The future of AI
in your contact centre

You asked the right question. But the answer starts further back than the contact centre.

What Your Customers Think

Voice-of-customer research revealed patterns across your brands. Here are three that summarise the experiences your customers talk about most.

The Nectar refund that breaks

A customer returns a faulty Argos product paid for with Nectar points and a debit card. The refund pushes points to a cancelled card. Argos says it's a Nectar issue. Nectar says it's an Argos issue. The customer waits months.

+

The return that depends on who's working

A customer returns an Argos item at a Sainsbury's store. One location accepts it immediately. Another refuses the same product, same packaging, same policy. The outcome depends on which staff member they get.

+
To be validated

The substitution nobody checked

A customer orders beef steaks. The fulfilment algorithm sends beef-flavoured dog treats. 32% of online orders contain a substitution. The logic that picks replacements cannot tell human food from pet food. The contact centre handles the complaint. The algorithm keeps running.

+

The Solution: Total Experience

In most organisations, these four disciplines evolved independently. Customer experience sits in one team, employee experience in another, digital experience in a third. Total Experience is the framework that would connect them into one system.

Customer Experience

A customer moves between Sainsbury's, Argos, Nectar, Tu, Habitat, and Smart Charge and is recognised as the same person everywhere. Their history, preferences, and loyalty status travel with them. Every interaction starts with full context already loaded.

+

Employee Experience

An agent sees the full picture across all brands on one screen. Nectar profile, Argos order history, grocery delivery status, all in one place. They have the tools and authority to resolve multi-brand issues in a single conversation.

+

User Experience

A customer resolves their own issue without calling. The app works. Search returns what they need. Self-service is intuitive and complete. When something goes wrong, the digital journey handles it before a call becomes necessary.

+

Multiexperience

A connected washing machine bought from Argos detects a fault and books its own repair through the Sainsbury's ecosystem. A customer moves from app to chat to phone and never repeats themselves. Every touchpoint, including the product itself, carries full context.

+

What Your Business Faces

Four pressures compounding at the same time.

£1bn

Margins are tightening

Sainsbury's has invested over £1 billion in price reductions across four years to match the discounters. Operating margins are under constant pressure. The room to absorb operational inefficiency across the group is shrinking every quarter.

70%

Competitors are moving

M&S improved contact centre routing accuracy by 70% with AI. Tesco personalises for 13 million Clubcard users at scale. Ocado cut 1,000 roles through AI productivity gains. John Lewis now tops the UK customer satisfaction index.

18m

Nectar data is underused

18 million active users. 260 million personalised offers generated every week. This data powers pricing and marketing brilliantly. It barely touches the contact centre. The richest customer dataset in UK grocery is underused where it matters most.

To be validated
£6.17

The cost of standing still

Sector average cost of a single voice contact. Voice accounts for over 80% of volume. The April 2025 National Living Wage increase adds over £615 per agent per year. Agent attrition runs above 30%. Every quarter without change costs more than the last.

Why Now?

Four shifts are happening at the same time, and together they make Total Experience unavoidable.

78%

Customer expectations have reset

of consumers expect seamless cross-channel experience

+
40%

The technology now exists

reduction in cost per interaction with AI augmentation today

+
60%

The agent experience is broken

of agent time spent searching for information instead of helping customers

+
87%

Leadership is paying attention

of CEOs now see customer experience as a direct driver of revenue

+

The Path to Convergence

Three forces must mature together for Total Experience to work. Technology moves fastest. Organisational convergence is the constraint.

Values represent relative maturity on a 0-100 scale.

AI Autonomy

The evolution from scripted chatbots to self-governing AI agents.

+

CX/EX Convergence

The unification of customer and employee experience into a single discipline.

+

Omnichannel Fluidity

The shift from disconnected channels to seamless, context-preserving interactions.

+

Sector Evolution Map

Every sector is moving toward Total Experience at a different pace. Pick a sector to see where it stands today, what changes in two and five years, and how ready it is for TX.

