Optimize your customer journey with AI: where to start and what problems to solve
We've all heard it before: "We should really do something with AI..." Yet despite the excitement and countless proof-of-concepts, many organizations struggle to generate meaningful value from their AI investments. The problem isn't AI itself, it's how we approach it.
At iO, we've learned through dozens of client projects that AI doesn't solve problems by itself. It's an opportunity to solve many problems, but only when we start with the right foundation: understanding real human needs first, technology second.
)
The human-first approach to AI
AI should solve human problems, not create new ones. Take Harry Eriksson, who is deaf, and his neighbour. Using an AI-powered transcription service called TERA, Harry can now have conversations with his neighbour for the first time. As Harry puts it: "It won't always be perfect with AI, but the alternative is doing nothing." This perfectly captures our philosophy at iO: AI needs to be business-driven and human-centred. We've seen too many organizations start with the solution ("Let's build an AI chatbot!") instead of the problem ("Why are customers abandoning their searches?").
Finding the right problems to solve
Before diving into AI, apply design thinking principles. Start by identifying the problem or opportunity, then design a solution to meet your goals. AI is just design material, another tool in your toolkit.
Key questions to ask:
What are your customer pain points through research?
Where do you see employee frustrations in daily work?
What process bottlenecks cause delays?
Is this the root cause or just a symptom?
What's the real cost of not solving this?
Why spend millions on AI chatbots when your real issue is unclear product information? We use various problem discovery methods: UX research, impact mapping, design sprints, and exploratory workshops to help clients identify the right challenges before building solutions.
Pro tip: Score your ideas using frameworks like the impact/effort matrix to identify quick wins versus major projects.
Real-world success stories
Stadsmissionen: 2.5x faster product uploads
Stadsmissionen wanted to increase e-commerce sales by 200%, but their manual method for uploading second-hand products severely limited capacity. We built an AI-driven solution where staff simply upload images and handwritten notes. The AI automatically identifies product attributes such as type, color, condition, warehouse location and populates all CMS fields.
Impact: Staff now complete product uploads 2.5 times faster than before, enabling the scale needed for their growth goals.
WWF: Automated accessibility compliance
WWF had approximately 9,000 images missing alt-texts, making accessibility compliance challenging. We created an AI-powered dashboard where editors can select up to five images at once, review AI-generated alt-text suggestions, and approve or edit them before automatic publication to WordPress.
Impact: Happier editors who gain time for higher-value work while improving website accessibility.
Using AI to discover problems
AI excels at detecting patterns in large datasets. You can leverage this to identify issues in your customer journeys:
Traffic analysis: Export data from GA4, run funnel analysis through ChatGPT, and get observations on user behavior across key website sections.
Support ticket analysis: Process large volumes of support issues to identify common pain points and address them proactively.
Content gap analysis: Analyze search data, competitor content, and customer questions to identify missing content opportunities.
Automated audits: We built an AI-powered audit tool for Husqvarna's B2B portal that analyzed thousands of pages across multiple languages, scoring each against strategic criteria and generating specific recommendations for editors.
Impact: 85% reduction in audit time compared to manual methods.
Where AI creates value in customer journeys
Based on our client work, here are proven high-value, low-risk starting points:
Awareness phase
Problem: Users overwhelmed by generic marketing noise
Solution: AI-generated, personalized marketing content
Proven impact: 70% reduction in content creation costs (source: SalesHub, 2025)
Consideration phase
Problem: Prospects struggle to find relevant product information
Solution: AI chatbots for personalized product recommendations
Proven impact: 23% higher conversion rates (source: Glassix, 2024)
Decision phase
Problem: Sales teams spend too long drafting proposals and quotes
Solution: AI-assisted proposal and quote generation
Proven impact: 30% faster deal closures (source: McKinsey, 2024)
Purchasing phase
Problem: Too many cart abandonments due to checkout friction
Solution: Conversational checkout assistants
Proven impact: 20% reduction in cart abandonment (source: Walmart, 2024)
Onboarding phase
Problem: New users get stuck and open support tickets
Solution: Personalized AI-generated onboarding guides
Proven impact: 40% fewer support tickets during setup (source: Zendesk, 2024)
Learning from failures
Not every AI project succeeds. We worked with a global MedTech company to build an AI-driven site search for healthcare professionals. Despite following best practices: validating the solution, writing acceptance criteria, ensuring compliance, the project was shelved.
What went wrong:
The company lacked AI maturity for customer-facing generative AI
IT wasn't fully onboard from the start
Legal concerns arose around the EU AI Act
No clear scaling plan for the expensive technology
Key lessons:
Do early risk assessment and avoid high-risk problems initially
Ensure IT and legal involvement from day one
Work with experienced partners who understand AI implementation challenges
Getting started: practical next steps
Start here (high value, lower risk):
1. Customer support/service: Deploy retrieval-augmented chatbots connected to your knowledge base to handle Tier 1 questions, cutting support costs by -30% while boosting customer satisfaction.
2. Product consideration pages: Implement lightweight, product-database-powered chatbots on high-traffic pages to handle feature comparisons and "which option fits me?" queries.
Essential steps for success:
Always start from real human customer needs, not technology
Plan for data, compliance, and stakeholder buy-in upfront
Tie each AI use case to concrete business outcomes
Build internal AI skills so teams can iterate independently
Focus on quick wins: high-time, low-value tasks first
Think long-term: agentic AI that acts autonomously is coming
The AI-native mindset shift
Most companies won't become "AI-first," but we can adopt an AI-first mindset. This means shifting from "How do we scale humans?" to "How do we scale decisions, creativity, and action with machines?" For Volvo Penta, we built a custom AI assistant trained on legal guidelines that gives copywriters real-time compliance feedback for sustainability claims. This transformed a manual, error-prone review process into an efficient workflow where content is compliant from the start.
Impact: Weeks of work saved and potential compliance issues avoided.
For Husqvarna, we created an AI-driven content playbook generation process that turned interviews and research into actionable guidelines, achieving a 50% reduction in information gathering effort and 75% reduction in content writing time.
Ready to optimize your customer journey?
AI implementations can feel like "hurry up and wait" situations. But when done right and starting with real problems, involving stakeholders early, and building incrementally, AI delivers transformative value.
The key is identifying where AI can solve genuine human problems in your customer journey, then building solutions that put experience first and technology second. At iO, we blend expertise across marketing, data, AI, infrastructure, and operations to help ambitious brands bridge the gap between customer expectations and digital capabilities. We've guided dozens of organizations through successful AI implementations, from quick wins to enterprise-scale transformations.
Sources: Based on presentation "Optimize your customer journey with AI" by Christofer Falkman, AI Strategy Lead at iO Nordics, and case studies from iO's client work with Stadsmissionen, WWF, Worldline, Volvo Penta, Husqvarna, and Polestar.