New Feature: Introducing District-Level School Holiday data for better demand visibility and forecasting

Published on October 04, 2021
Peter Jansen
Head of Product

PredictHQ’s school holiday category gets even more granular

There are many school holidays globally and they will impact your demand drastically if you aren’t aware of them. School holidays cover events like spring and summer break, as well as national holidays, for both primary and secondary schools throughout the U.S. and the U.K.  As families are more likely to depart for vacation, these holidays can cause a drastic incremental or decremental change in your business' demand.

School holidays have a massive impact on demand for accommodation, retail, transportation, and tourism. School holidays in the US and UK cause hundreds of thousands to millions of people to travel and spend money across the year. Family trips and holidays during the school holidays have a big economic impact and drive demand for many businesses.

If your business revolves around serving children and families or is located within a short radius of a school, tracking school holiday data should be a key part of your planning or forecasting. Whether you’re a coffee chain where parents visit after school drop-off or you’re a hotelier that sees a big spike in demand when students are on break, knowing when students will be on a break ahead of time can significantly alter your plans as a business

The problem with tracking school holidays is that they can vary drastically across countries, states, regions, and even districts. In order to solve this challenge and simplify the process, we’re excited to announce that we now track school holidays at a district level for both the U.S. and the UK. 

Why tracking holidays at a district level is necessary

We have made our school holiday event category more powerful. This release means that we have increased coverage and granularity of school holidays in the US and the UK by tracking school holidays at a district level. To give you context, we previously had 250-300 school holidays for the US annually and we now have 56,000-57,000 total school holidays annually across the US. 

District-level coverage means that different types of businesses can be aware of upcoming national and local holidays using the different features like holiday start and end times and the number of students affected per district.

  • A waterfront restaurant on the Jersey Shore beachline can anticipate increased demand and staff more employees to maintain quick service and generate more orders

  • A chain of coffee shops discovers that while destination locations in Miami and the Florida Keys are ramping up, their suburban stores have dropped in demand as parents are no longer going for coffee after drop-off’s

  • A boutique hotel knows they’ll experience an influx of demand, but needs to understand the start times of school holidays for the districts with the largest student populations nearby to appropriately price their rooms

District-level coverage can take you from a generalized suggestion of when school holidays are happening across an entire state, down to an acute understanding of how school holidays affect your locations.

The new advancements also include changes to how we were ranking school holiday events. We now rank all school holidays by student numbers on a logarithmic scale. What does this mean? Depending on the size of the district’s student population, the size of impact will differ. For example the Summer Half Term Holidays in Kent, UK affects 257,694 students and has a rank of 98. The Winter Holidays in Manchester affects 95,432 students, therefore has a slightly lower rank of 90. The changes to rank for school holidays means that you can have a much better understanding of which school districts will have the highest impact, therefore helping you figure out where to focus your efforts. Here is a refresh on how our ranks work if you’re interested in learning more. 

Remove manual effort so you can scale your visibility into demand drivers

Let’s walk through what this really means for you and your team. Districts can have holiday dates that differ significantly from neighboring districts with similar student counts. For example, the City of London breaks for Summer on July 7th, while neighboring North Somerset breaks for summer on July 22nd - a difference of over 2 full weeks.

Manual tracking and planning for fluctuations in demand becomes incredibly difficult and impossible to scale. PredictHQ provides a centralized place to leverage data for thousands of school holidays and the corresponding number of students they affect, ensuring that your business stays ahead of demand months in advance. Should any of them change, the events would be updated in our system immediately. And this is just one category of 19. The improvement of this event category, along with integrating other types of events into your strategies will ensure you and your team have the right visibility into upcoming external factors that may impact your demand. 

How to use PredictHQ school holiday data to better anticipate demand

Whether you’re using PredictHQ strictly for event awareness or if you have our demand intelligence data integrated directly into your forecasting models, district-level school holidays can be leveraged to improve your planning, strategies or forecasts. 

Let’s start with how you can surface up this data in our Control Center. Within control center, you can do a quick search for school holidays within a specific area of interest. Here we’ve pulled school holidays for Manchester in the UK.

Control center screenshot of school holidays at a district level in the UK

Now maybe you are a quick service restaurant and you have pizza shops distributed across the UK. Your locations closest to schools experience consistent demand when schools are in session. Think families who grab pizza for dinner after sports practices or order pizza for takeout, etc. You can easily search across different districts in close proximity to your locations to ensure that you are planning to have the appropriate staff on site or the right amount of stock for the expected demand. Now it's equally important to understand when schools are on holiday to reduce the amount of staff and stock to account for the decrease in expected demand. 

For another example, business traffic around schools is driven substantially by morning routines based on when families take their kids to school. During the week I might frequent a specific Starbucks that’s close to my child’s school, but during a holiday break, I’ll visit a Dunkin’ Donuts location that’s more convenient to my home. People's day-to-day schedules are connected to where their kids go to school and district school holidays can help you better plan as a business. 

For data scientists and companies who plan to use this data directly into your forecasting models, you can easily call the API to incorporate school holidays. Also, you can export data from Control Center to explore the data and for R&D. To find school holidays that impact your location you can use the standard features of our events API - as follows:

  • Perform a lat/long and radius search using the within parameter on the API. For this find the lat/long of your location such as a store, hotel or any other location. Search for events around that location - such as 5 miles around a location. 

  • You can use the places parameter in the events API to find events impacting a geographic location - for example all events impacting Bristol or all events impacting New York City. 

  • Alternatively, if you are downloading the data into a data lake you can use our location scopes from the place_hierarchies field with the places hierarchy endpoint to retrieve events that impact a geographic location.

  • For school holidays for the US we have polygons showing the geographic area covered by the school district. See the geo field in the events API. The same methods work for searching for events around your location for US school holidays. But we provide the additional information on the exact geographic area covered by the school district with the polygon information.

School holidays is one of many demand-drivers that can be used directly in ML models to improve demand forecast accuracy. See our data science guide on unattended events for Jupyter notebooks and resources to get you started with forecasting. 

School holidays at a district-level provides a more accurate picture into expected demand. With up to four years of historical data and one year future looking, businesses large and small can pick and choose timeframes that matter most to them to understand pinpointed impacts of school holidays before they happen for better forecasting and operational planning. Sign up for a free trial or reach out if you’d like to learn more!