The next chapter for HR is not about earning a seat at the table. It is about proving what HR changes once it is there.
For years, the HR profession has talked about becoming a strategic partner. I think that ambition has expired.
Not because strategic partnership is unimportant. Quite the opposite. It has become the minimum standard.
We have moved from employee-first thinking through operating-model redesign, journey and service design, knowledge management, agile operations, AI, agentic automation, modern architecture, personalization and the increasingly important boundary between machine speed and human judgement.
The more interesting question now is this:
“What exactly should HR do with all the capacity technology is supposed to give back?”
That question matters because the technology has moved on dramatically. Stanford’s 2026 AI Index reports that 88% of organizations were using AI in at least one business function in 2025 and 70% were using generative AI, yet deployment of AI agents remained in single digits across almost every business function. Meanwhile, Gartner predicts that more than 40% of agentic AI projects will be cancelled by the end of 2027 because of escalating cost, unclear business value or inadequate controls (Stanford HAI).
That is a useful reality check. We are simultaneously further into AI adoption than many HR operating models acknowledge and much earlier in genuinely autonomous work than the hype suggests. And this is where I believe the next HR transformation will be won or lost.
AI in itself doesn’t make HR strategic, but it does remove some of HR’s excuses for not being strategic.
If agents answer routine questions, initiate workflows, gather information, draft communications, route work and complete low-judgement tasks, HR should need fewer people spending their days moving information from one place to another. But releasing capacity is not the same as creating value:
The strategic test is what happens next.
That is the journey from provider to partner, and it requires HR to stop measuring the work it processes and start measuring the capacity it creates.
There is a line I would happily retire from the HR vocabulary:
“HR needs a seat at the table.”
A seat is furniture. It tells us nothing about influence. You can attend every executive meeting and still add little more than workforce commentary after the real decisions have been made.
You can also run an operationally excellent HR function and still fail to influence how the organization responds to AI, demographics, skills scarcity, management capacity or changing employee expectations.
The better question is:
“Did HR materially change the decision?”
That is a much harder standard. Consider a business function planning to introduce AI and reduce its operating cost by 20%.
Provider HR waits for the new organisation chart, updates roles, launches consultation, changes contracts, supports exits and eventually recruits the new skills requested.
Partner HR enters much earlier:
That is partnership. Not HR as a friendlier service desk. Not HR as an internal consultant waiting to be briefed. Not HR as the conscience that arrives after a decision.
HR as a participant in designing the economics and humanity of work.
This also means confronting an uncomfortable truth. The basic operational work does not become unimportant just because HR wants to become strategic. The opposite is true: a strategic HR function with unreliable payroll, inaccessible policies, poor case handling or confused employee journeys has a credibility problem. You cannot build strategic influence on top of operational distrust.
Think of HR service delivery as a stack:
The mistake has been treating those layers as a choice: operational HR versus strategic HR; efficiency versus experience; technology versus humanity; centralization versus personalization; automation versus people.
Most of those are false choices. A genuinely mature HR function has to do the basics extraordinarily well so that people do not need HR very often, while becoming exceptionally good when human expertise really is required.
That is especially important as AI spreads through the workforce. The International Labor Organization’s 2025 task-level analysis found that around one in four workers globally are in occupations with some exposure to generative AI, but concluded that transformation of jobs is more likely than wholesale replacement; only 3.3% of global employment falls into its highest exposure category (International Labor Organization).
In other words, the biggest HR problem is unlikely to be managing one enormous wave of redundancies but rather managing millions of smaller changes in what people actually do.
Traditional HR machinery is poorly designed for that. Job descriptions change slowly. Skills frameworks get refreshed periodically. Organizational design tends to happen in projects. Training is often offered after a capability gap appears. Workforce planning is still frequently headcount planning wearing a smarter jacket.
The emerging reality is different: tasks can change monthly, skills can depreciate quickly, and AI can alter the economics of one part of a role without touching the rest. Somebody who appears to occupy a “declining job” may possess exactly the judgement, customer knowledge or domain expertise needed in a growing one. The strategic unit of HR is therefore shifting:
That is a bigger transformation than automating HR.
