How we make this list: we compare pricing pages, product documentation and public user reviews. Ratings come from G2, Capterra or Trustpilot, with the date we checked them.
Key takeaways
- This is a practices list, not a product list. The entries are operating practices used by mature shared services and global business services teams, each with the body or source that defines it.
- Standardise a process before you automate or move it. Automating a broken local variation locks the problem in.
- A named global process owner, a published service catalogue, and a short KPI set are the three governance pieces most failed centres skip.
- Cost per transaction is the headline measure, but first contact resolution and cycle time are what the business actually feels.
- Mature centres climb from transactions to insight: reporting, analytics, and process improvement are where shared services earns its keep.
Quick comparison
Is your product missing or out of date?
We review this list regularly. Vendors can request a listing, send updated pricing or ask about featured placement.
The best shared services practices
#1
Standardise before you automate
Best for: Centres drowning in local variations
Visit apqc.org
Starting priceFree to adopt
Free planYes
The first practice is order of operations: agree one standard way to run a process, then move it, then automate it. APQC, the process benchmarking body, publishes process frameworks that give centres a common structure to standardise against. Centres that skip this step lift twenty local variants of invoice processing into one building and spend years untangling them.
The failure mode is standardising on paper only. If plants and country teams keep their side spreadsheets, the centre inherits the cost of both models. Standardise before you automate fits any centre still running inherited local variations, and it should be finished before an RPA or workflow tool is bought.
What we like- Removes duplicate process variants
- Makes automation and training cheaper
- Creates one version of the truth for policy
Watch out for- Needs executive backing to overrule local habits
- Slower at the start
- Some local rules are legal, not habit
Starting priceFree to adopt
Free planYes
A global process owner holds end to end accountability for one process, such as procure to pay or hire to retire, across every unit the centre serves. SSON, the shared services network, treats the GPO model as a marker of maturity because it ends the argument about whose version of the process wins.
The failure mode is a GPO with responsibility and no authority: change requests die in committee and local teams drift back to their own methods. Name a global process owner fits multi country operations where the same process runs differently in each region and nobody owns the whole.
What we like- One accountable owner per process
- Faster decisions on change requests
- Clear escalation path for the business
Watch out for- Owner needs real authority, not a title
- Can clash with country management
- Role is often under resourced
Starting priceFree to adopt
Free planYes
A service catalogue lists exactly what the centre delivers, in what form, at what turnaround, and what it needs from the requester. Shared Services Link, a practitioner community, pushes the catalogue and service level agreement pairing as the basic contract between a centre and its customers.
The failure mode is a catalogue written once and never maintained, while the real menu lives in email threads. Publish a service catalogue with SLAs fits centres fielding constant ad hoc requests, because it turns "can you just" into a defined service with a defined queue.
What we like- Business knows what the centre offers and what it does not
- SLAs make performance a fact, not an opinion
- Stops bespoke one off work
Watch out for- Catalogue needs regular updates
- SLAs set too tight punish the wrong behaviour
- Business will test the edges
Starting priceFree to adopt
Free planYes
Mature centres track a handful of numbers: cost per transaction, cycle time, error rate, first contact resolution, and customer satisfaction. The Hackett Group benchmarks these figures across thousands of organisations, which lets a centre say whether its cost per invoice is good or merely familiar.
The failure mode is measuring what is easy instead of what matters, such as tickets closed rather than problems solved. Measure a short KPI set fits centres that must justify their budget each year, because a benchmarked number ends arguments that anecdotes cannot.
What we like- Cost per transaction anchors the value story
- Benchmarks show where you stand against peers
- A short list keeps focus
Watch out for- Benchmarks cost money at full depth
- Bad data makes confident wrong numbers
- Vanity metrics creep back in
Starting priceFree to adopt
Free planYes
Tiering splits incoming work by complexity: self service and automation at tier 0, trained generalists at tier 1, specialists at tier 2, and retained experts or the business itself at tier 3. Deloitte's global business services research describes tiering as the structure that lets one centre serve both volume and judgment work.
The failure mode is dumping everything on tier 1 to hit a cost target, which burns out staff and ships errors downstream. Govern with tiered service levels fits centres whose queue mixes password resets with pension calculations, because one service level cannot honestly cover both.
What we like- Routine work runs cheap and fast
- Complex cases get expert handling
- Tiers make staffing and pricing honest
Watch out for- Triage rules need constant tuning
- Business users dislike hearing their work is tier 1
- Experts can become bottlenecks
#6
Automate in the right order
Best for: High volume repeatable tasks
Visit kpmg.com
Starting priceFree to adopt
Free planYes
Automation pays when it lands on a standard, stable, high volume process: invoice capture, journal posting, onboarding data entry. KPMG's shared services guidance puts automation after process redesign for this reason. A bot copying a bad process just produces the bad output faster and at licence cost.
