Lower Funnel Post-Purchase AI Strategy
Driving Value Beyond the Checkout Transforming Lower Funnel Post-Purchase Through Grounded AI Strategy
Summery
Organisational Impact
-
Cross-Functional Alignment: United UX Research, Design, Engineering, and Product Management under a shared AI discovery model.
-
Proactive Engagement: Shifted focus from reactive support deflection to building proactive gear experiences.
-
Support Load Reduction: Automated routine tracking, return, and assembly inquiries to lower support volumes.
-
Agent Productivity: Equipped support reps with AI co-pilots and clear policy citations to cut handle times.
-
Policy Governance: Forced cross-team alignment to create a single source of truth for post-purchase content.
-
Clear Ownership: Defined explicit boundaries between Design/UX and technical execution teams to prevent scope drift.
Role in the work:
-
Lead Author & Director: Co-authored the AI strategy and led the overall Product Design direction for Lower Funnel.
-
Research-First Strategy: Directed the initiative to validate user problems before jumping into interface or solution design.
-
Prioritization: Prioritized Lower Funnel post-purchase opportunities over Checkout and Gifting for Phase 1.
-
Cross-Disciplinary Alignment: Brought together UX Research, Product Design, Content Design, Engineering, and GPMs into a unified AI discovery process
. -
Problem Identification: Focused on fixing fragmented, unreliable content by establishing a single source of truth before automating support.
-
Vision & Concept Creation: Oversaw the transition of research findings into 7 validated AI vision concepts and storyboards.
-
Organizational Boundaries: Defined clear operational scope for Design and UX versus downstream technical, legal, and operational teams.
Financial Impact
-
15%–25%. Saves €1.05M–€1.80M
Automation routine WISMO, return, and setup inquiries to cut contacts by annually per 1M interactions.
-
15%–30%. Saves €250k–€400k
Equiping agents with AI Co-Pilots and quick policy citations to cut handle times
-
Saves €400k–€600k in fees per 50k
Preventing 5%–10% of setup-related returns via New Bike Day diagnostic guides. avoiding returns.
-
€1.2M–€2.0M in GMV per €10M in returns
Converting up to 60% of return requests into exchanges or store credit instead of cash refunds.
-
€1.5M–€2.5M per 500k active customers Automating reorders via Predictive Replenishment to drive a 3%–5% lift in repeat purchases.
-
NPS Lift
Drives a projected +3 to +7 point lift in NPS by removing friction. Cuts customer churn by 0.5%–1.0% to lower acquisition costs.
Context
As GenAI and autonomous agents redefine digital commerce, e-commerce leaders face a critical risk: building technically impressive features that are commercially irrelevant or erode customer trust (the "Apple Touch Bar" trap).
To prevent this, I spearheaded a research-led AI strategy for Decathlon’s Lower Funnel (LF) experience. By focusing on problem validation before solutioning, we prioritized the post-purchase phasewhere clear user friction, high Customer Relation Center (CRC) operating costs, and organizational readiness presented the highest short-term ROI.
Through a two-phase initiative combining global competitive benchmarking, CRC log analytics, cross-functional co-creation, and user concept testing, my team transformed raw customer friction into 7 high-impact, validated AI concepts. This strategy directly targets reduced support costs, higher NPS, improved customer autonomy, and increased Customer Lifetime Value (CLTV).
The Challenge & Strategic Vision
The Problem Space
Customers were losing trust in post-purchase operations not from a lack of support, but from unclear, fragmented, and conflicting information spread across channels regarding returns, tracking, setup, and care. Deploying conversational AI or chatbots on top of unreliable source content risked accelerating misinformation ("AI slop") rather than resolving friction.
Core Strategic Principle: AI should not replace clarity, t should amplify it.
Before scaling automation, we needed to establish a single, validated source of truth for all post-purchase content and policies. AI is was then to be leveraged to scale access, handle historical context, and respond intelligently within controlled, accuracy-preserving boundaries.
Key Business Metrics Targeted
-
CRC Cost Reduction:
Lowering contact rates per order across high-frequency inquiries (WISMO, returns, exchanges).
-
Customer Autonomy & Trust:
Enabling instant, friction-free self-service.
