Total Talent Is Still a Pipe Dream

I have been listening to people talk about Total Talent for more than a decade. It was a regular topic at CWS conferences, and I sat through countless presentations and discussions about breaking down the barriers between permanent and contingent talent, getting HR and Procurement to work together, connecting MSP and RPO, and eventually creating a workforce strategy that started with the work rather than the employment status of the person doing it.

I bought into the idea, and I still do. There is something fundamentally sensible about an organisation looking at the work it needs to get done, understanding the capability required and only then deciding how best to access it. Yet it is 2026, Staffing Industry Analysts was already discussing Total Talent Management in 2015, and I still don't think we have really cracked it.[1] That is a long time for an idea that has always sounded so intuitively right to remain something we are still trying to operationalise.

That bothers me because we certainly cannot blame a lack of effort or technology. Over those years we have seen increasingly capable VMS and ATS platforms, talent CRMs, MSP and RPO combinations, Direct Sourcing, talent marketplaces, skills platforms and countless attempts to connect one part of the workforce ecosystem with another. The components became more sophisticated and the language around them evolved, but the basic conversation at industry conferences remained remarkably familiar.

One experience in particular has stayed with me because it showed how quickly an idea that seems blindingly obvious can run into problems that have very little to do with whether the idea itself is good. It also taught me that getting everyone to agree on the ambition is not necessarily the hard part.

When everybody wants the same thing

In February 2018, I was involved in a workshop that brought Finance, HR, Legal, Procurement, IT and Operations into the same room to talk about the digital future of talent. We presented a broader approach to accessing capability and eventually arrived at Direct Sourcing, creating a community of people who might be interested in project work with the organisation.

The potential sources of talent were already around us. Line managers regularly received unsolicited approaches from people they could not employ at that moment but might want to work with later. Permanent recruitment produced silver medallists who had gone deep into a selection process and narrowly missed out. Contractors came to the end of assignments and disappeared back into the market despite having spent months or years learning the organisation. Alumni and retirees were another obvious population. Instead of repeatedly starting from scratch, why not retain those relationships and give people a place where they could see project opportunities and express an interest in them?

The response in the room was overwhelmingly positive. Line management in particular immediately saw the practical value. There would finally be somewhere useful to send good people who approached them, and somewhere to look when a new project required capability. As the discussion developed, the question became less about whether we should do it and more about why we weren't doing it already.

The answer turned out to be rather complicated. Procurement could see the savings potential immediately, because every suitable person engaged directly rather than through a traditional staffing channel potentially removed agency margin. But Procurement traditionally did not have a budget for building and curating talent communities. Its remit was to save money, not to spend money creating a capability that would generate savings somewhere down the road.

HR had a different problem. These people were not necessarily employees or even candidates for permanent employment, which made ownership difficult. Legal quite rightly raised questions about an organisation actively recruiting people who would ultimately be employed or subcontracted through a third party, particularly in a market with specific recruitment and labour-leasing regulation. IT could enable the technology but had no reason to own the proposition, while Operations wanted the outcome rather than responsibility for the infrastructure behind it.

What struck me was that nobody needed to be difficult for us to have a problem. Everyone could support the idea while being entirely rational about why it did not belong to them, and the idea could make perfect sense at enterprise level while fitting rather badly into the organisation responsible for delivering it.

We eventually found a rather clever way around the funding issue. The existing contingent workforce programme generated volume rebates when certain spend thresholds were reached, so those rebates were used to fund the Direct Sourcing capability. In effect, value generated by the existing programme financed an attempt to build the next iteration of it. We got past the budget problem and the programme went ahead, only to discover that finding the money had not resolved the more fundamental question of where this thing actually belonged.

The people did not fit our categories

For the community to become useful, it needed a steady flow of good people. Permanent recruitment was an obvious source. A strong candidate who had narrowly missed a permanent position might be very interested in a six-month project, provided they actively chose to enter the externally managed community and understood what that meant. Conversely, someone who had spent a year successfully delivering a project might eventually be an excellent permanent hire. Alumni and retirees might be interested in either, depending on what was happening in their lives.

The people themselves did not necessarily think in the categories we had created for them. Someone might want permanent employment at one point in their career, independent project work at another and permanent employment again later. The categories mattered enormously to the organisation because they determined budgets, processes, systems, legal relationships and ownership. They mattered rather less to the person deciding whether the work was interesting.