SECTOR OUTLOOK

Retail & E-Commerce

2-Year Outlook

AR virtual try-ons and staff mobile hubs syncing with online profiles. Staff see customer wishlists in real time, enabling hyper-personalised floor service alongside predictive inventory routing.

5-Year Vision

Zero-click commerce. Fully autonomous AI personal shoppers negotiate and purchase on behalf of the customer. Physical stores evolve into experience showrooms with biometric checkout as standard.

Innovations & Disruptors

TX Maturity Index

Two-year (inner) vs five-year (outer) maturity across six TX dimensions

2-Year 5-Year

The Size of the Prize

Five metrics show the commercial impact of Total Experience maturity across three stages: where most organisations sit today, what changes by 2028, and what full maturity delivers by 2031.

MetricToday20282031Change
Customer Retention Rate +70%81%91%+30%
Employee Tenure (avg years) +1.52.64.3+187%
First Contact Resolution +71%80%92%+30%
Cost per Interaction +$7.75$6.00$4.15−46%
Customer Lifetime Value +1.0x1.2x3.0x+200%

The TX Operating System

Total Experience needs an operating system. Three layers: the people who deliver the experience across Sainsbury's, Argos, Nectar, Tu, Habitat, and Smart Charge, the processes that connect them, and the technology they run on. All three must mature together.

PEOPLE
Customer-Facing Workforce
+
Operations & Fulfilment
+
Technology & Digital
+
Commercial
+
Corporate Functions
+
Leadership & Management
+
PROCESS
Service Resolution
+
Sales & Customer Growth
+
Product & Experience Design
+
Fulfilment & Delivery
+
Corporate Operations
+
Insight to Action
+
TECHNOLOGY
Contact Centre
+
Digital Platforms
+
Physical Infrastructure
+
Fulfilment Systems
+
Core Business Systems
+
Data & Intelligence
+
The contact centre is where all three layers meet the customer.

Why All Three Layers Matter

When one layer advances without the others, the system breaks in predictable ways. These are the three failure patterns.

Technology without People

You build the platform but nobody can use it. Advanced AI routing with undertrained agents creates faster misrouting. The technology works. The outcomes don't.

+

People without Process

You hire great talent but trap them in bad systems. Skilled agents fighting broken escalation paths, disconnected tools, and conflicting KPIs burn out faster than average ones.

+

Process without Technology

You design elegant workflows on legacy platforms. The process logic is sound but execution is manual, slow, and impossible to scale. Every efficiency gain hits a technology ceiling.

+

OS Maturity: Four Stages

Stage 1
Siloed

Each layer is managed separately with different goals, metrics, and leadership.

Stage 2
Connected

Data flows between layers. Shared metrics emerging. Gaps visible but fixed one layer at a time.

Stage 3
Orchestrated

Changes in one layer trigger planned changes in the others. The ceiling rule is understood and acted on.

Stage 4
Converged

One operating system. Friction in any layer triggers adaptation across all three. This is Total Experience.

The Maturity Gap

Organisations invest in technology first. Process and people follow later, if at all. The gap between them is where transformation programmes fail.

Values represent relative maturity on a 0-100 scale.

Technology

Moves fastest because it is the easiest to buy. Buying capability and assimilating it are two different things.

+

Process

Follows with a lag because it means dismantling decades of operational debt.

+

People

Trails furthest because cultural change moves at human speed, not machine speed.

+

The Contact Centre

Sainsbury's Group operates across grocery, Argos, Nectar, Tu, Habitat, and Smart Charge. The contact centre sits at the intersection of all of them. Within the TX model, five pillars define how it operates.

AI-Powered Automation

Nectar balance queries, delivery slot changes, Argos order tracking handled autonomously. AI reads intent, retrieves context across brands, and resolves without a human. Volume reaching agents drops. Every remaining interaction gets better.