HR teams are being asked to do more than ever. Everest's Tech Provider Spotlight explores how HR Service Delivery is evolving. Read Now.
There are at least four forces that make this new definition of partnership urgent.
The first is the sheer speed of skills change. The World Economic Forum’s Future of Jobs Report 2025 found employers expected 39% of workers’ existing core skills to be transformed or become outdated by 2030. In its 100-worker thought experiment, 59 people would need training by the end of the decade; 29 could be upskilled in their current roles, another 19 would need upskilling and redeployment, while 11 risk not receiving the training they need. Skill gaps were the most frequently cited obstacle to business transformation, named by 63% of employers (World Economic Forum).
That is a learning-and-development problem, sure. Perhaps more importantly It’ss a business-allocation problem. Who learns what? How early? For which future work? At whose cost? And what happens to the work they stop doing while they learn? Those are strategic choices.
The second force is demographics. Across the OECD, the working-age population is projected to shrink by 8% by 2060, while the ratio of people aged 65 and over to working-age people is projected to rise from 31% in 2023 to 52%. Without stronger participation, productivity and other responses, the OECD estimates annual GDP-per-capita growth could slow from roughly 1% to 0.6% (OECD).
This changes the automation conversation. In aging economies, doing more with fewer available workers is not necessarily an anti-employee agenda. It may become a basic condition of growth. The question is whether organizations achieve it by burning out a shrinking workforce or by intelligently removing low-value work, widening participation, retaining experience and increasing what each person can accomplish. That should put HR right in the middle of productivity strategy.
The third force is management capacity. Gallup’s State of the Global Workplace 2026 reports global employee engagement fell to 20% in 2025, while manager engagement fell to 22%. Gallup’s long-running research also finds managers account for at least 70% of the variance in engagement between teams (Gallup). At the same time, the CIPD’s 2025 UK Good Work Index found only 60% of managers said they had the training and information needed to manage well, and only 59% said they had enough time (CIPD).
We are giving managers more distributed teams, more well-being responsibility, more complex employment expectations, more organizational change, more data, more technology, more AI-enabled employees and potentially wider spans of control. And then we wonder why management is becoming a bottleneck.
This is one of the biggest opportunities for HR partnership: manager time may be one of the most valuable pools of capacity HR can release. Not HR time – manager time. Every unnecessary approval, badly explained policy, duplicate form, avoidable employee query, clumsy workflow and HR process that requires managerial translation consumes a little of it. Individually, these irritations look trivial. At scale, they become an organizational tax.
The fourth force is one I think has received far too little attention: AI may break the traditional apprenticeship model of white-collar work. For generations, people learned judgement partly by doing lower-risk, lower-complexity work first. Junior lawyers reviewed documents. Analysts assembled reports. Recruiters screened applications. HR advisers researched policies and drafted first responses. Managers learned by handling relatively ordinary people issues before facing the extraordinary ones.
Some of that work is exactly what AI is good at removing. PwC’s 2026 analysis of more than one billion job advertisements found that skills in the most AI-exposed roles are changing more than twice as quickly as in the least exposed roles. More strikingly, its US analysis found the most AI-exposed entry-level roles were seven times more likely to demand capabilities traditionally associated with senior work, including judgement and leadership (PwC).
That should set off alarm bells in HR because there is an obvious paradox: What happens when we automate the work through which people used to learn the work that cannot be automated? If AI writes the first draft, when does the graduate learn to write? If AI conducts the first analysis, how does the analyst learn what good analysis looks like? If an agent handles every straightforward employee case, where does the future HR expert learn the pattern recognition required for the difficult cases?
We cannot respond by keeping pointless work alive as a training exercise and neither can we assume judgement appears magically at age 11. This requires an entirely new learning architecture: simulation, supervised practice, rotational exposure, AI-assisted critique, shadowing, explicit apprenticeship in judgement, staged decision rights and better feedback.
That is strategic HR work. And it is exactly the kind of issue that becomes visible only when HR stops thinking about technology implementation and starts thinking about organisational capability.
I find it useful to think about the workforce through three currencies.
