GBS Insider Club Field Guide Free
Pillar 2 — Operational Excellence
Cluster 1 of 6
Pillar 2 · Operational Excellence · Cluster 1

Process and Knowledge Management A process exists in documentation. Without it, you have a habit.

Every GBS center runs on documented processes. But how that documentation is structured, maintained, and governed separates the centers that scale from the ones that break when someone leaves.

3 document layers in a well-structured GBS: SOP (what), DTP (how), controls (guardrails)
Annual minimum review cycle for SOPs and DTPs — more frequently after process changes or audit findings
GPO Global Process Owner — the role responsible for keeping process design current, consistent, and improvement-ready
Topic 01

What a process is — and why managing it matters

TL;DR

Everything you touch in GBS is a process — inputs, steps, outputs, repeated. Knowing which ones deserve standardization is the starting skill. The model is in THE FIX.

You do not have tasks.
You have processes.

2 min read · full theory in the expandable
The Problem
R
Ravi
AP analyst · Month 8 · Pune

Ravi describes his job as "handling invoices." A process consultant asks him to walk through one.

Receive, validate, match, resolve, post — five steps, three systems, two decision points.
He had never seen his own work as a structure. Just as a pile.

"I run a process 40 times a day. I just called it my job."

He feels curious about work he thought he knew.

The Trap

You execute steps without seeing the structure — and cannot improve what you cannot see.

The Fix

A process is three parts, repeated — and repetition decides its treatment.

INPUT → STEPS → OUTPUTThe universal shape. What arrives, what happens, what leaves.
STANDARDIZEHigh volume, high repetition. These earn documentation and controls.
LEAVE FLEXIBLERare, judgment-heavy work. Over-documenting these wastes effort.

Ravi sketches his invoice flow in five boxes. The pile becomes a structure — and structures can be improved.

What a process is — in depthTHEORY · 3 MIN

Before you document anything, it helps to know what a process actually is, and why some are worth standardizing while others are left alone. This is the ground the rest of the cluster stands on.

What a process is

A process is a set of steps that turns inputs into an output for a customer. In GBS, that customer is usually another part of the business. Service delivery is the day-to-day work of running these processes reliably, at the agreed quality and speed.

Take invoice processing. The input is a supplier invoice. The steps are matching it to a purchase order, checking it, and posting it for payment. The output is a paid supplier and a clean record for the business. Every task you do sits inside a process like this, whether or not anyone has drawn it out.

What process management is, and why it matters

Managing a process means owning how it runs: agreeing the steps, measuring how well it performs, and improving it over time. That is different from doing the work. Doing the work is completing today’s invoices. Managing the process is making sure every invoice, by every person, is handled the same reliable way.

This matters because work that is not managed drifts. Each person settles into their own version, quality varies by who is on shift, and errors stay hidden because there is no agreed baseline to compare against. You cannot improve, automate, or audit a process you cannot see. Managing it is what makes it visible.

Standard processes, and why they matter

A standard process is one agreed way of doing the work, written down and followed by everyone on the team. It removes needless variation so the output is predictable, and frees people from re-deciding things that are already settled.

Standardized work trains new joiners faster and makes errors easier to spot. It is also the precondition for automation. A bot or an AI assistant can only take over a process that runs the same way every time. Work that changes from person to person cannot be automated, measured fairly, or signed off in an audit.

Harmonized processes, and why they matter

Harmonization takes standardization one level up. It aligns the same process across locations, entities, or teams so they all work the same way, while allowing the few local differences that are genuinely required, such as tax, legal, or currency rules.

Standardization happens within a team. Harmonization happens across the organization. It matters because a global process owner can only manage, measure, and improve a process that runs consistently across every hub. Five sites doing accounts payable five different ways cannot be reported on as one process, let alone improved as one.

Process maturity

Processes tend to move through stages as an organization invests in them. A simple way to picture it.