+

The Super-Agent

The human who handles a customer whose Christmas Argos order failed, grocery delivery was wrong, and Nectar points are missing. One call, three brands, full resolution. AI handles the routine. Humans handle the moments that matter.

+

Real-Time Intelligence

Every contact generates a signal. Substitution complaints feed grocery product teams. Delivery failures feed Argos logistics. Nectar friction feeds programme design. The contact centre becomes your best sensor for what customers actually experience.

+

Preemptive Service

The system detects a grocery substitution the customer will reject and offers alternatives before the delivery arrives. An Argos parcel is delayed and the customer receives a message before they notice. The best service is the one the customer never has to ask for.

+

Built-In Trust

Nectar holds data on 18 million customers across grocery, Argos, and partner spending. As AI uses this data to personalise service, every decision must be transparent, explainable, and compliant with UK GDPR and FCA requirements.

+

The Three Horizons

Three horizons show how the contact centre matures from where it is today to where it needs to be in two years and five years.

The Fragmented Multi-Brand Contact Centre

To be validated

Separate Systems, Separate Brands

Pega handles 2 million complex cases annually. Engage Hub AI routing has cut AHT by 33%. But agents still toggle between Sainsbury's, Argos, Nectar, Tu, Habitat, and Smart Charge systems. Customers report being bounced between teams when an issue spans brands.

+
To be validated

Volume-Driven and Reactive

Voice accounts for over 80% of contacts at £6.17 each. The operation is measured on speed and cost. Customer Effort Score across the sector has dropped to 63%, meaning getting help takes too much work even when the outcome is good.

+
To be validated

Nectar Data Disconnected from Service

260 million personalised offers generated every week for marketing and pricing. Agents can see complaint history through Pega, but the full Nectar profile, price sensitivity scores, and predictive basket data are not yet powering the service interaction.

+

Multi-Brand Friction Everywhere

The returns lottery across Argos desks inside Sainsbury's stores. Nectar refunds pushed to cancelled cards. Customers told to call a different number for a different brand within the same group. Agents lack authority for cross-brand resolution.

+
The pace of that maturity depends on five capabilities that follow: the technology, the people, the intelligence, the service model, and the trust framework.

The Augmented Contact Centre

The contact centre has connected its data, deployed AI co-pilots across all brands, and begun measuring outcomes rather than handle time.

To be confirmed

AI Co-Pilots Across All Brands

The Microsoft Copilot rollout extends from the current 3,000 store colleagues to contact centre agents. Real-time knowledge surfacing, automated summaries, and draft responses. New agents reach competency faster with AI beside them from day one.

+
To be confirmed

Nectar Data Reaches the Agent

The Snowflake CDP connects to the agent desktop. When a customer calls, the agent sees their Nectar tier, recent purchases across all brands, personalised offer status, and points balance. Service becomes contextual for the first time.

+
To be confirmed

Cross-Brand Resolution

Unified customer identity means the agent handles the grocery complaint, the Argos return, and the Nectar query in one conversation. No transfers between teams. No "you need to call the other number." One call, full resolution.

+
To be confirmed

Predictive Routing and Proactive Outreach

If Nectar data shows an abandoned basket or a failed SmartShop scan minutes before a call, the IVR predicts the reason and routes directly to the right specialist. Delivery delays trigger proactive SMS before the customer calls.

+
The pace of that maturity depends on five capabilities that follow: the technology, the people, the intelligence, the service model, and the trust framework.

The Autonomous TX Hub

By 2031, the Sainsbury's Group contact centre is unrecognisable. Agentic AI handles the majority of interactions autonomously across all brands.

To be confirmed

Autonomous Resolution at Scale

AI agents handle Nectar queries, delivery tracking, slot changes, Argos order status, and standard complaints end to end. No human involvement for routine contacts. The volume reaching human agents drops dramatically.

+
To be confirmed

The Sainsbury's Super-Agent

Humans handle only the interactions that genuinely require a person. A customer whose Christmas was ruined across grocery and Argos. A vulnerable customer. A complex multi-brand complaint. The role is unrecognisable. Compensation reflects the skill.