At the same time, the permanent Talent Acquisition organisation was building its own talent community. To me, the obvious opportunity was for the two to feed each other. Permanent candidates who opted in could become visible for project work, while proven project talent could become visible for permanent opportunities. Both communities would become richer as a result, without pretending that the contractual and regulatory distinctions between them did not exist.

It didn't really happen, and even the intended link from the permanent careers environment into the externally managed project community proved difficult to get across the line. Looking back, I find that small detail more revealing than many of the larger debates we had. The technical connection between two talent ecosystems could have been achieved with a hyperlink, while the organisational connection was considerably harder.

There was also another group around the table, even when they were not physically in the room. Staffing agencies had invested in finding much of the contingent talent already working inside the organisation. If one of their contractors finished an assignment, entered the Direct Sourcing community and was subsequently re-engaged without the agency, the organisation might reasonably view that as retaining known talent and eliminating unnecessary margin. The agency might equally reasonably ask why its sourcing investment was being used to disintermediate it.

That tension extended to the MSP. Vendor neutrality is an important part of the MSP proposition, yet the same MSP was now involved in creating a sourcing channel whose success could reduce business flowing through the staffing suppliers it was expected to manage fairly. The enterprise economics made sense, but the interests of everyone participating in those economics did not necessarily line up.

None of this required a villain, which is why I find the usual language about “breaking down silos” slightly unsatisfying. HR, Procurement, Legal, the MSP, the agencies and the business could all have perfectly rational positions. The problem was that an idea designed around the needs of the enterprise crossed boundaries that had been created for perfectly legitimate reasons.

Getting closer exposed something else

Several years later, I saw another model that came much closer to what we had been talking about. HR, an MSP and a third-party RPO cooperated around a Blended Talent approach, and people could move much more freely between permanent and contingent populations. It addressed one of the things that had frustrated me earlier because talent no longer had to remain trapped in whichever channel had originally brought it into the organisation.

It worked, but it also exposed another weakness. Permanent recruitment increasingly found candidates in the contingent population rather than sourcing them independently. From the customer's perspective, that was difficult to argue with. A vacancy existed, a suitable person was already known somewhere in the wider talent ecosystem, and the organisation got the person it needed. That was, after all, supposed to be the point.

The problem was that the contingent pool had not magically populated itself. The MSP and its underlying staffing supply chain had invested the effort and economics required to put those people there. As permanent recruitment increasingly drew from that population, the argument shifted from whether talent should be allowed to move between different parts of the ecosystem to who was responsible for replenishing the ecosystem in the first place.

We had solved part of the access problem without solving the incentives underneath it. The customer still got people, at least to a point, but too much energy ended up being spent resolving conflict between parties whose commercial responsibilities had never become as blended as the talent model supposedly was.

My experience was hardly unique. Everest Group research published in 2024 found that half of organisations surveyed were still exploring ways to combine contingent and permanent talent approaches, while another 27% considered it a possibility as their organisations matured.[2] After all the years of discussion, technology investment and new workforce models, that feels remarkably early in the journey.

That experience made me even more cautious about the way we talked about Total Talent. We had spent years treating integration as though it were mainly a matter of connecting things: MSP with RPO, ATS with VMS, permanent with contingent, internal talent with external talent. Yet every time another connection became possible, something underneath it seemed to resist being connected quite so easily.

Starting one step earlier

There is another part of the equation that I think gets lost when Total Talent becomes a discussion about permanent versus contingent labour. Before looking for the right person, there is a more basic question about the work itself and whether the organisation should be looking for a person at all.

Some capability should be built. If it is strategically important, repeatedly required and central to how the organisation operates, developing or hiring that capability internally may make sense. Other capability is better borrowed because the requirement is temporary, specialist or uncertain. Then there is work where neither building nor borrowing makes much sense because an external provider can simply deliver the outcome.

Build, borrow or buy is not a particularly complicated idea, but it changes where the demand conversation begins. A genuine SOW or outsourced service can be exactly the right answer when the organisation wants to buy an outcome rather than assemble and manage the individual people required to produce it. If somebody else already has the capability, methodology and delivery infrastructure, there may be very little sense in recreating it internally.

