Graduate researcher and Shiplake Properties develop intelligent, demand-informed pricing system for Toronto apartment market
Project at a Glance: Setting the right rent for each apartment means understanding not just what makes a unit valuable, but when residents are actively searching. Jillian Menikefs built a scientific pricing system for Shiplake Properties that incorporates market demand patterns, seasonal rental cycles and unit-specific attributes that optimize asking rents. By applying advanced revenue management techniques borrowed from hotels and airlines, the system helps Shiplake capture value during peak demand periods while remaining competitive year-round.
When Timing Meets Technology: The Pricing Challenge
Data tells an interesting story about Toronto’s rental market. Search interest for apartments peaks sharply every summer and drops significantly each winter — a pattern that has held consistent over more than a decade.
Yet most apartment buildings set their asking rents without accounting for these predictable demand fluctuations. A unit listed in January faces a very different market than the same unit listed in July, but traditional pricing approaches don’t adapt accordingly.
Shiplake Properties has managed premium urban rental communities across Toronto for more than 70 years. With thousands of units spanning iconic buildings like the Torontonian at Yonge and Eglinton and modern developments like Lillian Park at Yonge and Davisville, the fourth-generation family business has decades of leasing data.
But their existing pricing model, built in-house years ago, failed to account for market dynamics, both seasonal patterns and the way individual unit attributes interact with changing tenant preferences.
The old system had multiple problems: inflexible formulas that couldn’t adapt to market changes, statistical issues with how variables interacted, and most critically, a trust problem. Leasing teams increasingly questioned the model’s recommendations. When frontline staff doesn’t trust your pricing tool, you face both a technical and organizational challenge.
The company needed a system that could learn from recent lease signings, incorporate market demand signals, adjust for seasonal patterns and update attribute weights as tenant preferences evolved. In the rental market, this is fundamentally a revenue optimization problem: what’s the right price for each unit at each moment to maximize long-term value while keeping occupancy strong?
The Team: Academic Expertise Meets Real Estate Innovation
Menikefs brought training in applied mathematics to her MScAC internship at Shiplake Properties. She completed her undergraduate degree in mathematics and engineering and has prior internship experience in financial quantitative modelling and solutions architecture. She is passionate about applying her technical knowledge to build and deploy machine learning models that address real-world challenges.
“I’m motivated by using math and data to improve everyday decisions and solve problems that impact people in my community,” she said. “Shiplake gave me the opportunity to do that in the city I’ve called home all my life. While the rental property sector was new to me, Shiplake’s collaborative culture and my MScAC training allowed me to be confident in my abilities and contribute meaningfully to different aspects of the business.”
The partnership was established through the MScAC internship program, with Menikefs supervised at the University of Toronto by Professor Mary Pugh from the Department of Mathematics. At Shiplake, she worked under industry supervisors Peter Brimm and Hank Latner.
The role offered high visibility with senior management through bi-weekly meetings with the head of leasing, chief operating officer (COO), vice president of finance, president, and chairman. These brainstorming and check-in sessions supplemented by external industry experts from multiple countries ensured her work was understood and contained the collective experience of the leadership team regarding attributes and leasing processes. The tight-knit team environment shaped Menikefs’s experience beyond the technical project, and executive meetings provided insight into real estate strategy. She effectively received a crash course in rental property fundamentals that shaped the ways in which the data could be interpreted.
Shiplake President Peter Brimm noted, “At Shiplake, one of our core values is innovation, and while we specifically look for team members that embody a commitment to innovation, it is important to us to partner with young, talented and ambitious people like those at the U of T MScAC program. Their fresh eyes and academic rigour help us focus on what will be most impactful.”
Building an Intelligent Demand-Informed System
Shiplake lacked a dedicated technical team for model development, giving Menikefs complete creative freedom while requiring the system to be understandable to users.
She drew methodology from revenue management research developed in hospitality and transportation, adapted for apartment rentals. Menikefs employed supervised learning to identify which unit attributes drive tenant decisions, unsupervised learning to discover hidden patterns in leasing data and reinforcement learning to develop a pricing policy that optimizes revenue while maintaining operational constraints.