PROCESS MATURITY higher maturity 1 Fragmented Everyone works their own way 2 Standardized One documented way to do it 3 Harmonized Same way across sites 4 Optimized Measured and improved 5 Automated Rule-based, AI-assisted

Knowing where a process sits tells you what to work on next. A fragmented process needs standardizing before it is worth automating — trying to automate chaos just makes faster chaos. Most of the documentation, ownership, and improvement topics in this cluster are the work of moving a process up these stages.

With that grounding in place, the next topics cover how you capture a process so it can be standardized and shared: the document layers, SOPs and DTPs, and the governance that keeps them alive.

Monday Move

Draw your most frequent task as boxes and arrows. Five minutes, one sheet.

Process Maps End-to-end flows What happens and when SOPs Standard procedures How to do it right DTPs Desktop procedures Step-by-step execution
The Three Document Layers
Topic 02

The three document layers — what they are and how they connect

TL;DR

Process documentation comes in three layers — map, SOP, DTP — from bird’s-eye to keystroke. Know which layer answers which question. The model is in THE FIX.

Three documents, one process.
Each answers a different question.

2 min read · full theory in the expandable
The Problem
P
Priya
Process SME · Migration + BAU · Bangalore

Priya’s migration needs documentation. The receiving team asks for "the documents" — singular confusion follows.

She sorts it: the process map shows the flow, the SOP sets the rules, the DTP walks the keystrokes.
Three layers, three audiences, three levels of zoom.

"A manager reads the map. An auditor reads the SOP. A new joiner needs the DTP."

She feels orderly — and so does the handover.

The Trap

You write one document trying to serve every reader — and serve none of them.

The Fix

Three layers, three zoom levels.

PROCESS MAPThe flow, visually. Boxes, arrows, handoffs — the bird’s-eye view.
SOPThe rules. Scope, roles, controls, exceptions — the blueprint.
DTPThe keystrokes. Step-by-step with screenshots — the execution manual.

Her handover package has all three layers, labeled. Every reader finds their altitude.

The three document layers in depthTHEORY · 4 MIN

Most GBS professionals use these documents every day without knowing their names. Understanding the hierarchy helps you know which document to look in — and which one to update.

Policy What and why Standard Operating Procedure (SOP) How — process steps and decisions Desktop Procedure (DTP) Exactly what to click — screenshots and keystrokes STRATEGIC OPERATIONAL TACTICAL
Policy → SOPDTP Hierarchy
L1
SOP

Standard Operating Procedure — the "what and why"

The high-level blueprint for a process — the overview a new manager reads to understand it, not the step-by-step instructions for someone executing it.

  • Describes what the process does and who is involved
  • Captures inputs and outputs, systems used, and which controls apply
  • Written for comprehension, not execution

Contains: process scope, inputs and outputs, process map (swimlane or flowchart), parties involved, systems used, controls and compliance references, exceptions.

Example: AP Invoice Processing SOP Shared with: auditors, new hires, BU stakeholders
L2
DTP

Desktop Procedure — the "exactly how"

The step-by-step execution manual — written for the person performing the task, not the manager overseeing it. If an SOP is the recipe, the DTP is the photo-by-photo cooking guide.

  • Includes screenshots and system navigation paths
  • Field-by-field instructions and error-handling guidance
  • Granular enough that a new joiner can execute without asking for help

Contains: numbered steps, screenshots, system navigation, error-handling instructions, keyboard shortcuts, approval routing paths.

Example: How to post an invoice in SAP — step 1 to step 24 Shared with: processors, new joiners during training
L3
Controls

Controls and compliance references — the "guardrails"

The rules that cannot be broken — referenced in the SOP but may also exist as standalone documentation, especially in regulated or SOX-compliant environments.

  • Segregation of duties requirements and approval thresholds
  • Mandatory two-person sign-offs and regulatory deadlines
  • Traceable to a control register and auditable on demand

Contains: control names and IDs, what triggers each control, who performs the control, evidence required for audit.