The complication is that policy has a habit of changing behaviour without necessarily changing demand. Tighten the controls around contingent labour sufficiently and the work does not disappear. Sometimes it simply moves. People who would previously have been visible as contingent workers can reappear inside SOWs, consulting engagements or subcontracted work orders, even when the underlying arrangement still looks remarkably like buying individual labour.

That can create the slightly absurd situation where an organisation believes it has reduced contingent labour while much of the same work is still being performed by external people, now sitting in another spend category and potentially with less visibility. The classification has improved according to one policy measure while the enterprise outcome may have become worse.

This sits slightly upstream of Total Talent rather than changing the concept itself. Getting the build, borrow or buy decision right simply means that the demand enters the talent ecosystem through the right door. Getting it wrong means we may already be trying to optimise the wrong solution before we even start deciding where the capability should come from.

Perhaps that should not surprise us either. HR has employees and permanent candidates. Procurement has external spend and suppliers. Talent Acquisition has requisitions. An MSP has contingent demand and a supplier ecosystem. An RPO has permanent hiring outcomes. Services Procurement has SOW and consulting spend. Each operates within its own economics, accountabilities and measures of success, while the manager somewhere in the business usually has a much simpler starting point: there is work that needs to get done.

If Total Talent really starts from that perspective, the concept still makes enormous sense to me. Understand the work first, decide whether the capability should be built, borrowed or whether the outcome should simply be bought, and then determine the best route to whatever talent is actually required. The problem is not the logic. The problem is everything that has to happen around it.

Better technology, same organisation

The workforce industry has spent years building increasingly sophisticated solutions to individual parts of this problem. Direct Sourcing matured considerably. MSP and RPO models evolved. Talent marketplaces and skills platforms added new ways of finding and understanding capability. Services Procurement technology became more sophisticated. The underlying ambition became broader, but the organisation around it remained largely divided into functions, budgets and systems designed for particular types of talent and particular types of spend.

Perhaps the most telling thing is that the language has barely changed. As recently as 2026, Total Talent commentary is still discussing the need to break down the traditional silos between HR and Procurement.[3] Those are essentially the same organisational boundaries we were wrestling with years earlier.

That is why I have become wary of the idea that the next technology will finally make Total Talent happen. We have already been through several generations of the next technology, and each one improved something without making the underlying organisational questions disappear.

Inevitably, the conversation has now moved to AI. There is no question that AI can improve important parts of the equation. It can infer skills, interpret requirements, identify adjacent capability, search enormous populations and make connections between people and work that would be impossible to replicate manually. If one of the historic problems of Total Talent was simply seeing all the capability available to an organisation, AI could become extraordinarily useful.

For AI to do that properly, however, it needs the data, and not just CVs. It would ideally want to understand employees and their skills, aspirations and project histories, people interested in internal mobility, permanent applicants and silver medallists, contingent workers, previous contractors, alumni, retirees, Direct Sourcing communities, RPO populations and perhaps talent sitting inside staffing supplier databases. The richer the information and the more populations it can see, the better its ability to make the connections Total Talent has always promised.

That sounds wonderful from the perspective of workforce optimisation. From the perspective of European data protection, it becomes complicated very quickly. GDPR requires personal data to be processed lawfully and transparently, for specified purposes and with appropriate data minimisation, while organisations must be able to demonstrate compliance with those principles.[4] Purpose limitation, transparency and accountability are not peripheral considerations when the entire proposition depends on information about people moving between different systems, organisations and employment contexts.

We were already dealing with a very simple version of that problem in 2018. If a silver medallist from permanent recruitment might be interested in joining an externally managed project community, we could not simply move their details across because it seemed sensible. The person needed a clear invitation to opt in and needed to understand that they were entering something different, managed by a third party.

Now extend that journey over several years. The same person could enter as a permanent applicant, opt into a project community, become a contractor, finish an assignment, return for another project and eventually become a permanent employee. Along the way, different systems and organisations may touch their information, new experience is added to the profile and AI may begin inferring skills or suitability that the individual never explicitly provided in the first place.

From the perspective of Total Talent, that is exactly the kind of fluidity we have spent years trying to create. From the perspective of data governance, the flow has to remain anything but casual. Where the information came from, why it was collected, what the person understood, who may use it and for what purpose still needs to be clear enough to withstand scrutiny.