-
NPS & CLTV Lift:
Turning post-purchase moments into continuous brand engagement opportunities.
Project Strategy & Methodology
We executed this initiative across two structured phases:
Phase 1: Foundational Research & Market Benchmarking
How?
-
CRC Log & Data Deep-Dive: Audited customer support tickets, chatbot analytics, and tracking interactions to map core reasons for contact.
-
Competitive Landscape Audit: Benchmarked 20+ global commerce leaders to isolate high-performing post-purchase AI patterns and identify cautionary risks.
Key Research Takeaways:
-
Proactive, accurate delivery prediction (EDD) dramatically cuts "Where Is My Order?" (WISMO) queries.
-
Returns should be designed as a reverse checkout focused on retention (exchanges, instant credit).
-
Automation must always feature a clear, friction-free escape hatch to human agents.
Phase 2: Design, Co-Creation & Validation
I brought together the UX Research, Product Design, Content Design, and Engineering, working with the leads in this ares, we facilitated cross-functional co-creation sessions utilizing:
-
HMW (How Might We) Mapping tied directly to verified CRC pain points.
-
Rapid Crazy 8's Sketching to explore broad solution spaces quickly.
-
Feasibility & Value Mapping to eliminate low-quality "AI slop" concepts early.
Key Solutions: The 7 Post-Purchase AI Vision Concepts
We translated vetted insights into 7 storyboarded concepts designed to transform transactional buyer journeys into long-term engagement loops.
01 - Product Assistant
Pain Points addressed
-
Limited post-purchase support likelihood to return & buyers remorse.
-
Customers don’t know how to get the best use of purchases.
AI use:
GenAI analyses products a customer has bought from Decathlon, uses structured product information to suggest ideal use, training plans or nutrition plans
Opportunity:
Maximum value from purchases to increase satisfaction and reduce returns, ↑ NPS & CLTV in the future.
02 - Maintenance companion
Pain Points addressed
-
Limited post-purchase support ↑ buyer’s remorse & returns.
-
Customers don’t know how to properly maintain products, leading to safety risks and costly repairs.
AI use:
GenAI analyses products a customer bought from Decathlon and uses structured product data to suggest personalised maintenance plans.
Opportunity:
Help customers maintain products correctly to extend product life and reduce returns.
Longer product lifespan ↑ sustainability, satisfaction, trust in Decathlon, and ultimately NPS & CLTV.
03 - Decathlon Coach
Pain Points addressed
-
Customers don't know how best to use products they bought and they are less likely to practise a new sport regularly.
AI use:
GenAI analyses products a customer has bought from Decathlon, uses structured product information to suggest ideal use, training plans or nutrition plans (or all three).
Opportunity:
Help customers get the most from their products and stay engaged in their sport.
Enable proactive nutrition replenishment while positioning Decathlon as a trusted sports companion, increasing NPS & CLTV.
03 - Trip Companion
Pain Points addressed
-
Customers don't know how best to use products they bought and are more likely to return items without relevant advice.
AI use:
GenAI combines structured product data, purchase history, weather and location data to recommend personalised packing lists and relevant gear.
Opportunity:
Help customers prepare for their trips with the right equipment.
Drive complementary product recommendations while strengthening Decathlon as a trusted sports companion, increasing NPS & CLTV.
04 - Delivery Intelligence
Pain Points addressed
-
Customers don’t know when orders will arrive or be ready for collection and often contact Customer Support.
AI use:
GenAI combines delivery partner API data with external factors (holidays, traffic disruption) to provide clearer, real-time delivery updates in natural language.
Opportunity:
Provide clearer and more relatable delivery information.
Reduce Customer Support contacts while improving trust, delivery transparency and NPS.
05 - Predictive replenishment
Pain Point addressed
Customers don’t know when products should be replaced and may continue using worn equipment, increasing injury risk and reducing performance.
AI use:
GenAI combines product lifecycle data, purchase history and activity data from connected apps (Strava, Garmin, etc.) to predict when equipment needs replacing.
Opportunity:
Proactively notify customers when products are nearing end of life and recommend suitable replacements.
Improve performance and safety while driving relevant product upgrades and deeper brand loyalty, increasing NPS & CLTV.
