Understanding the evolving consumer behind diamond purchases
Debeers came to us as a part of their strategic expansions plan. They were keen to expand across China, USA and India. The research study objective was to gauge an understanding of the local demography and identify if there was opportunity within the Indian market.
1 x Innovation Lead
1 x Business Analyst
1 x Business Intern
1 x CX Expert
Challenge
De Beers, as part of its strategic expansion, sought to enter key growth markets across China, the USA, and India, aiming to rejuvenate its appeal to new generations of diamond buyers. However, the brand faced the challenge of redefining how people connect with diamonds in an increasingly digital, experience-driven world. The diamond category was becoming commoditised, and customers were losing emotional connection to the product. De Beers needed to differentiate its offering beyond sparkle and price, creating an experience that celebrates provenance, authenticity, and purpose — all while being scalable across diverse cultural contexts.
Process
Opportunity Framing
Use case ideation & prioritisation
Stakeholder interviews to understand goals, systems, and constraints
User interviews across buyer, planner, and supplier roles
Industry trends & business model research (e.g., pharma workflows, volatility, regulation)
Persona creation & segmentation
Problem statements & need articulation
As-is journey maps to identify pain points and friction zones
Thematic synthesis of qualitative data
Key insights clustered into opportunity areas
Comprehensive research report delivered to cross-functional team
Rapid ideation workshops & concept sketching
Affinity mapping and desirability clustering
“How Might We” statements to guide solutions
To-be journey mapping with improved workflows and roles
Service blueprint mapping touchpoints, frontstage, backstage systems
Wireframes, interactive prototypes, and interaction flows
Desirability and usability testing with target users
Design iterations based on feedback loops
Cross-functional workshops to validate solution directions
Feedback incorporation from product, engineering, and business teams
Pilot plan with rollout stages
Design roadmap with short-term and long-term features
Foundational Research
Problem Definition
Synthesis & Insight Generation
Concept Development
Experience Blueprinting
UX/UI Design + Usability Testing
Stakeholder Workshop & Pitch
Readiness & Roadmap
Research & Define
We began with a mixed-method study combining maturity assessment and contextual inquiry. The maturity assessment mapped five core processes- demand planning, segmentation, inventory planning, replenishment, and monitoring- to uncover where automation broke down and where human intervention added value. In parallel, qualitative interviews with buyers and ROAR managers revealed emotional and cognitive pain points: time-consuming root cause analysis, hidden anomalies, and lack of real-time visibility. Motivations such as clarity, confidence, and proactiveness guided our opportunity framing. The research clearly highlighted a gap between data availability and decision usability- a space where design could bridge insight and action.
x 3 Buyers (Generic)
Replenishment forecasting
x 2 Buyers (Brand)
Replenishment forecasting
x 2 Buyers (OTC)
Replenishment forecasting
x 2 ROAR Managers
Replenishment forecasting
Relevancy of use case across user groups
Key Research Findings
07
Transfer & Safety Stock Optimisation
Transfers are often done without visibility into cost, geography, or urgency. Contractual thresholds are tracked manually, leading to risk.
We don’t always know the cost of a transfer vs. a new PO… would love to have that in the system.”
06
Inventory Visibility & prioritisation
Buyers lack real-time, network-wide inventory views and struggle to prioritize omits or transfers based on SKU criticality and dollar impact.
“We’re buying at the DC level… we don’t always see national stock, so we just keep placing orders.”
“We prioritise omits by quantity and dollar value, but maybe we’re missing out on more critical drugs.”
05
Trend analysis
Identifying and understanding anomalies like demand spikes is critical but time-consuming. Buyers need better tools for root cause analysis to act quickly and effectively.
“There's just a lot of shopping that goes on in generics... It’s doing that research on what is happening with this item today.”
“It would be easier if we knew the reason for the large spike or the large decrease without having to go in and manually research it.”
04
Slicing & Dicing Data
Buyers want flexible, intuitive tools to slice data across multiple dimensions (e.g., geography, customer, SKU type) to quickly identify and prioritise exceptions.
“I do export and slice and dice to figure out what item caused the drop in service level.”
“I use filters... based on inventory position less than 10 days... we do need a way to narrow this down.”
03
Next Best Action
Buyers want contextual recommendations when anomalies are flagged — not just alerts but suggested actions to resolve issues efficiently.
“I’d love to come in the morning and see: which items are spiking, why, who to contact, and what to do about it.”
“If I could get those actions pre-written — an email to the supplier, an alert to the customer team — that would be perfect.”
“Tell me who to contact or what action to take — not just that something is wrong.”
02
External Factors
Critical external factors (e.g., formulary changes, weather, outbreaks) are often invisible to buyers, causing them to miss key demand shifts.