AI does not somehow sit outside those requirements. The European Data Protection Board has made clear that the development and deployment of AI models involving personal data remains subject to GDPR, including questions around the lawful processing and subsequent use of that data.[5] The more ambitious we become about what AI should see across the workforce, the more demanding the data question becomes, not less.

This is where I struggle with the suggestion that AI will finally solve Total Talent. It will almost certainly make some parts of it considerably better, perhaps dramatically so. The conceptual process of matching capability to work can be almost obvious, but the compliant movement and reuse of the information required to make that happen is anything but.

Still the same conversation

What troubles me most is how familiar all of this feels. We were talking about Total Talent in 2015. In 2018 I sat in a room where almost everyone could see the value of a much more connected approach to talent, yet we struggled with ownership, funding, regulation and competing organisational priorities. Years later, I worked with a model that removed some of those barriers and watched different conflicts emerge underneath them.

Now it is 2026 and we have technology more capable than anything we could realistically have imagined when those conversations started. We can search, infer, match and analyse at extraordinary scale, but many of the difficult questions are still not technology questions at all. Who owns the capability? Who funds it? Who gets the saving? Who carries the cost of sourcing? What happens when one participant's efficiency becomes another participant's lost revenue? How does the talent move without the commercial model fighting back, and how does the data follow without losing the clarity and accountability that privacy law requires?

There is also the question sitting right at the front of the process that we sometimes skip in our enthusiasm to optimise talent. Should we be finding people to do this work in the first place, or should we simply be buying the outcome? Even when we answer that correctly, we still need the organisation behind it to behave consistently enough that work does not migrate between categories simply because one route has become administratively or politically easier than another.

I still believe in the ambition behind Total Talent. Starting with the work, deciding what capability is actually required and then determining the best way of accessing it makes too much sense to dismiss. There are undoubtedly organisations doing meaningful parts of it very well, and perhaps somebody will eventually bring all of this together in a way that genuinely works at scale.

Until then, I struggle to call Total Talent anything other than a pipe dream. Not because the idea is wrong, and certainly not because the technology is missing, but because after more than a decade we still have not worked out how to make ownership, economics, incentives, regulation and data move as freely as we would like the talent to. AI may make the matching smarter and the individual components better, but it does not remove any of those constraints, and in the case of data it may simply make one of the hardest ones more obvious.

TL;DR

Total Talent still makes perfect sense on a whiteboard. The problem starts when you try to operate it. Decide whether to build, borrow or buy. Look across every available source of talent. Let people move between permanent and contingent work. Align HR, Procurement, MSP, RPO and Services Procurement. Make the commercial incentives work. Keep suppliers onside. Stay compliant. Maintain a watertight trail of who agreed to what use of their data, where it came from and where it went. Then give AI enough access to make intelligent decisions without breaking any of the above.

We have been trying to make those pieces fit for more than a decade. AI will make some of them better. It does not make the contradictions disappear. That is why, for now, Total Talent remains a pipe dream.

References

[1] Staffing Industry Analysts, “Total talent: Past, present and the path forward”, 18 June 2025. A retrospective on the development of Total Talent, looking back at the industry's discussions and research beginning in 2015.
Staffing Industry Analysts: Total talent: Past, present and the path forward

[2] Everest Group, Total Talent Acquisition research, 2024. Research referenced in Smart Trends Moving Talent and Business Forward, including organisations' approaches to bringing permanent and contingent talent together.
Everest Group research report

[3] AMS, “Total Talent Management: The Future Is Now - Who's in Charge?”, 28 April 2026. Contemporary discussion of Total Talent and the continuing organisational divide between HR and Procurement.
AMS: Total Talent Management

[4] European Data Protection Board, “Basic principles”. GDPR principles including lawfulness, fairness and transparency, purpose limitation, data minimisation and accountability.
European Data Protection Board: Basic principles

[5] European Data Protection Board, Opinion 28/2024, 17 December 2024. Opinion 28/2024 on certain data protection aspects related to the processing of personal data in the context of AI models.
EDPB Opinion 28/2024

Contact

Reach out for tailored support across services procurement, external workforce strategies, workflow optimisation and SaaS evaluation.

© 2026. All rights reserved.