What is an HR Service Delivery Model, and How AI is Changing It

Key takeaways
- An HR service delivery model organizes who handles each employee request and at what level of expertise using tiers that escalate from self-service through shared services to specialist centers.
- The traditional tiered structure works on paper but creates delays at handoff points, where context is lost and rebuilt manually at every step.
- Organizations without deep legacy infrastructure now have a credible option: bypass traditional tiers entirely and invest in AI-enabled resolution that preserves context from first contact to answer.
HR has run on some version of the same structural idea for nearly thirty years: a three-legged stool of shared services, centers of excellence, and business partners, first popularized by Dave Ulrich in 1996. The stool itself has held up remarkably well. What actually sits on each leg, and how a request moves between them, is what has changed considerably.
This guide covers what an HR service delivery model is, why the traditional version is reaching real limits, what a request moving through it actually looks like, where technology genuinely fits, what changes for a global organization, and how AI is changing what the model should look like going forward.
What is an HR Service Delivery Model?
An HR service delivery model is the structure an organization uses to decide who handles an HR request and at what level of expertise, rather than routing every question to the same generalist.
The Ulrich model, sometimes called the three-legged stool, divides HR into three roles:
- Shared services handle transactional and administrative work, like payroll changes, employment letters, and benefits enrollment.
- Centers of excellence provide deep expertise on specific domains like compensation design, talent management, or employee relations.
- HR business partners consult directly with business leaders on people strategy, organizational design, and workforce planning.
Underneath these three roles sits a tiered support structure. Employees start with self-service (knowledge bases, portals, FAQs). If self-service doesn't resolve the issue, the request escalates to shared services and, from there, to a specialist if deeper expertise is needed.
The Traditional Model and Why It's Reaching Its Limits
The traditional HR service delivery model works well on paper. Each tier has a clear purpose, and the logic of escalation is sound: handle simple requests cheaply and quickly, and reserve expensive expertise for the questions that genuinely need it.
The real strain shows up at the handoff points between tiers. The systems supporting each tier were rarely built to pass context along automatically. When a request moves from self-service to shared services, the shared services agent often has to reread information that the employee already provided. When shared services escalates to a specialist, the specialist frequently asks for details that already exist in the HRIS.
This produces delay rather than outright failure. Each team may be performing well within its own tier, but the request accumulates a multi-day lag purely from sequential handoffs. LACE Partners' research on HR shared services also identifies persistent issues in mature operating models.
What a Request Actually Looks Like Moving Through the Model

One of the clearest HR service delivery examples is a benefits eligibility question.
An employee relocates from one state or country to another and asks whether a domestic partner is eligible for benefits under the new jurisdiction. Here is what happens:
- The employee checks the self-service portal. It returns general benefits enrollment content but doesn't answer the specific combination of jurisdiction, plan type, and employment status.
- The employee submits a case. Shared services checks the employee's profile, enrollment window, and relevant policy content.
- Shared services cannot resolve the eligibility question from the available policy information, so the case is escalated to a benefits center-of-excellence specialist.
- The specialist interprets the policy and sends the answer back to shared services.
- Shared services communicates the decision and next steps to the employee.
Following HR service delivery best practices, a well-designed model can handle such requests in under a day. A poorly integrated one can take several business days as context is rebuilt at each step.
Where Technology Actually Fits Into the Model
Technology's role has historically been narrow—powering the self-service tier and giving shared services a system to log and track tickets. That scope is useful, and it also means technology has mostly automated the easiest tier while leaving the actual coordination between tiers, like the handoffs, the context-passing, or the escalation logic, dependent on a person doing it manually.
The more meaningful role for technology is closing that coordination gap directly by making sure the context from an early interaction actually reaches the specialist who eventually resolves it, rather than requiring the employee or the shared services team to relay it manually at each step. That shift, from powering one tier well to connecting all of them, is where most of the real opportunity in modernizing HR service delivery actually sits.