Example: Four-eyes approval for payments above €50,000 Referenced in: SOPs, audit workpapers, risk registers
The real-world gap: Many organizations have SOPs. Fewer have current DTPs. Almost none have all three layers consistently maintained and linked together. The further from setup you get, the more the documentation drifts from reality. Processes change, systems are upgraded, and nobody updates the manual. This is a governance failure.
DOCUMENTATION LAYERS SOPs Step-by-step procedures Screenshots + system paths Exception handling Version controlled HOW TO DO IT PROCESS MAPS Visual workflow flows Decision points Handoff points Swimlane format HOW IT FLOWS DTPs Desktop procedures Quick-reference cards System shortcuts Cheat sheets HOW TO DO IT FAST

Documentation layers — process maps, SOPs, DTPs

Monday Move

Check your process: do all three layers exist? Name the missing one.

Two of those layers get confused daily. SOP vs DTP, settled.

GBS Process & Knowledge Management Ecosystem — documentation, centralized knowledge hub, E2E process ownership, and continuous improvement cycle

GBS process & knowledge management ecosystem: Document → Centralize → Own → Improve

Topic 03

SOP vs DTPDesktop Procedure — a step-by-step execution manual with screenshots, used by the person performing the task. More detailed than an SOP. — side by side

TL;DR

The SOP is the blueprint; the DTP is the execution manual with screenshots. Same process, different altitude, different reader. The model is in THE FIX.

Same process, two documents.
Stop confusing them.

2 min read · full theory in the expandable
The Problem
A
Amara
O2C analyst · Year 1 · Lagos

Day one, Amara learns her process from a DTP: every click, every screen, every field.

Month three, an auditor asks about controls and exceptions. The DTP has no answer — that lives in the SOP she never opened.
Two documents, one process, and she only knew half the pair.

"The DTP taught me the clicks. The SOP explains why the clicks exist."

She feels complete after reading both.

The Trap

You learn the keystrokes and skip the rulebook — until a question arrives that keystrokes cannot answer.

The Fix

One pair, cleanly split.

SOPWhy and what. Scope, roles, controls, exception rules. Audit-facing, stable.
DTPHow, exactly. Screenshots, clicks, field values. Training-facing, updated with every system change.
THE LINKDTP implements the SOP. If they disagree, one of them is wrong — flag it.

Amara reads the SOP behind her DTP. The next audit question gets an answer instead of a shrug.

SOP vs DTP in depth — side by sideTHEORY · 4 MIN

The same process has two documents describing it. Here is what belongs in each.

Dimension
SOP — Standard Operating Procedure
DTP — Desktop Procedure
Purpose
Describe what the process does and why it exists
Show exactly how to perform each step
Audience
Managers, auditors, new hires, BU stakeholders
Processors, analysts, anyone executing the task
Level of detail
High-level — process flow, not click-by-click
Granular — step 1 to step N, with screenshots
Format
Narrative + process map (swimlane, flowchart)
Numbered steps + screenshots + system navigation paths
Shared with
Internal and sometimes external (auditors, clients)
Internal only — operational teams
Update trigger
Process scope change, governance redesign, new controls
Any system change, UI update, step modification
Owner
Process Owner / GPO
Team Lead / Senior Processor — with GPO approval
Monday Move

Open the SOP behind your daily DTP. Find one rule the DTP never told you.

Documents written once. Kept alive is the harder part.

Topic 04

Governance: keeping documentation alive

TL;DR

Unmaintained documentation becomes a liability — failing audits and training people wrong. Ownership, review cycles, and one source fix it. The model is in THE FIX.

Three versions of the SOP.
All of them outdated.

2 min read · full theory in the expandable
The Problem
K
Klaudia
Senior associate · Year 3 · Krakow

Klaudia needs the current SOP. She finds one on the shared drive, one in email, one on the team site.

All three differ. None match the process as it runs today.
A new joiner trained on version two last month. The auditor sampled version one.

"We do not have documentation. We have documentation archaeology."

She feels exasperated — at a fully preventable mess.

The Trap

You treat documentation as a writing task. It is a maintenance contract — or it is decay.

The Fix

Living documentation needs three mechanisms.

OWNEROne name per document. Accountability for accuracy, not just authorship.
REVIEW CYCLEDated and triggered. Annual minimum, plus every system or process change.
SINGLE SOURCEOne location, everything else links. Copies are future contradictions.

One owner, one home, one review date per document. The archaeology ends; the audit trail begins.

