Shared Services

Building Shared Services as a CAO

Administrative shared services take work that is duplicated across departments, such as procurement, facilities requests, payroll support, and IT help, and deliver it once, to a defined standard, for the whole organization. For many Chief Administrative Officers it is the largest lever they control. This guide covers how to decide what to centralize, how to design the catalog and SLAs, how to fund the services, and why some shared-services programs stall.

Sector context matters. The healthcare, higher education, government, and nonprofit guides show how the patterns adapt.

Updated 2026-10-0211 min read

Start the build Service catalog template

Shared service goals

  • Standardization
  • Cost visibility
  • Service quality
  • Governance

Scope

What administrative shared services usually include

A shared service is any administrative capability delivered by one team to many internal customers under an agreed standard. The CAO's portfolio varies by organization, but the candidates below come up again and again. Not every item belongs in scope on day one; the point is to see the full landscape before choosing where to start.

Procurement and purchasing Accounts payable support Payroll and HR transactions Employee onboarding logistics IT service desk Facilities and workplace requests Records and document management Contract administration Fleet and asset management Mail, print, and logistics Travel and expense administration Internal communications support

Good candidates to centralize first

  • High-volume, rules-based transactions that every department performs.
  • Work where inconsistency creates audit, compliance, or safety exposure.
  • Services already partly centralized, where the next step is standardization.
  • Processes with an obvious pain point that customers will notice improving.

Services to leave local, at least for now

  • Work that depends on deep, unit-specific expertise or relationships.
  • Low-volume activities where centralizing adds handoffs without saving effort.
  • Processes that differ for good reasons, such as clinical versus academic units.
  • Anything that is about to be replaced by a new system or vendor.

Steps

How to build shared services

The four steps below are the core build sequence. Most programs loop through them more than once: a first wave of services goes live, the team learns, and later waves move faster.

Step 1: Define the service catalog

  • List all services by function, including work done informally inside departments.
  • Identify internal customers and estimate demand volumes from tickets, invoices, or staff time.
  • Assign a named service owner and delivery team to every service.

Step 2: Set SLAs and metrics

  • Document response and resolution targets based on measured performance.
  • Track quality and satisfaction alongside speed.
  • Publish dashboards so customers can see service performance without asking.

Step 3: Establish governance

  • Create escalation paths and a customer forum with decision rights.
  • Agree how service costs are funded and shown to budget owners.
  • Review service performance monthly and the catalog itself annually.

Step 4: Manage change

  • Communicate what changes for customers, and when, in plain language.
  • Train both the delivery team and requesters on new workflows.
  • Collect feedback in the first weeks and act on it visibly.

Maturity

The four maturity stages of shared services

Shared services rarely arrive fully formed. Most organizations move through recognizable stages, and the CAO's job is to be honest about the current stage and to fund the next one, rather than skipping straight to the end state.

Stage What it looks like Main focus for the CAO Signal you are ready to move on
1. Consolidated Staff from several departments are moved into one team, but each still works the old way. Stabilize service, protect continuity, and document current processes. Work is no longer dropped during the transition and volumes are visible.
2. Standardized One process, one intake channel, and one set of forms per service. Remove local variations, set baseline SLAs, and introduce a ticketing or request system. Most requests come through the standard channel and SLA data is reliable.
3. Service-managed A published catalog, owners, SLAs, cost transparency, and a customer forum. Manage performance, introduce showback or chargeback, and automate repetitive steps. Customers discuss service levels and trade-offs rather than individual complaints.
4. Business partner Multiple functions delivered through one front door, with analytics and continuous improvement. Use service data to advise leadership on demand, policy, and investment. Executives ask the shared-services team for insight, not only for transactions.

Skipping stage two is the most common shortcut and the most expensive one. Automating or charging for a process that still varies by department tends to lock in the inconsistency.

Catalog

What a useful service catalog actually contains

Many early shared-services efforts produce a catalog that is too long and too vague to use. A working catalog is structured around how internal customers consume the service, not how the team is organized to deliver it. The difference shows up in adoption.

