Revised September 2026: the Klarna case now includes the company’s 2025 course correction.
In brief: Products succeed only when Users, Buyers, Deciders and Influencers each have a reason to support them. This briefing helps leaders prepare an AI initiative in the terms each group needs—and warns why cost metrics alone are not enough.
AI may change how products get built. It does not change the gate every initiative passes through: the right people saying yes. In the PathPatron Compass, those people are your Targets:
- Users — must adopt the product in daily life or work.
- Buyers — must put money or budget behind it.
- Deciders — must prioritise it against competing bets.
- Influencers — must advocate for it when others hesitate.
Targets don’t just consume products. They define whether products launch, scale, or survive — and AI has raised what each of them expects. Users want real relief from real pain, and can tell an AI feature from a solved problem. Buyers want ROI they can defend, whether they’re external customers or the internal budget holder. Deciders weigh every proposal against strategy and resilience. Influencers shape the narrative around your initiative in rooms you’re not in.
Four cases
Users — Duolingo Max. When Duolingo shipped GPT-4-based conversation practice in 2023, adoption held because the AI interactivity was wrapped in the reinforcement loops users already valued. The AI was the ingredient, the product was still the product.
Buyers — Klarna, in two acts. Klarna’s AI assistant handled 2.3 million chats in its first month — the workload of roughly 700 agents — and internal Buyers funded it on immediate, measurable savings. By May 2025 the same Buyers were funding a correction: the CEO publicly conceded that cost had become too dominant a factor in evaluating the service, at the expense of quality, and Klarna began piloting human agents for the cases where judgment decides whether a customer stays. Bloomberg reported the 2025 change; Klarna disputes the “reversal” framing — the assistant covers more work than at launch, and the human pilot started small — and that dispute is itself part of the lesson. Volume and cost metrics won the original budget; the absence of equally rigorous quality metrics forced a correction whose very size is now argued about. Ask Buyers to fund both kinds of measurement from the start, so nobody has to argue afterwards about what the numbers meant.
Deciders — IKEA. IKEA greenlit AI demand sensing because it tied directly to fewer stockouts and to revenue — future-proof value in the language of strategy, not novelty.
Influencers — clinicians and diagnostic AI. In hospitals piloting diagnostic AI, budgets were approved and rollouts still stalled, because clinicians raised trust concerns that leadership couldn’t wave away. Influencers can amplify trust or quietly end an initiative, and they often hold no formal authority at all.
The same pitch, four languages
Suppose you propose an AI onboarding assistant. Users care that it removes drop-off friction. Buyers care that reduced churn is worth a specific number of euros. Deciders care how it positions the company against competitors over five years. Influencers care that the GDPR question has a real answer. Give all four the same pitch and each hears the part that worries them; resistance compounds. Tailor it, and the same facts land four times.
Targets and Transformers
Targets decide what should be done; Transformers decide how it gets done — and their trust is linked. Testers and Creators validate what Users will adopt. Owners and Organizers make the promises Buyers hear credible. Maintainers and Implementers carry the risk answers Deciders demand. And Influencers are often Transformers themselves, amplifying execution concerns — or successes — upward. Fail one group and you usually lose both.
AI can help you understand Targets faster: simulate stakeholder reactions before a pitch, cluster feedback to see what each group actually values, map perspectives side by side. Treat all of it as preparation for the real conversations, not a substitute. In AI-driven products, Targets don’t just influence adoption — they define survival.