Documentation governance in depthTHEORY · 4 MIN

Documentation that is not maintained becomes a liability — not just an asset. An outdated SOP fails audits. An outdated DTP trains people incorrectly. The governance cycle is what prevents both.

01

Review triggered

Annual cycle or event-driven: system change, process redesign, audit finding, new regulation

02

Team inputs collected

Processors flag what no longer matches reality. The best input comes from the people executing the work daily

03

Update drafted

Process Owner or Senior Processor drafts changes. Screenshots refreshed. Controls references checked

04

Reviewed and approved

GPO or Process Owner approves. For controls changes: Compliance or Internal Audit review may be required

05

Published and trained

Updated document shared with the team. Brief retraining session held. Acknowledgment logged for audit trail

What gets checked during an audit
  • Version control: is the document dated, versioned, and clearly showing what changed from the prior version?
  • Approval evidence: is there a signature or system record showing who approved the current version?
  • Training records: is there evidence that the team was trained on the current version — not an older one?
  • Controls completeness: does the SOP reference all relevant controls? Are they traceable to the control register?
  • Currency: does the document reflect how the process actually runs today — including any system changes since the last update?
GBS Insider Club Insights
  • The best input for updating a DTP comes from the people using it every day. When a process changes, the processors know before the document does. Building a simple feedback mechanism — even a shared comment in the document itself — surfaces gaps faster than any top-down review cycle.
  • An outdated DTP is not a neutral document. It is actively training new people incorrectly. In a high-turnover GBS environment, an unreviewed DTP from two system upgrades ago becomes the source of the errors that show up in your FPY metrics six months later.
  • SOPs and DTPs should be in the same review cycle as each other — and the same cycle as the SLA review. If the process changes and the SLA target doesn't, or the SLA changes and the SOP isn't updated to reflect the new service definition, you have a governance gap.
CENTRALIZED KNOWLEDGE HUB SINGLE SOURCE OF TRUTH One location · One version · One owner SOPs + DTPs Procedures + quick refs PROCESS MAPS Flows + swimlanes TRAINING Onboarding + L&D SHAREPOINT · CONFLUENCE · NOTION · CUSTOM WIKI

Knowledge hub — centralized, searchable, governed

Monday Move

Find your process’s SOP copies. Kill all but one; make the rest links.

Governance per document. One role owns it per process — globally.

Topic 05

The Global Process OwnerGPO — Global Process Owner. Accountable for end-to-end process design, standardization, governance, and improvement across all GBS and business unit touchpoints for a given process. — what the role actually is

TL;DR

The Global Process Owner is accountable for one process end-to-end, across every hub — design, standards, improvement. Know yours. The model is in THE FIX.

Your process runs in four hubs.
One person owns its design.

2 min read · full theory in the expandable
The Problem
P
Peter
Team lead · Year 2 · Budapest

Peter’s team wants to change a process step. Site leadership approves. Two weeks later, the change is reversed.

The reversal comes from a role Peter had never dealt with: the Global Process Owner.
The process runs in four hubs — and local changes break global standards.

"We changed our copy of the process. The process has one owner, and it was not us."

He feels enlightened about a layer of the org he had only heard named.

The Trap

You change your local slice of a global process and discover the owner by collision.

The Fix

The GPO holds end-to-end accountability — three things follow.

DESIGNThe process blueprint is theirs. Across every hub, every variant.
STANDARDSDeviations need their yes. Local optimization is global fragmentation without it.
YOUR ROUTEImprovement ideas flow to them. A good idea with the GPO’s backing scales to four hubs.

Peter re-submits the change through the GPO — and it ships globally, with his team credited as origin.

The GPO role in depthTHEORY · 4 MIN

The GPO title sounds clear. In practice, what it means varies enormously by organization. Here is what the role should do — and the different ways it gets set up.