Capacity. Capability. Confidence.
Together they form what I call the Three-Currency Model of HR Partnership.
You can express the idea as a deliberately simple diagnostic:
Workforce value ≈ Capacity × Capability × Confidence
It is not meant to be an exact accounting formula but rather make one important point: the relationship is multiplicative, not additive. Have brilliant people with no capacity and little gets done. Create capacity without capability and you get spare time but not better performance. Build capability without confidence and employees may resist the tools, distrust the change or leave. Build confidence without sufficient capability and good intentions fail.
Capacity means time, attention, cognitive bandwidth and organisational room to act. AI makes this particularly interesting because many business cases start here.
Fine. But saved from whom? To do what? Those two questions should appear in every automation business case.
Consider the difference:
Technology case: “The AI assistant saves managers 30 minutes a week.”
versus:
Workforce case: “The AI assistant releases 27,600 manager hours a year, of which we intend to redirect 40% into coaching, workforce planning and customer-facing leadership.”
Those are not the same proposition. The second one has a destination, and without a destination it’s largely theoretical.
Capability is not the same as “skills training”. It is the organisation's ability to do the work its strategy will require. That may involve technical skills, but the evidence increasingly suggests AI also raises the value of distinctively human capabilities. PwC’s 2026 Jobs Barometer found that roles in which AI amplifies professional expertise were growing faster and experiencing stronger wage growth than roles where AI mainly makes existing work easier to perform. It also found specialist AI roles growing far faster than the wider jobs market, alongside increasing demand for judgement, creativity and leadership (PwC).
Capability is not the same as “skills training”. It is the organization's ability to do the work its strategy will require. That may involve technical skills, but the evidence increasingly suggests AI also raises the value of distinctively human capabilities. PwC’s 2026 Jobs Barometer found that roles in which AI amplifies professional expertise were growing faster and experiencing stronger wage growth than roles where AI mainly makes existing work easier to perform. It also found specialist AI roles growing far faster than the wider jobs market, alongside increasing demand for judgement, creativity and leadership (PwC) Capability is not the same as “skills training”. It is the organization's ability to do the work its strategy will require. That may involve technical skills, but the evidence increasingly suggests AI also raises the value of distinctively human capabilities. PwC’s 2026 Jobs Barometer found that roles in which AI amplifies professional expertise were growing faster and experiencing stronger wage growth than roles where AI mainly makes existing work easier to perform. It also found specialist AI roles growing far faster than the wider jobs market, alongside increasing demand for judgement, creativity and leadership (PwC).
This is why “AI skills” is far too narrow a workforce strategy. Some employees need to understand how to use AI. Far more need to become good at the things that sit around AI: framing problems, challenging outputs, recognizing exceptions, making trade-offs, communicating difficult decisions, building trust, combining incomplete evidence, and knowing when not to automate.
The third currency is confidence. I deliberately prefer this to “engagement” or “trust” alone. Confidence asks whether people believe they can safely participate in the new operating model:
The CIPD found in early 2025 that 63% of surveyed workers trusted AI to inform important workplace decisions but not make them; only 1% said they trusted AI to make those important decisions (CIPD). Those numbers will evolve as tools improve, but the principle will not; a technically excellent transformation can be organizationally unusable if people do not have confidence in it.
That is why Applaud’s “People 1st, Tech 2nd” philosophy still matters in an agentic era: technology should remove noise and unnecessary work without allowing the human consequences of automation to become an afterthought.
This is where partnership becomes concrete. When a business leader says “we need 15% more productivity” HR should not instinctively answer with a recruitment plan, a skills programme or an automation proposal. It should ask:
That conversation is very different from traditional HR planning… and potentially far more valuable.
The most dangerous metric in an AI business case may be “hours saved” with no destination attached. I would ban it from executive papers unless it is followed by a second line: “Hours saved will be used for: __.”. There are actually three completely different kinds of value hiding inside most productivity claims.
Hard value is money genuinely removed or avoided: an external contract cancelled, overtime reduced, recruitment avoided because capacity can absorb growth, or a role genuinely no longer required.