What Changes When the Model Has to Work Globally
A global HR service delivery model has to handle everything the domestic version does, plus a layer of complexity that compounds at every tier. Different labor laws, benefit structures, payroll practices, languages, and employee expectations all need to be accounted for when requests are routed through the same escalation structure.
Organizations typically land on one of two approaches:
- A single centralized team that supports all regions, offering standard processes, consistent measurement, and consolidated knowledge management.
- Regional hubs organized by geography (Americas, EMEA, APAC), which provide closer local knowledge, time-zone coverage, and language familiarity.
Neither approach removes the coordination problem. It only changes where the handoff happens. A centralized model still needs regional expertise for requests outside global policy, while regional hubs still need coordination for cross-border employees and global mobility cases.
Should You Fix the Tiers or Skip Past Them?
This is the choice organizations face today—improve the traditional tiered model with better handoffs, or bypass the structure for something built differently from the start.
For organizations with mature, deeply embedded shared services operations, the practical path is often incremental. That can mean better case management, smarter routing, and tighter integration between the HRIS and ticketing systems. These improvements reduce friction without requiring a wholesale redesign of established roles and processes.
But LACE Partners' research makes a compelling case for the other path. Organizations without deep legacy infrastructure have the opportunity to leapfrog traditional tiered structures entirely, investing directly in AI-enabled resolution rather than replicating a thirty-year-old structure by default.
Ema's Employee Experience Suite is built around this leapfrog path. Its AI Employees interpret a request, pull context automatically from connected systems, and resolve it directly across HR, IT, and payroll, collapsing several manual handoffs into a single resolved interaction. Its Generative Workflow Engine™ orchestrates these workflows, mediates data between steps, and stores results for downstream use and audit. Rather than bolting automation onto existing tiers, Ema's approach treats context preservation and intelligent routing as the foundation and not as an afterthought.
From Processing to Something Else Entirely
The three-legged stool has held up for thirty years because the underlying idea—matching a request to the right level of expertise—is genuinely sound. What is changing is the assumption that a person has to manually carry context between each leg of the stool for it to work. An HR service delivery model built around automatic handoffs, not just automated tiers, resolves the coordination problem that has quietly limited the traditional structure since Ulrich first described it.
If you're ready to see what collapsing those handoffs looks like in practice, explore how Ema's AI Employees resolve requests across HR, IT, and payroll without the sequential delays that tiered models produce.
Frequently Asked Questions
Is the HR service delivery model the same as an HR operating model?
Not exactly. An HR operating model describes how the HR function is structured, governed, staffed, and connected to business strategy. The service delivery model is one component within it, focused on how HR services and requests reach employees through tiers, channels, and escalation paths. You can redesign service delivery without overhauling the full operating model.
How do you know if your current HR service delivery model needs a redesign?
Watch for practical signals. Employees routinely bypass self-service and go straight to an HR person. HR business partners get pulled into administrative cases that should be resolved at lower tiers. Simple requests take days because context is lost between handoffs. LACE Partners identifies persistent budget pressure and difficulty attracting skilled talent into service center roles as signs.
What's the difference between HR shared services and an HR shared services center?
HR shared services is the operating approach—consolidating repeatable HR work so it can be handled consistently across multiple business units. An HR shared services center is the organizational unit that delivers those services. A company can adopt shared services as a capability without a single physical center, especially if delivery is virtual, regional, or distributed across intelligent platforms that integrate with existing HCM systems.
Can a small company use a tiered HR service delivery model, or is it enterprise-only?
A tiered model is not enterprise-only. In smaller companies, employees might start with self-service, then go to an HR generalist, with an external advisor handling complex benefits, legal, or employee relations questions.
How long does it typically take to redesign an HR service delivery model?
A narrow redesign of intake, ownership, and escalation paths can move relatively quickly. A global redesign involving HR roles, regional hubs, technology selection, service catalogs, governance, and change management takes considerably longer. LACE Partners emphasizes building a practical, evidence-based roadmap anchored in current reality rather than treating the redesign as a one-time org chart exercise.