What a GPO is accountable for — in any model
  • Process design and standardization: defining how the process runs globally — the target operating model — and driving harmonization across regions and entities
  • Documentation governance: ensuring SOPs and DTPs are current, approved, and accessible — the governance cycle described above is the GPO's responsibility to enforce
  • Continuous improvement: identifying where the process underperforms, sponsoring improvement initiatives, and tracking the outcomes
  • Change management for process changes: when the process design changes, the GPO owns the communication, retraining, and adoption — not just the updated document
  • Metrics ownership: defining what KPIs the process should be measured by and ensuring the right data is captured to report them reliably
  • Escalation point: when operations teams and BU stakeholders disagree about how a process should work, the GPO is the arbiter
Process Design Standardization & harmonization Doc Governance SOPs & DTPs current & approved Cont. Improvement Sponsor & track initiatives Change Management Comms, retrain, adoption Metrics & Escalation KPI ownership · final arbiter GPO SCOPE
GPO owns design through escalation end-to-end.

Research by EY across 250+ client interviews identifies four main GPO setup models. Each has different reporting lines, authority levels, and operational intensity. No single model is universally correct. It depends on organizational structure and GBS maturity.

Business-unit-centric model
  • GPO embedded in or reports to a business unit
  • Strong connection to BU priorities and local context
  • Risk: GBS operations teams may not feel the GPO's authority
  • Common in early-stage or decentralized organizations
  • GPO may be part-time, combined with a functional leadership role
Function-centric model
  • GPO sits within the function (Finance GPO inside Group Finance)
  • Strong process expertise and functional credibility
  • Dotted-line relationship to GBS operations teams
  • Works well when the function retains significant process authority
  • Common for F&A and HR processes in mid-maturity organizations
GBS-centric model
  • GPO sits within the GBS organization itself
  • Direct authority over GBS operations teams
  • Clear accountability — one owner for design and delivery
  • Requires GPO to maintain strong BU relationships despite reporting into GBS
  • Common in mature, well-established GBS organizations
Enterprise-centric model
  • GPO reports directly to C-level (CEO, COO, or CFO)
  • Highest authority — can drive standardization across all BUs and GBS
  • Required for true enterprise-wide process harmonization
  • Rare — most common in organizations with strong centralization mandates
  • Best-practice model per Genpact and EY research
GPO engagement vs. change intensity — how the role scales
CHANGE INTENSITY IN ORGANIZATION GPO ENGAGEMENT LEVEL Stable operations Light governance role Process improvement Active involvement Transformation / migration: full GPO LOW MEDIUM HIGH LOW MED HIGH

GPO engagement is not constant. During stable operations the GPO plays a lighter governance role. During major transformation, migration, or tool deployment, the GPO becomes a central driving force. The intensity of the role follows the change curve.

The GPO controversy — why the role often underdelivers

The GPO title frequently exceeds the authority granted.

  • A GPO with no solid-line to operations, no budget, and no dedicated time cannot drive the standardization the role requires
  • Research shows the most effective GPOs have direct accountability to senior leadership — not just a dotted-line to GBS
  • Without real authority, the GPO becomes a process documentarian rather than a process owner
END-TO-END PROCESS OWNERSHIP TRIGGER PROCESS HANDOFF OUTCOME Request received Steps executed Cross-team transfer Delivered to customer ONE OWNER ACCOUNTABLE FOR THE ENTIRE CHAIN

End-to-end ownership — GPO accountability model

Monday Move

Find out who owns your process globally. One name. Write it down.

Processes owned. Services get a menu.

? CHALLENGE YOURSELF click to expand
  • When was the last time you reviewed and updated an SOP for your process? Is it current, or would a new joiner struggle to follow it?
  • Does your team have a single source of truth for documentation — or do SOPs live in emails, shared drives, and people's heads?
  • Could you map your end-to-end process from trigger to outcome right now? Where are the handoff points, and who owns each?
Topic 06

The service catalog — GBS's menu of services

TL;DR

The service catalog is GBS’s menu: what is offered, to whom, on what terms. Twelve services stop being folklore when they are written down. The model is in THE FIX.

Twelve services, zero menu.
Stakeholders order off-card anyway.

2 min read · full theory in the expandable
The Problem
M
Miguel
New team lead · Week 6 · Manila

Miguel inherits a team delivering "about twelve" services. Nobody can list them precisely — including the stakeholders consuming them.

