problem is that the nature of the skills gap has changed. Skills now depreciate on a timeline that human capital planning was never designed for. The World Economic Forum estimates that 39% of core job skills will shift by 2030, not disappear entirely, but shift enough that a hire you make today for a specific technical profile could be working with an outdated toolkit within three years. Meanwhile, 94% of C-suite leaders say they are already facing shortages in AI-critical roles. The gap is not ahead of us. It is here now.
This is exactly why a different framework has taken hold in serious HR circles: Build, Borrow, Bot, or Buy. It is a way of forcing a more precise question before you default to posting a job. Instead of asking who do we need, you start by asking what kind of problem do we actually have, and then match the solution to that.
Build: The Patient Strategy That Pays Off
Building means developing your existing people. That sounds obvious, but most companies underinvest in it badly, usually because the timeline feels uncomfortable. Training someone takes months. Posting a job takes a week.
What gets missed in that reasoning is what you actually get on the back end. An internally developed employee brings something an external hire never can: they already know how your organization thinks. They know which processes are documented and which ones live in someone’s head. They know why the last three initiatives failed. That context is not transferable. You cannot hire it.
The argument for building has also gotten stronger as automation accelerates. When a meaningful chunk of any given role’s tasks will look different in three years, developing people for adaptability beats hiring people for a specific current skillset. You want a workforce that can keep learning, not one that was technically perfect on the day they joined.
Building requires lead time you do not always have, and it requires leadership that actually commits to it rather than treating L&D as a line item to cut. Companies that build well treat learning as infrastructure.
Borrow: Speed Without the Strings
Borrowing means bringing in talent on a contingent basis: freelancers, contractors, project consultants, interim executives. The workforce platform market has grown substantially over the past decade, and what you can access on a project basis now would have required a full-time hire five years ago.
The value of borrowing is precision. You are paying for a specific outcome over a specific window, not betting on someone’s five-year trajectory at your company. If you need a machine learning engineer to work through a data pipeline project, or a financial controller to cover a restructuring, or a subject matter expert for a product launch, borrowing lets you get the right person at the right time without the full cost of permanent employment.
The limitation is equally straightforward. Contractors have other clients. They are not building a career inside your organization, which means they are unlikely to go out of their way to fix problems they were not explicitly hired to fix. For roles that require real continuity: deep institutional knowledge, long-term relationships, cumulative judgment, bringing in contractors creates gaps that are hard to see until they matter.
Bot: The One Everyone Is Getting Wrong
Automation has absorbed a huge portion of what used to require dedicated human headcount. Invoicing, reporting, scheduling, routine customer communications, data entry, quality checks, etc. The list of tasks that software now handles is long and growing. One global survey found that 92% of senior executives reported up to 20% workforce overcapacity in legacy roles, almost entirely because automation had absorbed tasks those roles used to own.
Where organizations consistently go wrong is treating Bot as a headcount reduction strategy and stopping there. The more accurate way to think about it is: Bot reshapes roles. When you automate the repeatable parts of a job, the human doing that job does not disappear, they shift toward the parts that actually require judgment, relationship management, creative thinking, and oversight. If you do not redesign the role around what the person does after the automation handles the routine, you end up with confused employees who do not know what their job is anymore.
Agentic AI, systems that can reason across multiple steps and execute longer-horizon tasks, is pushing this further. The question is no longer just which tasks can be automated. It is which parts of a role are genuinely, irreducibly human, and which parts just felt human because we had never built anything good enough to do them.
Buy: Still Right, But Not by Default
Hiring externally is still often the right answer. When you are building something entirely new: a function that did not exist before, a capability your organization has never had, a market you have never operated in, an external hire brings more than skills. They bring perspective you cannot grow internally because you have no internal vantage point from which to grow it.
The risk of defaulting to Buy, which is what most organizations do, is that it papers over problems that sourcing cannot fix. If you keep hiring for specific technical profiles in a market where those profiles have a three-year shelf life, you are running an expensive cycle.
The better frame for Buy in 2026 is not which skills does this person have, but how fast can they learn, how do they handle ambiguity, and how do they perform when the roadmap shifts. That has always been true of great hiring.
The Combination Is the Strategy
No organization should be running on a single one of these. The companies doing this well are using all four simultaneously, mapped to different roles, different timelines, and different business objectives.
You Build the AI literacy your entire workforce needs over the next two years. You Borrow the specialist you need for next quarter’s product launch. You Bot the customer intake process that does not require human judgment. You Buy the person who will lead the function you are standing up from scratch.
Don’t Buy because that is what HR has always done. Don’t skip Building because the ROI is harder to quantify in a quarterly review. Do not resist Bot because it feels threatening rather than clarifying.
The framework is not complicated. The execution is.
The workforce question has changed. It used to be: who should we hire? The better question today is: what kind of capability do we need, how long do we need it, and what is the right way to get it? Sometimes that is a new employee. Sometimes it is a software system. And sometimes it is the person who has been sitting three desks away for four years, waiting for someone to invest in them.