+
To be confirmed

Preemptive Service Across the Ecosystem

The substitution algorithm detects a mismatch and offers alternatives before delivery. A connected Argos appliance flags a fault and books its own repair. Nectar data predicts churn and triggers retention before the customer considers leaving.

+
To be confirmed

The Intelligence Engine

Every interaction feeds real-time insight to grocery product teams, Argos merchandising, Nectar programme design, and the executive board. The contact centre is no longer a cost line. It is the group's primary source of customer truth.

+
The pace of that maturity depends on five capabilities that follow: the technology, the people, the intelligence, the service model, and the trust framework.

The Super-Agent

As AI takes over routine volume, the human role transforms into three distinct tiers. Today shows what the workforce looks like. Two years and five years show what it needs to become.

To be validated

Task Volume Distribution

Sector benchmark. Percentage of total contact centre volume handled by category. Sainsbury's actual split to be confirmed.

Routine (AI) Complex (Co-Pilot) High-Stakes (Human)

Real-Time Intelligence

Every interaction generates data that the contact centre can act on. Today shows what is being used. Two years and five years show the intelligence it needs to be producing.

The Sainsbury's Group contact centre handles more real-time customer interactions than any other function in the business. The 22 billion rows in the Nectar CDP, the Pega case data, the Engage Hub routing signals. The opportunity is moving that intelligence from the queue dashboard to grocery product teams, Argos merchandising, and the executive board.

Technology Foundations

The first capability. Every domain on this page powers something on the pages that follow. Today shows where each one stands. Two years and five years show where they need to be.

Preemptive Service

The most powerful service interaction is the one the customer never needs to make. Today shows how far this has got. Two years and five years show what it looks like when it works.

Reactive Service

The customer discovers the problem, contacts Sainsbury's, and waits for a resolution. Every step adds cost and erodes trust.

To be validated

Customer discovers issue

The grocery delivery arrives with beef-flavoured dog treats instead of beef steaks. The Argos fridge is marked "delivered" but never arrived. The Nectar points have gone negative after a return. The customer has to notice it themselves.

+
To be validated

Customer initiates contact

They call the Argos line and wait. The phone cuts off after extended hold. They try social media. They call a different number for a different brand within the same group. The effort falls entirely on them.

+
To be validated

Explain and repeat

They describe the problem to an offshore agent who lacks systemic permissions. They get bounced between Argos and Nectar teams. Neither can override the other's system. Context is lost at each handoff.

+
To be validated

Eventual resolution

The agent offers the next available delivery slot, often a week later. Or suggests click-and-collect at a store miles away. Or the customer gives up and initiates a Section 75 chargeback, permanently severing the relationship.

+
Reactive service is expensive, slow, and corrosive to trust. At an estimated £6.17 per voice contact, every inbound call is a signal that something upstream failed.

Built-In Trust

As AI handles more interactions, trust determines whether customers adopt, stay, or leave. Today shows the current state. Two years and five years show what needs to be in place.

Sainsbury's holds one of the largest first-party data assets in UK retail. The Nectar CDP, the Pega case history, the Engage Hub routing data. Customers who trust the group with that data will stay. Those who lose trust will leave and take their lifetime value with them. No amount of technology investment compensates for lost confidence.

The Evidence

The five pillars deliver measurable results. Current state numbers are sector benchmarks. Sainsbury's actual data to be confirmed.

Shift in Interaction Models

As AI matures across Sainsbury's, Argos, Nectar, Tu, Habitat, and Smart Charge, the balance between reactive, proactive, and preemptive service shifts dramatically. The baseline is a sector estimate. The trajectory shows where the Sainsbury's Group should be heading.

To be validated
The 80% reactive baseline is a sector estimate. Sainsbury's has already shifted some volume through Engage Hub (500,000+ automated enquiries) and app-based self-service for low-value grocery refunds. The actual reactive/proactive/preemptive split across all group brands to be confirmed.