What each catalog entry should answer

  • What the service is, in one sentence a non-specialist can understand.
  • Who can request it, and how: the entry point.
  • What the customer gets back, and within what time.
  • What it costs, or how it is funded if not chargeable.
  • Who owns it, and how to escalate if the SLA slips.

What to leave out of the catalog

  • Internal team structure, which is irrelevant to the customer.
  • Tool names, unless the customer interacts with them directly.
  • Process steps that happen behind the scenes.
  • Services that have never been delivered and have no realistic delivery plan.

Keep the first version short

A first catalog of a few dozen clearly described services is more useful than an exhaustive list of hundreds. Add entries as the team proves it can deliver them. The template below gives a ready structure for each entry.

Catalog template

SLAs

Setting SLAs that hold up in practice

SLAs are easy to write and difficult to defend. The most common mistake is setting targets based on what the team wishes were true rather than what it can repeatedly deliver. Start from current performance, set targets a step above that, and tighten them only once the new level is stable.

Three SLA layers

  • Response time: how quickly the customer hears back.
  • Resolution time: how quickly the work is finished.
  • Quality: whether the work was right the first time.

What undermines SLAs

  • Targets set higher than the team's measured baseline performance.
  • No tracking of misses; without measurement there is no enforcement.
  • One SLA covering services with very different demand profiles.
  • No escalation path when the SLA is missed twice in a row.
  • Clocks that keep running while the team waits on the customer for information.

Example SLA structures

The targets below are illustrative formats, not benchmarks. Replace each value with one based on your own measured baseline and the needs of your customers.

Service Response target Resolution target Quality measure
IT service desk: password or access issue Acknowledged within one business hour Resolved same business day Share of tickets reopened within a week
Purchase order for a catalog item Confirmed within one business day Order placed within three business days Orders returned for missing or wrong information
New-hire onboarding package Request confirmed within two business days Equipment, access, and workspace ready on the start date New-hire and manager feedback after the first week
Facilities repair request (non-urgent) Triaged within one business day Completed within an agreed window by priority level Repeat requests for the same issue
Contract review for a standard agreement Assigned to a reviewer within two business days First markup returned within an agreed number of days Contracts requiring a second full review

Funding

Chargeback models: how shared services get paid for

How a shared service is funded shapes how customers use it. Free services tend to be overused and undervalued; heavily charged services tend to be avoided or duplicated. Most organizations settle on a model that matches their budgeting culture and their stage of maturity.

Model How it works Works best when Watch out for
Central funding (overhead) The service is funded from a central budget with no charge to departments. Early stages, compliance-driven services, or very small organizations. No price signal, so demand grows unchecked and value is hard to show.
Showback Departments see what they consumed and what it cost, but are not billed. Building cost awareness before introducing real charges. Reports are ignored if no decision ever depends on them.
Allocation by driver Total cost is split using a simple driver such as headcount, square footage, or budget size. Services with steady, broadly shared demand, such as payroll or facilities. Disputes over whether the driver is fair to units with unusual profiles.
Usage-based chargeback Departments pay a unit rate per transaction, ticket, or hour consumed. Mature, well-measured services with variable demand. Administrative overhead of billing and incentives to avoid the service.
Hybrid (base plus variable) A base allocation covers core capacity; optional or premium work is charged by use. Catalogs that mix essential services with discretionary ones. Unclear boundaries between what is base and what is premium.

Rules that keep chargeback credible

  • Publish the rate methodology, not just the rates.
  • Set rates once per budget cycle and avoid mid-year changes.
  • Charge for things customers can control, such as volume or optional services.
  • Agree with the CFO how over- or under-recovery is handled at year end.

Public sector and nonprofit considerations

Government and grant-funded organizations often face rules on how indirect costs can be allocated to funds, programs, or grants. Work with finance early so the allocation method is defensible to auditors and funders, and see the government and nonprofit guides for sector detail.

Governance

Governance forums that keep shared services accountable

Governance is how customers get a say in a service they can no longer run themselves. Without it, every complaint escalates to the CEO. Keep the structure light and tie each forum to specific decisions.