He writes the menu: each service named, described, scoped, with its turnaround and its consumer.
Two surprises: one service nobody remembered promising, one delivered for a team that no longer exists.

"You cannot manage a menu that lives in people’s heads."

He feels clear-eyed about what his team actually is.

The Trap

You deliver services from memory — and scope, demand, and credit all leak.

The Fix

A catalog entry is four lines per service.

WHATThe service, named and described. Specific enough to point at.
WHOThe consumers. Which teams may order it.
TERMSTurnaround and boundaries. What is included, what is extra.
HOWThe intake path. Where requests enter — and where they do not.

The dead service retires; the forgotten one gets terms. Demand finally has a front door.

The service catalog in depth — the menu modelTHEORY · 4 MIN

Think of the service catalog as the GBS equivalent of a restaurant menu. It tells stakeholders what is available, what is not, who can order it, and what standard comes with it.

What a service catalog contains
  • Service name and description: plain-language explanation of what the service does — not internal process jargon
  • Scope and eligibility: which business units, entities, or geographies can use this service
  • What is included — and explicitly excluded: the service boundary. What GBS will do and what remains with the business
  • Inputs required from the business: what the business must provide for GBS to execute — data formats, approval chains, timing
  • Deliverables and outputs: what GBS will return — reports, processed transactions, resolved queries
  • Performance standards: the SLA metrics and targets that apply to this service
  • Escalation and exception handling: who to contact, and through what channel, when something falls outside standard
  • Cost or allocation model (if applicable): how the service is charged to the business unit, if a chargeback model is in place
When it matters most

Onboarding new processes

When a new process is migrated to GBS, the service catalog entry sets expectations before anyone starts working. It is the shared reference for what was agreed.

When it matters most

Managing stakeholder expectations

When a BU asks for something GBS doesn't do, the catalog is the objective reference. "That isn't in our service catalog" is cleaner than "we don't do that."

When it matters most

Resolving scope conflicts

When GBS and a BU disagree about who is responsible for something, the catalog — combined with the SLA — is the primary evidence. Undocumented scope is always disputed scope.

In practice: Many GBS organizations do not have a formal, standalone service catalog. The service definition exists as part of the SLA or the transition documentation from when the process was set up. This works until scope creep accumulates, new stakeholders join, or processes are restructured. The service catalog should be reviewed in the same cycle as the SOP, not treated as a one-time setup document.
Monday Move

List your team’s services from memory, then verify. The gaps are the finding.

The menu says what you serve. XLAs ask how it tasted.

Topic 07

XLAs — measuring how the service feels, not just how it performs

TL;DR

SLAs measure whether targets were hit; XLAs measure how the service felt to receive. Green SLAs with unhappy customers is the gap XLAs close. The model is in THE FIX.

The SLA was green.
The customer was still cold.

2 min read · full theory in the expandable
The Problem
A
Amara
O2C analyst · Year 1 · Lagos

Amara’s dashboard: 96% on-time, targets met, everything green.

Her stakeholder’s reality: chasing updates, re-explaining context to every new handler, effort on their side the SLA never counts.
Compliant service. Poor experience. Both true at once.

"We measured our promise. Nobody measured their experience."

She feels unsettled by a green that lies.

The Trap

You defend the SLA while the stakeholder describes the experience — two conversations passing in the dark.

The Fix

XLAs add the receiving side to the measurement.

SLADid we hit the target? Time, volume, accuracy — the provider’s view.
XLAHow did it feel to receive? Effort required, clarity, confidence — the consumer’s view.
TOGETHERGreen needs both. A hit target with a burned stakeholder is half a service.

One experience question joins the scorecard. The dashboard starts telling the whole truth.

XLAs in depth — experience-level agreementsTHEORY · 4 MIN

SLAs measure whether GBS hit its targets. XLAs measure whether the people receiving the service actually had a good experience. These are not the same thing.