Capacity value is time returned to people. This is valuable, but it is not automatically cash. If an employee saves three hours and spends three hours doing something else, the company has not put three hours of salary back in the bank.
Strategic value is what the organization achieves because that capacity has been redirected: more customers served, faster innovation, better management, shorter time-to-productivity, fewer errors, reduced risk, faster redeployment or stronger retention of critical talent.
Those distinctions matter enormously. Otherwise we end up adding imaginary savings from 25 different AI initiatives and wondering why the P&L has not moved.
This is what I use with an executive team.
The crucial step is in the middle: “Name its destination.” Many transformation programs are relatively sophisticated above that line and remarkably vague below it. They know precisely what technology will do. They know much less precisely what humans will do instead. That needs to reverse.
Imagine an organization with 10,000 employees and 1,200 managers. Suppose better HR knowledge, simpler workflows and agent-assisted administration genuinely remove 30 minutes of HR-related administration from each manager’s average working week for 46 weeks of the year. That returns 27,600 manager hours annually. At an illustrative fully loaded managerial cost of $60 an hour, someone could very easily write $1.656 million of productivity value on a PowerPoint slide.
I would not. Not yet. Because unless managers work fewer paid hours, the organization has not saved $1.656 million in cash, it has acquired $1.656 million-worth of optionality. That is potentially more interesting.
Suppose the organization explicitly converts 40% of those recovered hours (11,040 hours) into structured coaching, team development, workforce planning and customer-facing leadership. Now we can test an outcome.
The original automation has now become part of a causal chain rather than a standalone saving. That is what HR should be measuring.
One metric I would add to the CHRO dashboard is:
Capacity Conversion Rate = released capacity demonstrably redirected to an agreed priority ÷ total capacity released
If HR reports 100,000 hours “saved” but only 8,000 can be connected to changed work, the organization has a measurement issue.
Over time I would track capacity in three populations separately:
| Capacity pool | What to measure | What good reinvestment might look like |
| Employees | Time previously lost navigating work and HR friction | Productive work, learning, customer activity |
| Managers | Administration and avoidable people-process effort removed | Coaching, performance, planning, change leadership |
| HR | Repetitive service, coordination and information work removed | Workforce design, complex cases, manager support, skills and organizational change |
This changes the strategic conversation around HR service delivery. A knowledge article is no longer successful because somebody viewed it. Self-service is not successful because a ticket did not get created. An agent is not successful because it “contained” a conversation. The question becomes: “What human capacity did this service create, and where did that capacity go?”
That does not mean abandoning operational metrics. Accuracy, resolution, effort, repeat contact, case cost, service quality and employee satisfaction remain essential. It means connecting them to something bigger.
Modern HR service platforms can already provide much of the raw material: search failures, knowledge gaps, journey drop-off, case volume and cost, AI resolution, escalation, agent cost, workflow performance and employee interaction funnels. Applaud is one example of a system designed to bring those signals together across knowledge, AI, journeys and cases.
But no platform can make the strategic leap for you. The dashboard can tell you capacity was released; leadership must decide what the capacity is for.
For every material automation initiative, I would require a one-page agreement between the business, HR and finance before implementation.
It should answer five questions.
Then review it six months later. This sounds obvious. It is surprisingly rare. And it is the difference between installing AI and redesigning an organization.
We should be careful with corporate AI case studies. The market is full of heroic productivity claims, self-reported savings and pilots whose long-term economics remain unclear. Stanford’s 2026 data show widespread AI use but much more limited agent deployment, while Gartner’s cancellation forecast is a reminder that impressive demonstrations do not guarantee scalable business value (Stanford HAI).
There are, however, some useful stories precisely because they illustrate different choices about human capacity.
One of my favourite examples predates the latest agentic-AI boom but captures the principle beautifully. Ingka Group, the largest IKEA retailer, said in 2023 that its Billie chatbot had handled 47% of call-centre queries over the preceding two years. During the same period, it had trained 8,500 call-centre workers to become interior-design advisers. Its people leader said the company was not then seeing AI lead to headcount reduction and emphasised employability, lifelong learning and reskilling (The Korea Times).
