Forecasting that worksin the real world
Pryor Analytics builds practical forecasting tools powered by Bayesian modelling, helping businesses turn uncertainty into clear, confident decisions. By combining mathematical rigour with real-world operational insight, our models adapt to business data and deliver actionable forecasts that help cafés, restaurants, and other SMEs plan ahead with clarity.
A message to owners and managers
An introduction to Pryor Analytics and our mission to make forecasting more practical, affordable and useful for small businesses — starting with BayesBlend, our café demand forecasting service.
BayesBlend: Demand Forecasting for Cafés
BayesBlend uses Bayesian modelling to help cafés forecast demand more clearly, understand uncertainty, and make better day-to-day decisions.
With clear, readable forecasts, uncertainty ranges, and practical recommendations, BayesBlend helps owners and managers to plan more effectively around stock, staffing and preparation, without needing expensive software or a dedicated analytics team.
Key Benefits
- Bayesian demand modelling that learns from your data
- Clear, readable forecasts with uncertainty ranges for better planning
- Flexible demand measures: BayesBlend can forecast demand using the measures that matter most to your business, such as customer numbers, daily takings, transactions, covers, or items sold
- Practical recommendations for staffing, stock, and preparation
- Weekday, seasonal, and weather-aware pattern detection
- Biweekly reports delivered straight to your inbox
- No expensive software required
- Simple, affordable monthly subscription (FREE during pilot)
- Cancel anytime
- Now seeking early adopters for the BayesBlend pilot
Seeking early development pilot restaurants
Sample Report Coming Soon
BayesTable: Demand Forecasting for Restaurants
BayesTable uses Bayesian modelling to help restaurants forecast covers, understand uncertainty, and make better operational decisions around each service.
By combining historical trading patterns with information such as bookings, walk-in demand, weekday, service period and weather, BayesTable is being developed to give restaurant owners and managers a clearer picture of what demand is likely to look like before service begins.
With clear forecasts, uncertainty ranges and practical recommendations, BayesTable aims to support better decisions around staffing, kitchen preparation, stock and service planning without the need for expensive software or a dedicated analytics team.
Key Benefits
- Bayesian demand modelling designed around restaurant services
- Forecast expected covers separately for lunch and dinner
- Clear, readable forecasts with uncertainty ranges for better planning
- Compare expected demand with usual service levels
- Estimate booked demand and expected walk-in demand
- Identify unusually busy and unusually quiet services
- Weekday, seasonal and weather-aware pattern detection
- Support planning around staffing, kitchen preparation, stock and service capacity
- Clear operational recommendations based on forecast demand
- No expensive software required
- FREE during the early development pilot
Now seeking a small number of restaurants to join the BayesTable early development pilot.
Participants will receive sample forecasts using their own data and help shape how BayesTable develops by providing feedback on which forecasts, insights and recommendations are most useful for day-to-day restaurant operations.
Forecasting made simple.
Provide your data
Upload your historical sales or POS exports — nothing complicated.
Our Bayesian engine models demand
We capture patterns, uncertainty, and real-world variability.
You get clear, actionable forecasts
Easy-to-interpret outputs that support planning, staffing, and ordering.
Why Bayesian?
Because real businesses don't behave like spreadsheets.
Traditional forecasting assumes certainty and fixed patterns. But cafés and restaurants operate in the real world — where weather shifts, events happen, and demand fluctuates.
Bayesian models embrace uncertainty, adapt as new data arrives, and provide a more honest, flexible view of the future. The result is forecasting that's both rigorous and practical.
Hi, I'm Mark
I studied Mathematics at the University of Exeter, with a particular interest in statistical theory and how statistical tools can be applied to real-world decision-making.
Before founding Pryor Analytics, I worked in hospitality, where I saw first-hand how difficult it can be to plan around uncertain demand. Staffing, stock control, preparation, waste, and profitability are all affected by not knowing exactly what is likely to happen next.
That experience led me to build Pryor Analytics: a business focused on turning data into clear, practical insight for owners and managers.
I'm especially interested in Bayesian statistics because it helps quantify uncertainty in forecasts, rather than reducing them to a single fixed estimate. This enables us to provide realistic forecast ranges, helping business owners and managers make operational decisions with greater confidence.
Pryor Analytics combines rigorous statistical training with real operational experience to create tools that are practical, readable, and genuinely useful for small businesses.
University of Exeter
BSc Mathematics
Hospitality Industry
Operational experience
Interested in testing BayesBlend?
We're looking for a small number of cafés to help test and shape BayesBlend, our forecasting service for café operators. The early adopter pilot is free of charge and includes clear demand forecasts, uncertainty ranges, and recommended actions around stock, staffing, preparation, and planning.
By submitting this form, you agree that Pryor Analytics may use your details to respond to your enquiry. We do not sell your data or share it with third parties for marketing purposes. See our Privacy Notice for more information.