Lessons from the Female Founders in Hospitality Summit

Stephanie Smith of Cogwheel recently joined an incredible panel at the Female Founders in Hospitality Summit in New York, moderated by Saadiq Rodgers-King, an AI Transformation Consultant, alongside Kim Bennett of AtlasGuru and Katrina Stalcup of Fora Travel.

Here were a few of my biggest takeaways from the discussion.

Start with the Problem, Not a Tool

One of the biggest reasons AI pilots fail is that companies start with technology instead of a business problem.

“I see this pattern over and over in my consulting work. Organizations say ‘We need an AI strategy.’ But what they actually need is clarity on what problem they’re solving. The intent you bring determines what you get out of it.” — Saadiq Rodgers-King

Across the panel, the most successful examples of AI adoption came from solving clear operational challenges—whether that was scaling travel advisor support, generating travel itineraries, or analyzing hotel marketing performance.

AI is not a strategy; it’s an accelerator. If you don’t know where you’re going, it just gets you lost faster.

Kim spoke to the lessons learned from building a consumer travel platform and evolving it into an AI-powered B2B product. It turned out the bigger business problem to solve was that travel companies are sitting on enormous content libraries with no scalable way to turn them into actionable trip plans with a high-quality, traveler-friendly UX that goes far beyond what a chatbot can offer. The AI use case became much cleaner, the customer much clearer, and the value proposition much easier to articulate. 

“The lesson for us was about not falling in love with the solution you built. Sometimes the best AI lesson is a business lesson first.” — Kim Bennett of AtlasGuru

Data Strategy Comes Before AI Strategy

Another critical lesson: AI is only as good as the data behind it.

“You can’t build an AI strategy on top of messy data. If your data isn’t clean, centralized, and standardized, AI will just amplify the chaos.” — Stephanie Smith

In hospitality, data is often spread across dozens of systems. Before organizations run with AI, they need to walk through the foundational work of data governance, consistent metrics, and standardized processes and documentation. 

“AI without context is just autocomplete. The organizations getting real value are the ones feeding AI with information no one else has.” — Saadiq Rodgers-King

Guardrails Matter (Especially with MCPs)

AI integrations like Model Context Protocol (MCP) can connect AI to real business systems. But with that power comes responsibility.

“Just because AI can take action doesn’t mean it should. The smartest implementations define clear boundaries between what AI suggests and what humans approve.” — Stephanie Smith

Thoughtful permissions and review processes are essential as organizations begin connecting AI to operational systems.

“We’re moving from AI that can answer questions to AI that can take actions. MCP lets AI connect to your actual business systems. That’s incredibly powerful, but it also means you need to be intentional about what you let it do versus what still needs a human decision.” — Saadiq Rodgers-King

AI an Iterative Process – There is no “Easy” Button

All panelists talked about the iterative and training process that AI needs to go through. There is no such thing as build once and be done.

Any agent or chatbot needs more training than you think it will, and needs to be able to learn from a designed feedback loop, to continue to improve.

“There is a false assumption that you can just plug in an AI tool and it works out of the box. The reality of what it takes to train and tune AI on proprietary data is an iterative process.” — Katrina Stalcup of Fora Travel.

Stephanie Smith of Cogwheel shared a time they opened up Claude MCP to their workflow system and it completely rewrote workflows in a matter of seconds. While easy, it was completely wrong and needed the “undo” button.

Make Time for AI Experimentation

One of the most practical recommendations for leaders: give teams (and yourself!) time to experiment.

“If your team doesn’t have time to play with AI, they won’t adopt it. Curiosity and experimentation are what actually drive innovation.” — Stephanie Smith

Organizations that create space for testing, sharing prompts, and learning together will move faster than those waiting for the “perfect” AI strategy.

One of the easiest ways companies can start using AI today is by encouraging teams to create custom GPTs or Gems trained on their internal knowledge.

“With the right culture, AI helps employees solve problems and vibe coding (like Cursor AI) has sped up development and expands their knowledge base which in turn makes them valuable.”— Kim Bennett of AtlasGuru

These internal assistants can help with SOPs, onboarding, marketing planning, and knowledge sharing.

Stephanie Smith, CEO and Digital Matriarch, Cogwheel Marketing & Analytics, HSMAI Marketing Advisory Board Member. Connect with Stephanie on LinkedIn.

Source: View the original article at Cogwheel Marketing.