Service review (monthly)

Service owners review SLA performance, backlog, and customer feedback, and agree fixes for repeated misses.

Customer council (quarterly)

Department representatives weigh in on priorities, catalog changes, and service-level trade-offs.

Executive steering (semi-annual)

The CAO, CFO, and senior sponsors decide on scope changes, funding model, and major investments.

Catalog refresh (annual)

Retire unused services, add proven ones, and reset rates and SLAs for the next cycle.

The operating model guide shows how these forums fit into the CAO's wider cadence, and the operating cadence checklist helps put them on the calendar.

Success

Shared services success metrics

Track a small set of indicators across service performance and financial impact. Report them the same way every month so trends, not single data points, drive the conversation.

Service performance

  • SLA compliance rate
  • Service satisfaction score
  • Average resolution time
  • Backlog aging
  • First-time-right rate

Financial impact

  • Cost per service unit
  • Cost avoided or saved through consolidation
  • Budget variance by service line
  • Vendor performance outcomes
  • Share of requests through the standard channel

The metrics dashboard guide and the executive dashboard metrics file give definitions you can adapt.

Failure modes

Why shared services fail to land

Even a well-designed shared-services model can fail. The patterns below account for many of those failures, and each can be avoided with deliberate design.

Customers feel worse off

When a centralized service replaces a local one, the customer often perceives a loss of attention. If the new service does not match or beat the previous experience, customers route around it. Make the first ninety days of a new shared service visibly faster, more reliable, or easier than what came before.

The team becomes invisible

A shared-services team that talks only in tickets and dashboards loses the political support it needs at the next budget cycle. Regular communication, such as short release notes, named owners, and a clear roadmap, keeps the team visible to the executives who fund it.

Lift and drop

Moving people into a central team without changing processes, systems, or intake simply relocates the problem. Plan the standardization work and budget for it before the reorganization is announced.

A cost-only business case

Programs sold purely on savings come under pressure the moment savings arrive slower than promised. A business case that also covers control, consistency, and service quality gives the program a reason to exist beyond year one.

Shadow teams rebuild locally

When departments quietly hire their own coordinators to do work the shared service is meant to cover, it is a signal the service is not meeting a real need. Treat shadow capacity as feedback, ask what it is compensating for, and fix that gap rather than simply banning it.

Lost expertise in the move

Experienced staff in departments often hold knowledge that is not written down. If they leave during consolidation, service quality drops. Document critical knowledge early and give key people a clear role in the new model.

FAQ

Shared services questions CAOs ask

What are administrative shared services?

Administrative shared services are support functions, such as procurement, payroll support, IT help, facilities, and records, that one team delivers to many internal departments under a common standard, instead of each department running its own version.

Does the CAO usually own shared services?

In many organizations, yes. The CAO is commonly the executive accountable for administrative shared services, while finance-specific services may sit with the CFO and people policy with the CHRO. The exact split depends on how the leadership team divides its portfolio.

Which services should be centralized first?

Start with high-volume, rules-based services that every department performs, that carry compliance risk when done inconsistently, and where customers will notice an improvement. Leave specialized or low-volume work local until the model is proven.

What is the difference between chargeback and showback?

Showback reports what each department consumed and what it cost without billing them. Chargeback actually moves the cost into the department budget. Many organizations use showback first to build cost awareness, then introduce chargeback for mature, well-measured services.

How long does it take to set up shared services?

It depends on scope and starting point. A single, well-defined service can often be standardized within a few months, while a multi-function shared-services organization is usually built in waves over several years. Plan for a first wave that proves the model before expanding.

How do you measure whether shared services are working?

Track SLA compliance, customer satisfaction, resolution time, backlog aging, and cost per service unit, and watch whether requests flow through the standard channel. If departments are rebuilding the service locally, that is a strong signal it is not meeting their needs.

Download the service catalog template

Use the template to define shared services quickly. Pair it with the operating model for governance and the metrics dashboard for ongoing oversight.

Get the template