The new remote customer-meeting channel, covering products and services sold via phone and video, generated €1.3 billion in IKEA’s 2022 financial year, equivalent to 3.3% of Ingka revenue, and the company said it wanted that share to reach 10% by 2028 (The Economic Times).
We should not make a causal claim that the chatbot “created” €1.3 billion. The remote-design business involved many factors. But the strategic choice is what matters. IKEA could have defined automation success as: “fewer people required to answer customer questions.” Instead, it also asked: “what more valuable customer problem can these people now solve?” That is the Capacity Conversion Loop in practice.
Automation absorbed repetitive demand. Human capability moved closer to judgement, creativity, relationship and revenue. There is a lesson for HR service delivery here. If an HR agent resolves 50% of routine contacts, the exciting question is not how many HR roles can disappear. It is what the HR team can now do that employees and the business were previously not receiving.
Klarna provides a useful counterweight. Its AI customer-service assistant produced extraordinary early numbers. In 2024 the company said the assistant had engaged with millions of customers and was expected to generate $40 million in annualised savings; its initial launch was widely reported as handling around two-thirds of service conversations and doing work equivalent to hundreds of full-time agents (Klarna).
Then the story became more nuanced. In 2025, chief executive Sebastian Siemiatkowski acknowledged that cost had been given too much weight in the customer-support model and that this could lead to lower quality. Klarna moved to strengthen access to human service rather than abandon AI, and its current customer-support model continues to offer both an AI assistant and human telephone support (TechCrunch).
That is not evidence that AI failed but that single-objective optimization fails. If you tell a transformation to optimize cost, it will optimize cost. If you tell it to optimize containment, it will contain. If you incentivize deflection, it will deflect. But organizations do not actually want the cheapest interaction. They want the best economics of the whole relationship, and that includes cost.
It also includes quality, loyalty, trust, successful resolution, brand and the ability to recognize when a human conversation is worth more than the saving generated by avoiding one. HR leaders should pay particular attention because many employee interactions carry far higher emotional and personal stakes than ordinary customer service.
Do not mistake the disappearance of human effort for the creation of value.
There is another uncomfortable side to this transition, and HR should not sanitise it. In February 2025, Singapore’s DBS said it expected around 4,000 temporary and contract roles across its markets not to be renewed over the following three years as AI took on more work. Permanent staff were not part of the announced reduction, which the bank expected to occur as temporary roles naturally rolled off. At the same time, DBS expected to create around 1,000 AI-related positions (Reuters).
The bank also said it had identified about 13,000 employees for upskilling or reskilling, with more than 10,000 already beginning AI- and data-related training (The Straits Times). Outgoing chief executive Piyush Gupta captured the difficulty unusually honestly: “For the first time, I’m struggling to create jobs.” (Reuters)
I respect that honesty. Not every automation story can end with every person moving neatly into a better-paid creative role. Some work will shrink. Some skills will lose market value. Some organizations will require fewer people in particular areas.
The International Labor Organization's evidence suggests job transformation will be more widespread than full replacement, but that does not mean displacement disappears for individuals (International Labor Organization).
A credible employee-first HR philosophy cannot promise that nobody will ever lose a job. It can promise something more meaningful:
That is partnership too. Sometimes HR’s strategic contribution is not making a difficult decision disappear. It is helping the business make it intelligently, honestly and humanely.
So what would I actually do as a CHRO or Head of HR Service Delivery preparing for this next phase? Not launch another transformation program. Not create a “Future of Work Center of Excellence”. And definitely not begin by buying more AI (though if you don’t already have Applaud, you probably should).
I would introduce a relatively simple operating discipline:
Ask the executive team: “Where is the organization currently unable to perform because of a people constraint?”
That might be:
Pick three. Those become HR priorities. Not employee experience, AI transformation, or future skills. Specific constraints. The closer HR gets to the actual constraint on business performance, the easier strategic partnership becomes.
For each priority area, stop looking only at organisational charts and map the work.