The failure mode is a bot farm built on screenshots and hope, where every system update breaks ten automations. Automate in the right order fits centres with proven standard processes and clear volume, where each automated hour can be counted and the savings shown.
What we like- RPA and workflow tools cut manual keying
- Automation after standardisation actually sticks
- Freed hours fund improvement work
Watch out for- Tool licences cost real money
- Automating exceptions creates maintenance debt
- Staff fear must be managed openly
#7
Build continuous improvement in
Best for: Year two and beyond
Visit asq.org
Starting priceFree to adopt
Free planYes
Continuous improvement means the centre treats its own processes as a product under development: problem logs, root cause work, and small tested changes every month. ASQ, the quality association, supplies the methods, from basic problem solving to Lean and Six Sigma, that centres borrow for this.
The failure mode is a launch event and no rhythm after it. Improvement survives when team leads own a monthly slot and report one fixed problem, however small. Build continuous improvement in fits centres past their migration phase, when the easy savings are gone and the remaining waste is structural.
What we like- Small fixes compound into real savings
- Staff closest to the work spot problems first
- Improvement record supports the centre at budget time
Watch out for- Needs protected time or it dies under queue pressure
- Suggestion schemes without follow through breed cynicism
#8
Run it as a business
Best for: Centres serving many business units
Visit pwc.com
Starting priceFree to adopt
Free planYes
The final practice is posture: a mature centre prices or at least shows the cost of its services, reports against its catalogue, and plans capacity from demand forecasts. PwC's shared services research describes this as the shift from cost centre to service business, and it is what keeps business units from rebuilding their own back offices in secret.
The failure mode is behaving like a monopoly supplier to a captive customer. Units that feel trapped route around the centre, and the shadow cost lands back on the company. Run it as a business fits large multi unit organisations where the centre must keep earning its mandate every budget cycle.
What we like- Transparent chargeback or showback builds trust
- A service mindset cuts shadow back offices
- Planning follows demand, not headcount habits
Watch out for- Chargeback models can drive gaming
- Internal marketing takes real effort
- Requires finance partnership to price services
How we chose
We built this list from the frameworks and benchmark sources that define shared services practice: APQC process frameworks, SSON maturity research, Hackett Group benchmarks, and the global business services studies published by Deloitte, KPMG, and PwC. We favoured practices that appear across several sources and that a centre can adopt without buying software. This article lists practices, not products, so the field format is adapted honestly: the website is the defining source, and the price fields state that adoption is free while noting the real cost, which is management time.
How to run shared services well
Sequence matters more than selection. Standardise the top processes first, because every later practice assumes one way of working. Put governance in place early: a process owner per process, a catalogue the business can read, and service levels you can actually measure. Only then automate, in volume order, and publish the savings in the same KPI set you chose at the start. Protect improvement time in the roster, not in a memo. Finally, review the catalogue with your business customers twice a year and retire services nobody uses. Centres fail from drift, not from a single bad decision, and the practices above are mostly anti drift habits.
Frequently asked questions
What is the difference between shared services and outsourcing?
Shared services keeps the work inside the company under its own management, usually in a dedicated centre. Outsourcing hands the work to an external provider under contract. Many companies run a mix, keeping judgment work in house and sending volume work out.
What is a global business services model?
It is the mature form of shared services: one organisation covering several functions, such as finance, HR, IT, and procurement, often with process owners and tiered delivery. The practices in this guide, especially process ownership and tiering, are its building blocks.
How is shared services success measured?
Cost per transaction, cycle time, error and rework rate, first contact resolution, and customer satisfaction. Benchmarks from firms such as Hackett show how those figures compare with peer organisations, which is what turns a number into a judgment.
What processes move to shared services first?
High volume, rule based, repeatable work: accounts payable, payroll processing, travel and expense, employee data changes, and IT service desk requests. Work needing local judgment or legal knowledge usually stays in the business longer.
Why do shared services centres fail?
The common causes are moving unstandardised processes, weak governance, no agreed service levels, and treating the business as a captive customer. Each maps to a practice on this list, which is why the basics matter more than the tooling.
Does shared services need special software?
Not to start. It needs workflow and ticketing once volume grows, plus automation tools for capture and posting later. Buying software before standardising the process is the most expensive order of operations in this field.
Sources we checked (0)
Is your product missing or out of date?
We review this list regularly. Vendors can request a listing, send updated pricing or ask about featured placement.
Shihab Mia · Founder, Rankite
Shihab has 5+ years in B2B and SaaS SEO. He writes Rankite's software buying guides from pricing pages, documentation and public user reviews. More about Shihab.
Name a global process owner
Publish a service catalogue with SLAs
Measure a short KPI set
Govern with tiered service levels