TX Value Impact (Index)

Sector evidence indexed to 100. Investment in employee experience and technology drives measurable improvement in customer experience, while operational costs per contact decline. This pattern holds across industries and is the basis for the Sainsbury's business case.

Note the strong correlation between the rise in EX driven by new tools and the subsequent trailing rise in CX. The investment sequence matters. Sainsbury's Microsoft Copilot rollout and AI graduate programme represent the early EX investment. The CX improvement follows.

The Metrics That Matter

Current state numbers are UK retail sector benchmarks. The projected state shows what the TX programme should deliver. Once confirmed, the left bars become Sainsbury's actual numbers and the gap between left and right becomes the business case.

To be validated
CSAT 72% is Sainsbury's own Which? in-store rating. FCR 65% is a sector estimate (UK retail best-in-class is 70-75%). Agent retention 70% reflects the sector average 30% attrition rate. AHT 8 minutes is the upper cross-sector range (retail baseline is 3-4 minutes). Once confirmed, all four become Sainsbury's actual numbers and the gap to the right becomes the business case.

The Challenges

The technologies that power the TX vision also create new vulnerabilities for the Sainsbury's Group. As AI, automation, and data flows scale across Sainsbury's, Argos, Nectar, Tu, Habitat, and Smart Charge, three structural risks emerge.

Generative Fraud and Deepfakes

Voice cloning and synthetic identities are scaling faster than defences. The Argos high-value appliance line and the Nectar points balance are both high-value targets. What Sainsbury's has deployed for caller authentication and deepfake detection is unconfirmed. The ICO mandates explicit consent for voice biometrics with a genuine alternative.

+

The Privacy Paradox

Customers demand hyper-personalised service but are increasingly hostile toward the data collection that enables it. AI needs data to deliver better journeys and experiences. Customers do not want to share it. The DUAA 2025 has liberalised automated decision-making, but the ICO is actively enforcing. Balancing personalisation with privacy is one of the hardest design challenges in the TX model.

+

The Empathy Deficit

As AI absorbs routine volume, 100% of the human agent's day becomes emotional, complex, or escalated. The easy calls disappear. What remains demands a level of resilience and skill that most organisations have yet to invest in. The sector attrition rate is already 30%. The agents who stay need fundamentally different support, training, and compensation.

+

AI Threat Scenarios by Brand

Each Sainsbury's Group brand creates a different attack surface when autonomous AI takes over customer interactions, refunds, and proactive service. Six brands, six threat landscapes.

BRAND-SPECIFIC AI THREATS · 2028-2031

Sainsbury's

Algorithmic Manipulation and Computer Vision Exploitation

ACROSS ALL BRANDS

The Privacy Paradox

Every AI capability described above requires customer data to function. Customers demand personalised, frictionless service and simultaneously resist the data collection that enables it. This tension applies equally to every Sainsbury's Group brand and intensifies as AI autonomy increases.

ACROSS ALL BRANDS

The Empathy Deficit

As AI absorbs routine volume across every brand, 100% of the human agent's day becomes emotional, complex, or escalated. The easy calls disappear. What remains demands a level of resilience and skill that requires fundamentally different support, training, and compensation structures.

Threat Landscape and Mitigations

Threat assessment based on sector research and documented incidents. Bubble size reflects breadth of exposure across the group. Click a bubble to see which mitigations address it.

Direct AI Attack Vector Systemic Exploitation Workforce Impact
ACTIONABLE MITIGATIONS

Zero-Party Data Strategy

+

Human-in-the-Loop AI

+

AI Model Governance

+

Adversarial and Agentic Defence

+

Infrastructure Modernisation

+

Workforce Resilience Framework

+

The Roadmap

Six capability pillars assessed at three milestones on the journey to the five-year vision. The same six dimensions, measured at today, two years, and five years. The distance between where you are and where you need to be is the work.

