VarietyIQ Staff
VarietyIQ helped Literati both improve demand forecast accuracy by 25% and sharpen its allocation strategy. Together, these changes are projected to drive a 5%+ seasonal net-sales lift without additional inventory. The result: higher revenue and better capital efficiency from smarter use of existing stock.
Executive Summary
Literati, a fast-growing children's literacy company, is scaling its nationwide book fair business. To succeed, each school must receive an assortment aligned to students’ interests — all while managing limited inventory that must be allocated across thousands of fairs.
VarietyIQ partnered with Literati to improve both personalization and allocation intelligence. Within just one fair season, we delivered three significant enhancements:
- An improved book demand forecast solution, increasing predictive accuracy by 25% (backtest accuracy).
- New optimization-driven allocation methods, projected to drive 5%+ seasonal net-sales lift without requiring additional inventory.
- A streamlined code foundation, improving runtime and memory efficiency while also improving long-term maintainability.
These improvements deliver a two-fold benefit: increased revenue from better product-market fit, and reduced capital requirements by making the most of existing inventory.
VarietyIQ's team brought deep retail and machine learning expertise that let them ramp up incredibly fast onto our complex and unique business and technical problems. They're easy to work with, laser-focused on driving real impact, and they built solutions that are both maintainable and ready to scale with us.
About Literati
Literati is a children's book company on a mission to ignite a love of learning and a love of life. The company believes that every child can develop a love of reading when they’re given the right books at the right time. Literati focuses on helping kids grow into their best selves by building reading skills and making the discovery of great books easy and inspiring.
Their rapidly expanding school book-fair program brings curated book assortments directly to schools — supporting literacy, fundraising, and joyful reading experiences for families and educators.
The Challenge
At its core, allocation is a global optimization challenge. There is a finite inventory pool, and any unit sent to one location is no longer available elsewhere. The objective is to maximize the total impact of that limited inventory by distributing units intelligently across fairs.
Literati’s allocation challenge is further complicated by its dynamic nature. Each book-fair season spans several months, meaning inventory that does not sell at one location may return and become available again before the season’s end. Effective personalization helps the business understand true demand for each title at each fair, and the allocation system then helps make the best possible use of the available units in this dynamic environment.
As its fair business continued to scale, Literati chose to invest in a concentrated push to maximize performance — sharpening prediction accuracy, improving allocation outcomes, and streamlining core components to accelerate future development. They tapped VarietyIQ to lead both the R&D and the implementation of these enhancements.
The Solution
We collaborated with Literati’s data and operations teams across two core workstreams, integrating directly with existing pipelines to ensure smooth adoption and zero disruption to ongoing processes.
Demand Personalization
We enhanced the ML models predicting a school’s title demand by leveraging:
- An expanded feature set with richer historical purchasing behavior and external indicators of book popularity, strengthening school-level demand alignment signals.
- Improved model architectures, identified through extensive backtesting and comparative evaluation.
- An end-to-end pipeline rewrite to improve memory efficiency (enabling larger datasets), long-term maintainability, and deployment flexibility.
Result: A 25% improvement in predictive accuracy across backtests, strengthening confidence in title-level recommendations.
Optimization-Based Allocation
We introduced a scalable optimization framework that addresses two key challenges:
- Dynamic sell-through optimization — We improved how the models account for early-season returns and their impact on inventory availability across fairs, enabling more efficient sell-through of titles by end of season.
- Substitution-aware allocation — We incorporated the influence of substitute titles directly into allocation decisions, allowing better use of both primary and substitute inventory to drive stronger overall performance.
Simulation results suggest that these improvements could together drive a 5%+ improvement to seasonal net sales. This lift requires no additional investment in units, but instead comes simply from smarter allocation decisioning.
Results
With better personalization and optimized allocation planning, Literati is positioned to:
- Ensure more students find books they are excited to take home.
- Increase revenue without increasing inventory.
- Continue scaling fair coverage efficiently with their small but mighty ML and planning team.
The enhanced approach strengthens how the system performs and will continue to improve as their fair volume grows.
Conclusion
Through enhancements to both personalization and allocation, Literati is now positioned to get more of the right books into the hands of eager young readers. We strengthened predictive accuracy, improved how finite inventory is deployed across a dynamic multi-month season, and streamlined the supporting code to ensure long-term stability and speed. All of this drives meaningful commercial benefit without requiring additional inventory.
VarietyIQ will continue to partner closely with Literati as the book-fair business expands, supporting planned growth and advancing toward a fully optimized allocation system. We’re excited to continue this collaboration, improving outcomes for schools, families, and students nationwide.
Why VarietyIQ
- Retail-focused ML and allocation expertise.
- Proven partnership with existing planning and engineering teams.
- Ability to deliver measurable ROI without requiring additional headcount or inventory.
VarietyIQ helps retailers make smarter allocation decisions — so every unit counts.
Send us a message to learn how optimization-driven planning turns current inventory into measurable growth.
With thanks to Daragh Sibley for his partnership at Literati.