The critical addition is this: “which tasks are developmental?” Before automating junior work, ask whether it currently functions as part of the organization’s apprenticeship system. If it does, design another route to mastery first. That may mean scenario-based learning, supervised case review, AI-generated practice cases, job rotations, shadowing or explicit coaching.
Do not accidentally automate the bottom rung of the career ladder and then complain that nobody is ready for the top.
Apply the capacity reinvestment contract. Do not approve a significant automation initiative that merely promises “efficiency”. Name the people, hours, destination and result. And distinguish hard financial saving from released capacity. Finance should love this because it makes AI business cases more honest. HR should love it because it makes workforce consequences visible before deployment. Employees should ultimately benefit because it prevents efficiency from defaulting silently into workload intensification.
One of the dirty secrets of productivity programmes is that time saved often gets filled immediately with more work. That may sometimes be appropriate, but it should be a decision, not an accident.
The WEF expects 50% of employers to transition people from declining to growing roles as part of their response to labor-market change, while 85% plan to prioritize workforce upskilling (World Economic Forum). The important word is transition.
A catalogue of courses is not a transition system. A genuine transition pathway looks more like this:
That requires HR, learning, workforce planning, recruiters and business leaders to work as one system. It also requires honesty about distance. An accounts administrator is not automatically six weeks of e-learning away from becoming an AI engineer. But they may be much closer to an operations analyst, customer adviser, process specialist or AI quality role than a job-title comparison suggests.
This is why skills should increasingly be treated as a redeployment language, not merely a learning taxonomy.
I would make manager capacity a board-level people metric. Not manager satisfaction with HR. Actual manager capacity.
Gallup’s evidence on the relationship between manager engagement and team performance makes the leverage obvious, while the CIPD data show many managers do not feel sufficiently equipped or have enough time to manage well (Gallup).
This is where agentic HR could eventually deliver enormous value. Managers need less organizational drag. A good HR service architecture should make ordinary people management easier, reduce the number of times managers become human middleware between employees and systems, and reserve their judgement for the things that genuinely require it.
If HR gives every manager back an hour but does not help them become a better manager with it, we have only completed half the transformation.
HR and Finance should agree in advance what counts as:
Do not combine them into one heroic ROI number. A useful value ledger might look like this:
| Value type | Example | Can it go into cash ROI immediately? |
| Hard saving | Outsourced service removed | Yes |
| Avoided cost | Growth handled without five planned hires | Usually, if the hiring plan was real |
| Capacity release | Managers save 30 minutes a week | No, track separately |
| Productivity outcome | Same workforce processes 12% more volume | Yes, with a credible baseline |
| Workforce outcome | Critical-role attrition falls | Translate carefully into economic value |
| Experience outcome | Employee effort falls | Strategic indicator; do not invent cash without evidence |
| Risk outcome | Materially fewer payroll/compliance errors | Quantify using expected risk/cost where defensible |
This will make some business cases look smaller. Good! A smaller number you can defend is more strategically useful than a $12 million “saving” nobody can find later.
Finally, I would replace the usual aspiration to “be a strategic partner” with a small partnership scorecard This is not a survey asking executives whether they like HR. It’s real evidence.
Track five things:
That last measure is difficult. It should be. Strategic partnership is supposed to be difficult. A CHRO should be able to point to decisions and say:
“We entered that market differently because of our workforce analysis.”
“We redesigned that function rather than making the original number of redundancies.”
“We slowed that automation because the capability pipeline would have collapsed.”
“We accelerated this one because the employee effort data showed the existing process was destroying thousands of hours.”
“We invested in manager capacity because it was constraining growth.”
“We built internally rather than bought externally, saving recruitment cost and retaining institutional knowledge.”
That is evidence of partnership (not merely meeting attendance).
Learn how to transform HR into a people-first function that builds trust, designs better experiences, and drives real business results in this interactive, 10-minute guide. Read Now.
There is a temptation to finish with a prediction.
I am increasingly wary of predictions: the last few years should have made all of us more humble. AI capabilities will improve. Some agentic systems will become remarkably useful. Others will fail. New job categories will appear. Some old ones will shrink. Employee expectations will change. Regulation will evolve. Operating models will be redrawn again.