“Formulary changes blindside us — we only hear about them after the fact.”
“The measles outbreak drove demand crazy... it took time to catch up.”
01
Root cause analysis
Most exception handling is reactive. Buyers need proactive, contextual insights that explain the “why” behind changes in demand or supply.
“Sometimes I’m just staring at a spike and thinking, ‘Okay… now what?’ It’d help to have a prompt.”
“It would be easier if we knew the reason for the large spike or decrease without having to manually research it.”
Ideation & Conceptualisation
We conducted cross-functional ideation sessions, bringing together design, data science, and supply chain SMEs from Accenture’s Dock and Barcelona Innovation Hub. Feature brainstorms translated research insights into tangible opportunities; AI-driven anomaly detection, next-best-action guidance, and proactive alerting systems. These were captured in concept cards, each detailing user value, data dependencies, and business feasibility. Rapid desirability testing with internal stakeholders refined prioritisation, leading to a clear design direction: empowering buyers through visibility, foresight, and autonomy rather than replacing their judgment.
Service & Interaction Design
We mapped the “as-is” and “to-be” service journeys to understand how data, roles, and systems interact across the supply chain lifecycle. This work culminated in a service blueprint connecting forecasting, replenishment, and inventory monitoring around a central orchestration layer. Interaction design focused on surfacing actionable intelligence in context: dashboards revealing network-wide stock patterns, root cause trails linking data to decisions, and simulation tools for forecasting “what-if” scenarios. Each feature was anchored in cognitive simplicity—AI insights presented as transparent, explainable prompts rather than opaque automation.
Half year checks
Exceptions
Demand Data
NLP model
Root cause model inputs
External data sources / newsfeed
Coverage risk flags
Allocation data
Forecast data
Related SKU usage
Inventory position
Supplier contact data
External comms channel integration
NLP model for attribute extraction
Historical demand trends
Forecast data
Event triggers
Customer behavior data
Forecast models
Lead times
Current inventory levels
Service targets
System-recommended buy logic
SKU risk flags
Suggested quantity logic
Action log
Task status tracker
User-specific queue data
Kirk logs into his personalized dashboard, tailored to his portfolio. Actionable widgets surface key insights - from material performance to priority exceptions - showing the most important information first. It’s a clear, guided starting point that helps Kirk focus, decide, and act without the noise.
Login
Review visual summaries of stock levels and forecast accuracy
Review actionable insights / widgets
System highlights next-best-action: Exception handling
After resolving all exceptions, Kirk reviews the planning considerations on the homepage, which highlights materials needing attention based on forecasts and trends. A flu season alert stands out. The platform shows which materials are likely to be affected, using historical data to forecast demand shifts and typical customer buying patterns during this period.
Open Planning Considerations panel
Review flu season alert
See affected materials and forecasted demand shifts
Analyze historical purchasing patterns
At the end of the day, Kirk reviews the system daily summary: exceptions resolved, substitutes ordered, and a new product onboarded. He checks a few flagged tasks for tomorrow - all low priority. With a clear queue and everything critical covered, Kirk logs off confident he’s ahead of what’s coming.
Review system summary of completed actions
Check flagged tasks for tomorrow
Log off with clear queue
Kirk reviews and sends the AI-generated email to the supplier, requesting confirmation of a potential disruption. When the supplier replies, the system automatically extracts and logs key details - confirming an API-related issue and the expected delay. Kirk receives a smart alert summarising the insight, allowing him to move forward without manual review.
Kirk reviews the API issue confirmation and sees that no replenishment is possible. The system surfaces next-best-actions: reallocation guidance for remaining stock, internal notification templates, and suggested flags for related SKUs that might be affected. Kirk confirms the exception as resolved - with root cause documented - and logs any adjustments made. The system updates inventory plans and notifies downstream teams.
Send AI-generated email to supplier
Receive supplier response via email
System extracts and logs issue details (e.g., confirmed API disruption, delay)
Receive automated summary alert
Review next-best-actions
Resolve of exception
Kirk launches the scenario planning tool from the flu season alert to explore a +10% demand increase. The system projects expected coverage, highlights stock gaps, and shows order implications. Two materials are flagged as at risk.
Based on the scenario analysis, the system recommends early orders for the two flagged materials. Kirk reviews the suggested quantities and submits the purchase orders, which are automatically linked to the seasonal planning record - ensuring they’re prepared ahead of the seasonal spike.