SLA — what we have always measured

Service Level Agreement

  • Measures technical performance: response time, accuracy rate, turnaround time
  • Binary — either the target was hit or it wasn't
  • Does not capture how easy the service was to use
  • Does not capture how the business felt about the interaction
  • A ticket answered in 24 hours but unhelpfully is SLA-compliant but experience-poor
  • The floor — necessary but not sufficient
XLA — what the experience actually is

Experience Level Agreement

  • Measures the end-user's perception: satisfaction, ease of use, emotional response
  • Captured through surveys, sentiment analysis, and structured feedback
  • Surfaces the gaps that KPIs cannot see — effort required, interaction quality
  • Used alongside SLAs, not instead of them
  • Originated in IT service management (ITSM/ITIL4) — now expanding into GBS
  • Still rare in GBS practice — but gaining traction in mature organizations
SLA vs XLA — what each measures
SLA MEASURES Turnaround time hit? Error rate within threshold? Volume processed on time? SLA compliance %? Output — did it happen? + XLA MEASURES Was the service easy to use? Did the interaction feel helpful? Would you recommend GBS? Was the outcome what you needed? Outcome — how did it feel?
XLA in practice — what it looks like when it works
  • Short pulse surveys sent to service recipients after key interactions — not annual exercises that nobody reads
  • Questions designed around outcomes ("Did this resolve your issue?") not process steps ("How would you rate our response time?")
  • Results shared with the GBS team — not just management — so the people delivering the service understand how it is being received
  • Results used for improvement, not for blame. A low XLA score is a signal to investigate, not a disciplinary trigger
  • Discussed in service review meetings alongside SLA data — giving both the quantitative and qualitative view of performance
Monday Move

Ask one stakeholder: "How much effort does working with us cost you?" Log the answer next to your SLA.

Experience measured once is a snapshot. Surveys make it a signal.

Topic 08

Feedback and satisfaction surveys — doing it right

TL;DR

Surveys build trust or destroy it depending on one thing: whether anything visibly changes. Scores are the start; comments and follow-through are the value. The model is in THE FIX.

They filled in the survey.
Then watched nothing change.

2 min read · full theory in the expandable
The Problem
M
Miguel
New team lead · Week 6 · Manila

Miguel inherits last year’s stakeholder survey: decent scores, twenty written comments, zero follow-up actions.

This year’s response rate is half. The message was received: feedback goes nowhere here.
He runs it differently — reads every comment, picks three fixes, and reports back what changed.

"The survey is a promise. Answering it costs them trust — spend it on something."

He feels accountable to twenty comments nobody had answered.

The Trap

You collect feedback and archive it — training everyone that honesty here is wasted effort.

The Fix

Feedback that builds trust runs a closed loop.

READ THE WORDSComments over scores. A 3.8 says little; the comments say everything.
PICK VISIBLYThree fixes, announced. "You said, we did" — named and dated.
REPORT BACKClose the loop publicly. The next survey’s response rate is your grade.

Response rates recover the following cycle — because this time, answering did something.

Feedback and satisfaction surveys in depthTHEORY · 4 MIN

Whether called NPS, CES, or just "the annual survey" — asking for feedback is necessary. How it is designed and used determines whether it builds or damages the service relationship.

What makes a GBS feedback survey useful
  • Ask what the business actually cares about: not what is easy for GBS to measure. If the business is frustrated by query resolution quality, asking about turnaround time misses the point
  • Keep it short: 3–5 focused questions get higher response rates than 20-question forms. The goal is signal, not data volume
  • Make results visible to the team: sharing results only with management creates a feedback loop that never reaches the people who can act on it. Teams that see their own results improve faster
  • Discuss outcomes — do not just report them: a survey result that is presented in a meeting and then filed away changes nothing. Outcomes need to become actions with owners and deadlines
  • Design for improvement, not vindication: a survey designed to prove GBS is performing well will produce results that confirm that — and miss the real gaps. Design it to find the friction
  • Protect the relationship: surveys can damage the service relationship if used for retaliation, blame, or creating adversarial dynamics between GBS and the business. The survey is a tool for getting better — not for settling scores
GBS Insider Club Insights
  • NPS was designed for B2C customer loyalty — it translates imperfectly to GBS. Asking an internal finance team "how likely are you to recommend AP processing to a colleague?" is odd. What matters in GBS is effort (how hard was it to get this done?) and outcome (did the service solve the problem?). Customer Effort Score is often more meaningful for internal service contexts than NPS.
  • The survey result is not the insight — the conversation about the survey result is. A score without a discussion about what drove it produces no action. Build the debrief into the process.
  • If you are a GBS analyst or specialist: survey results that land on your team are feedback about your work. Treat them as developmental data, not as criticism. The organizations that improve fastest are the ones where front-line teams can hear and act on feedback without it becoming political.
Monday Move