The exact shape of HR in 2030 is unknowable. But I think its test is remarkably stable:
Does HR help people and the organization succeed together?
That sounds almost too simple. It is not. There will be moments when employee and employer interests genuinely diverge. There will be restructures, performance problems, difficult investigations, automation decisions, pay constraints, job losses, AI errors. And trade-offs between individual preference and organizational need.
Employee-first HR does not mean employees get everything they want and strategic partnership does not mean HR simply gives executives everything they ask for. The role is harder than either of those caricatures; HR sits at the point where business economics meet human reality. That is why this profession matters. And paradoxically, the more transactions machines can perform, the more visible that responsibility becomes.
When an employee no longer needs an HR adviser to explain a holiday policy, submit a simple request, find a payslip, update an address or start an ordinary workflow, we have not made HR less important. We have removed some of the work that obscured what HR was for.
The same is true inside the HR function: if AI can create the first draft, synthesise the data, surface the anomaly, route the request, monitor the workflow and answer the standard question, then the human contribution becomes easier to see:
That is part of organizational performance. The labor market already gives us hints of the direction. AI exposure is rising, yet the most recent research points towards widespread job transformation rather than simple mass replacement. Employers expect rapid skills change. Aging economies need more productivity from scarcer labor. Managers are under strain. And the most AI-exposed work is beginning to put a premium on judgement and leadership earlier in careers (International Labor Organization).
So perhaps the final evolution is not really provider to partner. Perhaps it is:
Provider → Partner → Architect of organizational capacity
The provider gives people what they ask for. The partner helps leaders solve problems. The capacity architect goes further. They continuously examine where human effort is being wasted, where capability will be needed next, where technology can expand what people can do, where trust is fragile, where work should be redesigned and where the organization is quietly consuming the very capacity its strategy depends upon.
They understand that capacity without capability is wasted, capability without confidence is stranded, and confidence without performance is unsustainable. And they understand one other thing: the point of eliminating transactional HR was never to eliminate transactions. It was to create room for something better.
That is the promise technology has been making HR for decades. Shared services promised it. Self-service promised it. Cloud HR promised it. Automation promised it. Generative AI promised it. Agentic AI promises it again.
This time, HR should hold itself accountable for where the time goes.
Call it strategic when it converts that escape into something the organization and its people can actually feel: better work; better decisions; better managers; faster adaptation; more opportunity; less wasted human effort; and a workplace in which increasingly intelligent technology gives people more room to be intelligent themselves.
The future of HR service delivery is not ultimately about service delivery; it is about what great service delivery makes possible. An employee who can get help without fighting the organization. A manager who has time to manage. An HR professional who can spend more of their day on questions that genuinely deserve professional judgement. A business that can adapt without treating people as inventory.
And an organization capable of improving continuously because it listens to what work actually feels like from the employee’s point of view, rather than simply admiring how elegantly the process appears from the center. That employee-led principle has been the thread running through Applaud’s own philosophy of designing HR technology from the employee’s point of view.
There will always be another technology, another operating model, another workforce trend, another reason to transform. The challenge for HR is not to predict all of them but instead build a function capable of adapting to them without losing sight of why it exists.
People first. Technology second. Value as the consequence of doing both well.
And perhaps that is the real meaning of partnership: not standing between the employee and the business, but helping each succeed because of the other.
How Applaud Helps You Make It Happen
At Applaud, we believe employees are a company’s most important customers. That’s why our technology is built entirely from the employee’s point of view—delivering more human, intuitive, and rewarding HR experiences that empower HR teams to do more for their people.
If you’re ready to turn employee-first HR from vision to reality, we’re here to help. Get in touch to see how Applaud can transform your HR Service Delivery and create a workplace where employees truly thrive.
Duncan Casemore is Co-Founder and CTO of Applaud, an award-winning HR platform built entirely around employees. Formerly at Oracle and a global HR consultant, Duncan is known for championing more human, intuitive HR tech. Regularly featured in top publications, he collaborates with thought leaders like Josh Bersin, speaks at major events, and continues to help organizations create truly people-first workplaces.