This article was written for CustomerThink by Greg Kihlström. The Medallia 2026 State of Customer Experience Report reveals a disconnect between how brands perceive the quality of experiences they provide and the consumer reality. While 66% of customer experience practitioners believe experience quality is on the rise, only 17% of consumers agree. This perception gap suggests that many enterprise teams operate in an echo chamber.

What if your AI system is working exactly as designed… and still making your business worse? If you can’t clearly measure AI performance, you can’t confidently say it’s working. Many organizations deploy AI with strong pilots and impressive dashboards but lack a consistent evaluation framework once the system is in production.

In short, "creation sprawl" refers to the AI-driven proliferation of small tools and automations across marketing that outpaces oversight, leading to inconsistency, risk and rework. Let's explore how we got here, and what can be done about it

After decades of "spend-to-learn" advertising, are you still paying for consumer insights that AI can generate for free, before a single dollar of media spend?

What if your “high-performing” AI is quietly dragging down your business, and the only reason you don’t see it yet is because you’re measuring the wrong thing? The latest in the Confident Nonsense series by Greg Kihlström.

This article was written by Greg Kihlström for MarTech. High-maturity brands spend nearly twice as much on GEO as their peers. Here’s what it takes to win visibility in AI search before competitors become the default answer.

Customers care whether switching channels makes them start over. If someone has to repeat the issue, re-enter the same details, and re-authenticate twice to complete one task, your internal handoffs are not staying internal.

This article was written by Greg Kihlström for CMSWire. The company’s innovation roadmap reflects a broader shift in CX from analysis to accountability.

If you don’t own evaluation, you don’t own outcomes. You own activity, which looks great right up until it doesn’t. Vendor dashboards and “model quality” metrics are not the same thing as operational performance across real workflows.

Personalization is what you infer. Customization is what customers choose. If you want durable engagement, treat preferences as a first-class product surface, not a settings page nobody trusts.

This article was written by Greg Kihlström for CMSWire. What failed as consumer hype is reemerging as infrastructure for AI agents, autonomous transactions, and machine-to-machine CX.

Agentic AI in marketing is not about “more content with fewer people,” but about how fast you can respond, how relevant you can be, and how safely you can operate across your most important domains. The meaningful frame for executives is simple: focus on a small, governed portfolio of agentic use cases that clearly move speed, personalization, and control, not on a long list of tools and pilots.

This article was written by Greg Kihlström for MarTech. Salesforce data shows where AI agents drove growth, conversion and efficiency during the holidays — and what marketers should do next.

Have you ever shipped an AI capability that looked great on launch day, then quietly started making your customer experience worse, one “personalized” suggestion at a time?

This is the part of the AI story that rarely shows up in vendor demos. A model can be well-designed, thoroughly tested, and responsibly approved, and still become wrong. Not because the team did anything reckless, but because the world it learned from moved on. Catalogs change. Inventory fluctuates. Promotions end. Policies update. Customer behavior shifts with seasonality and economic pressure. Your model keeps optimizing for last quarter’s reality, while your customers live in today’s.

This is a case written and researched by Greg Kihlström for Sage Business Cases. The pressure to incorporate and effectively utilize artificial intelligence (AI) in organizations is growing. This case discusses how The Hershey Company adopted AI to accelerate innovation in product development and creativity in marketing new and existing products.

This article was written by Greg Kihlström for CustomerThink. If your customer stays with you in a downturn, is it because they love your brand? Or is it because they just don’t have the energy to try someone else again?

This article was written by Greg Kihlström for MarTech. Agentic AI is intermediating the consumer-brand relationship, changing how products are discovered, evaluated and purchased.

AI doesn't correct bad data. It does produce confident output built on whatever you feed it. Because of this, most “AI failures” are actually upstream data governance failures hiding behind polished outputs. Leaders should treat data readiness as a product, with clear owners, thresholds, and controls that prevent confident nonsense from scaling.

This article was written by Greg Kihlström for CMSWire. As customer expectations accelerate, CX leaders are turning to synthetic research and agentic AI to test experiences before customers ever feel the friction.

Customers stay engaged when they can see progress. Open world games like Assassin’s Creed Odyssey excel at this through visible milestones, layered goals, and consistent reinforcement. Brands can apply the same mechanics by adopting clear progress indicators, cross-channel consistency, and reward structures that evolve over time. When customers feel their actions matter, retention, loyalty, and lifetime value all improve.