Launch scenario planning from alert
Select +10% demand scenario
View coverage, gaps, and order impact
System recommends forward buys for at-risk materials
Review of suggested quantities
Submit POs tied to seasonal plan
Kirk navigates to the exception screen, where AI has already filtered and prioritized exceptions based on where his actions can have the most impact - such as preventing stockouts or reducing overstock. Each exception is clearly tagged, helping Kirk quickly understand and act on the right issues.
Kirk clicks on the top exception to view the material overview. The system highlights a demand spike and low coverage risk. It runs a root cause analysis and surfaces an API disruption, identified through recent news, as the likely cause. AI presents next best actions: draft a supplier email, suggest alternative materials, or run an impact analysis, enabling Kirk to respond quickly and confidently.
Navigate to exception screen
Review AI-prioritized exceptions based on potential Buyer impact
View system-tagged exceptions (e.g., half-year checks, upward/downward demand spikes)
Navigate to material overview
Review exception details
System surfaces root cause
AI suggests next best actions
IDENTIFY INVENTORY RISKS AND EXCEPTIONS
MONITOR LONGER-TERM PLANNING SIGNALS
NAVIGATE TO EXCEPTION SCREEN
UNDERSTAND CAUSE AND REVIEW SOLUTIONS
CONFIRM SUPPLY STATUS WITH SUPPLIER
REVIEW PLANNING CONSIDERATIONS
TEST SEASONAL COVERAGE SCENARIOS
REPLENISHMENT FORECAST USE CASE
The comprehensive journey map follows a Buyer as he navigates his day using his company’s new Gen AI powered platform. Using Gen AI and advanced features, the platform helps streamline his workflow - from handling exceptions and supplier issues to planning orders and onboarding new products. The platform supports faster and more efficient decision making, and reducing complexity in the Buyers day-to-day.
Kirk is a Buyer, working in the Brand Portfolio. He’s responsible for ensuring pharmaceutical inventory is accurate and balanced - meeting demand without overstocking. His role requires close coordination across teams to anticipate needs, manage exceptions, and maintain the right stock at the right time.
HOW TO READ THIS DOCUMENT
AI-drafted email to supplier
Possible alternative material
Understand developing and actualized disruptions to the customer experience in real time.
Navigate to inventory planning
Inventory OH analysis
Review trend analysis
Leverage the Gen AI scenario planning tool to understand the transfer options and rebalance the Stock
Review supplier and Gen AI-created order group for cost and confirm transfer requirements (MOQs etc)
Communication of what changes are happening across functions
Utilise pre-written email
Navigate to Ai tool
Set up of a dedicated AI agent to monitor the material profile and global supply disruptions to plan for an inventory reduction back to normal levels, when disruption is mitigated
Navigate to the scenario planing tool
Move to inventory management screen to make the DCs leaner
Use the What if tool to assess how this could be done
End of day review and close out
Navigate to homepage and then the inventory management
Review prioritized omits
Digests the high-level context for each omit
Review a specific omit along side omit details
Make a decision on what the best course of action is
Increase Safety Stock
Review next best actions and communication recommendations
Anna logs into the centralized “supply chain orchestration command center” that allows her to understand how inventory is moving throughout the network, from inbound supply, across DCs, and out to customers and where risk to key OKRs may be developing/have developed.
The highest priority omit, “developing overstock situation”, shows how a DC is overstocked, more inventory is on order, but the rest of the network is short on supply. The related decision reflects increased inventory levels in response to recent global supply disruptions. The inventory shortage across the rest of the network is at risk of leading to omits. The system recommends proceeding with the order while rebalancing stock and adjusting safety stock levels.
Anna uses the offered link to the Scenario planning tool where she is able to run a simulation of rebalancing existing OH inventory and inbound purchase orders across a MEIO scenario, evaluating transfer metrics (e.g. cost effectiveness, timing, etc). The analysis reveals the opportunity to change the allocation at the NDC of replenishment materials and transfer OH inventory at the DC in scope to rebalance material across the network, ensuring omits can be avoided.
Gen AI gives a suggestion on communication of impact across the different functions of the business. Providing a pre-written email to send to warehouse ops and transportation ops.
The risk of the omit is now mitigated, and moving forward Anna is urged to use a dedicated AI agent to ensure the material is appropriately distributed throughout the network. Despite the “overstock position,” the buyers can shift focus to other prioritized activities for the day.
After resolving the urgent issues, she shifts her focus to inventory planning. She accesses the planning console to optimise incoming inventory and prevent overstock situations. She makes use the What if tool to assess the affects of making their stock of a specific material even leaner, noting that they have maintained 100% SL YTD and the analysis identifies there is little to no risk of optimizing inventory moving forward.
At the end of the day, Anna reviews all created orders and order groups to ensure they were placed correctly. She saves a summary of her daily activities, including inventory optimisation planning, for inclusion in their quarterly report.