Find last cycle’s comments. Pick one fix. Tell the commenters it happened.

JT
FIELD NOTES
  • In every GBS team, there is one person who knows everything and never writes it down. That person is a hero today and a single point of failure tomorrow. If you become the person who documents what they know, you become indispensable — and promotable.
  • SOPs are not bureaucracy. They are the difference between "we can scale this process to 3 new hubs" and "we need 6 months to figure out how it works." The teams with clean documentation move faster on every transition, every audit, and every new hire ramp-up.
? CAREER CHECK click to expand
  • Is process documentation something you see as busy work — or as a career differentiator? How do the strongest performers on your team treat it?
  • Have you ever created or improved a process document that was adopted beyond your team? What impact did that visibility have?
  • Could you articulate your E2E process ownership in a way that a hiring manager would understand? What scope and complexity would you describe?
GBS Insider Club learning paths offer structured career frameworks, practical templates, and guided exercises tailored to your GBS role — from entry-level to leadership.
Glossary

Key terms in this cluster

Underlined terms throughout this page link here. Full cross-pillar glossary at the GBS Insider Club Field Guide glossary.

SOPStandard Operating Procedure — the high-level process blueprint covering scope, inputs, outputs, parties, systems, and controls. Describes what the process does and why. Audience: managers, auditors, new hires.
DTPDesktop Procedure — step-by-step execution manual with screenshots and system navigation paths. Describes exactly how to perform each step. Audience: processors and analysts performing the task.
GPOGlobal Process Owner — accountable for end-to-end process design, documentation governance, standardization, and continuous improvement across all GBS and BU touchpoints for a given process.
XLAExperience Level Agreement — a commitment to measuring and improving how service recipients feel about the service they receive. Complements SLAs by capturing satisfaction, ease of use, and perceived value.
NPSNet Promoter Score — a loyalty metric measuring how likely users are to recommend a service. Scale 0–10. Promoters (9–10) minus Detractors (0–6) = NPS. Originated in B2C; used in GBS with mixed results.
CESCustomer Effort Score — measures how much effort a user had to expend to get a service delivered or a problem resolved. Often more meaningful than NPS in internal GBS service contexts.
Service catalogThe documented menu of services GBS provides — including scope, inputs required, outputs delivered, eligibility, performance standards, and escalation paths. The primary reference for managing expectations and resolving scope disputes.
E2E ownershipEnd-to-end process ownership — accountability for a process from its initial trigger through to its final output, crossing all team and system boundaries. The GPO role is built on this principle.
What this cluster covers — and what comes next
  • Document layers — SOP (what), DTP (how), controls (guardrails)
  • Documentation governance cycle — review, update, approve, train, repeat
  • Global Process Owner — four setup models, engagement vs. change curve
  • Service catalog — what it contains and when it matters most
  • XLAs — experience measurement beyond SLA metrics
  • Service Management — ITIL, COBIT, service transitions — Cluster 2

Want the full breakdown on video?

Process and knowledge management in GBS covered in depth on the GBS Insider Club YouTube channel.

▶   Subscribe on YouTube
Theory done. Now make it count.

Knowing the frameworks is the entry ticket. Applying them — visibly, at your actual job — is what gets you promoted.

The GBS Insider Club Career Playbooks turn this theory into a guided 90-day program for your role: self-assessment, practical exercises, templates, and Julian's unfiltered practitioner playbook.

Explore the Career Playbooks → Back to Operational Excellence
Leave a comment
← Previous Location Strategy and Hub Design Next → Service Management