Today

Where does each pillar stand right now against the benchmark? Green means it meets the standard. Amber means it partially exists. Red means it is absent. This is the baseline that everything else is measured from.

Two Years

How much of the two-year journey is covered by confirmed plans? Each capability is either on the roadmap or it is not. The ratio of confirmed to unconfirmed tells you how well the system is being planned as a whole.

Five Years

The same assessment at the full vision horizon. The bar is higher. Capabilities confirmed for two years may fall short of what the five-year vision demands. What was green at two years could be grey at five.

The six pillars assessed across all three milestones: Technology Foundations, Super-Agent, Real-Time Intelligence, Preemptive Service, Built-In Trust, and AI Threat Mitigations. The first five build the capability. The sixth defends against what that capability introduces.

Roadmap Assessment

Six capability pillars benchmarked against where Sainsbury's should be today. Green means the capability meets the benchmark. Amber means it partially exists. Red means it is absent. All scores are based on external research and internal assessment.

The Starting Line

The way organisations deliver experience is changing fundamentally. Those that move first will define the standard. Three things to know before you start.

The shift has started

AI is already reshaping how organisations interact with their customers and employees. Expectations are resetting across every channel and function. Your competitors are moving. This is a now decision.

+

The plan must be whole

Technology, people, and process belong in one strategy. Organisations that design them separately will watch them fail separately. Start with the full picture.

+

The delivery is iterative

Large-scale, big-bang programmes fail at twice the rate of incremental approaches. Structure delivery in 90-day cycles. Prove value early. Scale only what the evidence supports.

+

Five Traps

The technology usually works. These five patterns are what kill transformation programmes.

01

Leading with technology

Buying platforms before defining what the operation needs to become. The technology works, but nobody uses it the way it was intended.

+
02

Measuring the old world

Tracking AHT and cost-per-contact while expecting agents to deliver empathy and lifetime value. The metrics fight the strategy.

+
03

Piloting without a path to scale

Running proof-of-concepts that prove the concept but never connect to a plan for the rest of the operation.

+
04

Treating AI as a cost play

Deploying automation to reduce headcount instead of to elevate the quality of every interaction. Short-term savings, long-term erosion.

+
05

Ignoring the human cost

Expecting agents to absorb wave after wave of change without investing in their skills, their wellbeing, or their career progression.

+

The Next Step

The roadmap has identified where Sainsbury's stands today and where the gaps are. Some capabilities are on track. Others are missing entirely. The AI threat landscape adds a new dimension that requires immediate attention. Before any of this can become a plan, the assessment needs to be validated.

Validate the maturity assessment

Every score in this roadmap is based on external research. The red, amber, and green ratings reflect what we can see from the outside. The roadmap coverage reflects what we have been able to confirm. A structured assessment validates these scores against reality, confirms what is genuinely in flight, and quantifies the gaps that need to be addressed.

The output is a validated baseline across all six pillars, a confirmed picture of what is on track and what is missing, and a prioritised set of actions that connect directly to the two-year and five-year vision.

This is the foundation everything else is built on. The strategy, the pilots, the scaling plan. None of it holds without a validated starting point.

Next Steps

Every organisation's starting point is different. These three moves, taken in sequence, reduce risk and build the momentum needed to scale.

01

Know where you stand

Technology. How close to cloud-native? Where does AI sit today?

People. Equipped for empathy-led roles or still measured on speed?

Process. Optimising for AHT or for customer lifetime value?

The gap sets every priority that follows.

+
02

Prove it in one place

Scope. One operation, one geography, one team.

All three tracks together. AI copilot, new agent roles, CLV metrics.

Measure. What actually moved? What didn't? Why?

Build the evidence before asking for enterprise commitment.

+
03

Scale what works

Governance. A structure to expand what the pilot proved.

Cadence. Coordinated rollout across vendors and regions.

Specialist capability. Analytics, CX design, AI toolkit.

The $407B market belongs to those who can do this repeatably.

+

Next Step

Content to be developed.

The future is already here.
It's yours to shape.