Continuing to deal with individual omits, Anna goes back to the home screen and is immediately drawn to the omits that are ranked and prioritised based on dollar value, volume, and therapeutic characteristics. Each omit has a brief description providing immediate context.
For each omit, Gen AI provides a summary potential influences, KPIs affected, and context. The top priority case involves a material with repeated omissions over 6 weeks due to high demand, leading to unmet service level agreements. Gen AI suggests two options: increase safety stock or boost alternative inventory and communicate to customer teams a potential strategy to shift demand to the substitute material code if customers are interested in maintaining full supply
Taking the action of increasing safety stock, the buyer evaluates the % increase in average purchase order quantity that will hit the manufacturers. Based on previous analysis around purchase order variability and an analysis and confirmation of limited supply disruption risk, the system confirms that the supply should be able to fulfil the increase in OH inventory moving forward. A new purchase order is drafted as is communication, offering to give the supplier a confirmation that the increased safety stock level decision is being taken for the indefinite future.
Inventory levels
Service levels
Purchase order data
Customer orders
Inventory parameters
MEIO vs SEIO analysis
Forward Looking Forecast
Purchase orders
External data sources (e.g. N-Tier Visibility Platform)
Forward looking forecast
Planned allocation
Purchase orders
Inventory parameters
Foreward Looking Forecast
Transportation impacts
NDC xDock impacts
Forward looking forward
Replenishment risk due to disruption
Material characteristics (e.g. disruption risks, therapy & patient characteristics, etc)
Omit flags & description
Disruption cause
Customer SL’s / KPIs
Priorities Heuristics
Possible influencer outcomes
SUPPLY CHAIN ORCHESTRATION MANAGEMENT
COMMUNICATE ACROSS FUNCTIONS
STOCK REBALANCING & SAFETY STOCK PLANNING
REVIEW REMAINING PRIORITY OMIT
ACTION BASED ON GEN AI SUGGESTIONS
INVENTORY MANAGEMENT USE CASE
This end-to-end journey maps the process a Buyer follows throughout their day using her company’s new Gen AI-powered platform. Leveraging the platform's advanced Gen AI features, the Buyer navigates through Omits and creates order groups, streamlining their workflow. From creating action plans to adjust MOQs with clients to generating rebalancing plans and identifying specific SKU agents, the platform supports faster, more efficient decision-making, reducing complexity in the Buyer's daily tasks.
Anna is a Buyer, working in the Brand Portfolio. She’s responsible for ensuring her portfolio of pharmaceutical inventory is accurate and balanced - meeting demand without overstocking. Her role requires close coordination across teams to anticipate needs, manage omits, and maintain the right stock across the network at the right time.
HOW TO READ THIS DOCUMENT
Cora
Stay ahead of stock-outs and overstocks with smarter, seamless control.
Simulate disruptions before they unfold and act with clarity, not crisis.
Turn scattered data into sharp, reliable forecasts that guide every decision with confidence.
Desirability & Usability Testing
Low-fidelity prototypes were tested with buyers to validate mental models and assess trust in AI-driven insights. Users responded positively to contextual recommendations but emphasized the need for transparency—wanting to understand why a system suggested an action. Iterations focused on explainability, traceability, and clear information hierarchy. Usability testing also revealed design refinements, such as how exceptions are visualised and prioritised, which directly informed the high-fidelity prototype. This phase was pivotal in transforming abstract intelligence into an interface that felt dependable, human, and intuitive.
Stakeholder Workshop
A multi-day workshop at the Barcelona Innovation Hub brought together client stakeholders, designers, data scientists, and strategists to co-create the future state. Teams collaboratively reviewed service blueprints, prioritised feature roadmaps, and assessed AI integration readiness. The workshop bridged perspectives between technical capability and user desirability—aligning on which Gen AI and Agentic opportunities could scale sustainably. Live walkthroughs of concept flows led to rapid iteration and buy-in, setting the foundation for Cora’s MVP development and enterprise adoption strategy.
Impact
Cora evolved into a proprietary Accenture asset that won a $2million pilot deal that transforms how pharmaceutical buyers interact with data and decisions. It integrates diverse data sources—external, operational, and contextual—into a single layer of actionable intelligence. The design’s impact was twofold: functionally, it improved exception visibility, decision speed, and cross-team coordination; strategically, it shifted Cencora’s operating mindset from reactive replenishment to proactive orchestration. Beyond workflow optimization, Cora also drove talent evolution—defining new buyer skillsets around AI collaboration. Its success positioned the client as a pioneer in human-AI supply chain design, proving that intelligence, when designed around people, can amplify care continuity across the ecosystem.
_gif.gif)