Time-series forecasting captured seasonal rental patterns, quantifying what Google Trends data revealed: demand fluctuates predictably throughout the year. Regression and econometric techniques identified which attributes drive willingness to pay. The critical 60-day notice window — when residents signal their intention to leave — became a key pricing touchpoint within the unit life cycle, creating decision moments where pricing directly impacts revenue.
Menikefs accessed decades of historical lease data including signed leases and rejected offers. This rejection data revealed what prices the market wouldn’t accept, providing crucial boundaries for optimization.
The system weighs dozens of attributes: bedrooms, floor plan, window orientation, balcony size, floor level and more. As seasonal demand shifts or tastes change, the model adjusts automatically. Competitor pricing data ensured recommendations remained competitive within Toronto’s localized rental markets.
Critically, the system was designed to earn trust from leasing teams. Their frontline observations about resident preferences provide qualitative insight that pure data analysis misses.
Transforming Daily Operations: The Results
The new pricing model delivers demand- and seasonality-aware recommendations on a suite-by-suite basis, accounting for unique unit characteristics while responding to market signals. Unlike the previous rigid system, it continuously incorporates recent lease data and external supply and demand information, automatically refining its understanding of market preferences.
“Rather than provide a ‘black box’ system, Jillian also spent time with stakeholders designing an output that made the pricing both transparent and understandable,” said Brimm. “As a result, the team has adopted the recommendations more readily and have been able to provide clear actionable feedback for model improvements.”
The framework operates across the entire unit life cycle — from notice through the 60-day notice window, vacancy and move-in. This perspective means pricing decisions account for full revenue implications.
For a portfolio of 1,833 units across multiple buildings, systematic and data-driven pricing represents significant operational improvement. Leasing agents now work with recommendations grounded in comprehensive market analysis.
Professor Pugh commented, “Jillian’s work is especially interesting because, unlike in dynamic pricing models used by platforms like Uber or travel websites — where one immediately sees when a customer clicks away because a price is too high — the information available from prospective renters is far more one-sided.”
“This creates complex and fascinating modelling challenges that require the ability to think of and test multiple fundamentally different approaches,” she said. “That flexibility of thought, and the willingness to explore varied solutions, are core aspects of rigorous mathematical training.”
A Culture of Innovation
Founded in 1948, Shiplake Properties is a fourth-generation, family-owned real estate firm in Toronto’s residential rental sector.
As a purpose-built rental owner-operator, Shiplake maintains direct control over the resident experience through professional on-site management, a 24/7 Resident Services Desk, a resident portal, and participation in the City of Toronto’s RentSafeTO: Apartment Building Standards Program.
“Shiplake has a generational culture of innovation. Each generation has made different efforts to understand the ever-changing real estate landscape,” said Hank Latner, chairman. “At some points, innovation was about new construction techniques or products. Today, it’s shifting more to the adoption of big data products and creating competitive advantages with artificial intelligence.”
What This Means for Real Estate
This project demonstrates that revenue management techniques proven in hospitality and transportation can adapt effectively to residential real estate. As rental markets become more competitive and data-rich, property operators need tools that respond to demand signals and market timing.
The methodology, combining time-series analysis, machine learning, and reinforcement learning with operational constraints and leasing expertise, offers a template for other property management companies. The emphasis on interpretability and trust addresses a critical challenge: user adoption and confidence.
For Toronto’s rental market, more sophisticated pricing could contribute to market efficiency. When prices reflect demand and supply conditions accurately, both operators and renters benefit from clearer market signals.
By the Numbers
- 2,000 units in the GTA
- 70+ years in Toronto’s multi-family residential market
- Supervised, unsupervised, and reinforcement learning methods used
- Strong seasonality observed: apartment search interest peaks in summer and drops in winter
Contact: For media enquiries, please contact MScAC Partnerships at partners@mscac.utoronto.ca. For more information about Shiplake Properties, visit www.shiplake.com.
