Startup Playbook: How We Used Dynamic Pricing to Decode Customer Behavior at NaHora.com
- Maria Alice Maia
- May 5
- 3 min read
Every empty seat on an airplane is a perishable good. The moment the cabin door closes, its potential value drops to zero. The central idea behind NaHora.com, the first last-minute ticket sales company in Latin America, was to build a market in those final, fleeting hours of value decay.
This wasn’t just about offering discounts. That’s the easy, and lazy, way. A fixed 20% off for every unsold seat leaves a massive amount of value on the table. The real challenge, and the true opportunity, was to decode the complex behavior of the last-minute traveler and build a system that could price their intent in real-time. This is the story of how we used data to measure what seemed immeasurable and, in doing so, created a market from scratch.
The Challenge: Pricing a High-Stakes Impulse
The core problem was one of extreme uncertainty. Who is buying a plane ticket for a flight that leaves in six hours? A student going to a music festival on the last-minute? An executive in an emergency? A couple on an impulsive weekend trip triggered by discounted price? Each of these situations implies in a radically different willingness to pay. A one-size-fits-all pricing strategy would fail them all—and our business.
To succeed, we had to move beyond guesswork and build a model from the ground up. Our central hypothesis was that demand for last-minute travel was not a monolith; it was highly elastic and predictable, if only we could find the right signals to measure. We needed to decode the subconscious drivers of that last-minute purchase.
The Playbook: Measuring Elasticity to Build an Algorithm
This became our obsession, following a rigorous cycle of hypothesis, testing, and iteration.
Decoding "Value" with Demand Elasticity: Our first step was to build a model to measure demand elasticity for these specific tickets. We collected data not just on flight routes and historical prices, but on the external triggers we believed were influencing purchase decisions. Our model included variables like:
Time Decay: The number of hours (not days) until departure.
Contextual Triggers: Day of the week, proximity to holidays, and major events (concerts, conferences, festivals) in the destination city.
Route-Specific Demand: The inherent popularity and historical load factor of a given flight path.
Building the Dynamic Pricing Engine: The elasticity model was the brain of our operation. It fed a dynamic pricing algorithm that constantly adjusted the discount based on the real-time context. A ticket to Florianópolis on a rainy Tuesday in May had a very different elasticity—and therefore a different optimal price—than a ticket to Salvador on the Friday before Carnaval. The algorithm wasn't just a pricing tool; it was an automated system for decoding human behavior at scale.
Iteration and Growth: We were relentless in testing. Every transaction, every user search, every abandoned cart provided new data to refine the model. This data-driven feedback loop became the engine of our growth, allowing us to get progressively smarter at matching the right price to the right customer at the perfect moment.
The Results: From Model to Market
The results validated our approach. In just the first 12 months of bootstrapped operation, NaHora.com:
Reached 100,000 users and BRL 1.5 million in revenue.
Sustained a 28% month-over-month growth rate.
Earned recognition that validated our innovation, including the Prêmio Santander de Empreendedorismo, Startup Chile, and Mulheres Tech Sampa.
After reaching these milestones, we completed a successful exit of the business.
The Primary Lesson
The enduring lesson from NaHora.com is one that has defined my entire career: you don’t simply find product-market fit. You often have to create it by decoding the underlying needs and behaviors of your customers with data. Our dynamic pricing algorithm did more than just sell leftover seats; it created a new, reliable, and efficient market where one hadn't existed before.
This same mission—to measure the immeasurable and replace intuition with evidence—is what drove my work building People Analytics at Ambev and what now fuels my research into public policy. The context changes, but the core principle remains the same.
If you are an entrepreneur, product manager, or data scientist wrestling with how to use data to build something from nothing, I invite you to join my email list for more practical, no-nonsense playbooks.
And if you have a specific growth or data challenge you’re facing, let's talk. I'm opening up a few 20-minute consultation slots to help you decode your own path to growth.
