What this help guide covers:
Sales Forecasts
Sales forecasts are generated from your historical sales data and form the foundation for demand-based rostering in Tanda. These predicted sales are then used to calculate a recommended number of staff to roster, helping you match labour to demand and avoid over/under-staffing.
You can also adjust predicted sales to reflect real-world factors like public holidays, weather conditions, or special events, ensuring your recommended staff counts stay accurate even when trading conditions change.
To learn about importing sales data to Tanda, refer to our help guide Configure your Sales, Revenue & Demand Data (Data Streams).
Forecasting Methods
Tanda’s primary forecasting method is Average of Past Dates, which uses sales data from the previous three weeks to estimate future demand. If you already use an external forecasting process, you can also choose to apply your own sales forecasts instead.
Check your forecasting method in Settings > Integrations > Demand Forecasting:
Some POS integrations automatically import a predicted sales value alongside actual sales data. When these values are available, Tanda will use them in place of the selected forecasting method.
If you prefer to provide your own forecast, you can import predicted sales via CSV, SFTP, or API by selecting Use Your Forecast on this page.
To modify your predicted demand data, see Modifying Predicted Sales in the Roster below.
Configuring Recommended Staff Counts
Recommended staff counts use a ratio system to determine how many staff should be rostered based on the revenue amount. Think of it as '1 staff member in this team can handle this amount of sales in a 15-minute period'. Once the value entered has been exceeded for that 15-minute interval, an extra staff member will be recommended.
To set this up, follow the steps below:
Navigate to Settings > Integrations > Roster Ratios and click Add Ratios.
Choose the teams that you are setting up this ratio for and choose the stat type. Teams will only appear here if they are attributed to an existing Data Stream.
Choose how many sales should be received to recommend rostering an additional staff member in that team. This is for each 15-minute period.
In the above example, for a 15-minute segment with less than 50 sales, 1 staff member for each team will be recommended. For a 15-minute segment with more than 50 sales but less than 100 sales, 2 staff members will be recommended for each team, and so on.
The minimum number of staff that will be recommended per location is the number of different team joins configured.
Entering a value of 0 will not recommend any staff.
💡 Tip: If you’re finding it difficult to set the right ratios when configuring recommended staff for the first time, a useful approach is to go to a day where staffing aligned well with sales, and adjust the team join amounts until the recommended amounts are similar to the actual roster amounts.
Viewing Forecasted Sales and Staff Counts
Sidebar & Weekly Views
Overall predicted sales and recommended staff hours can be viewed in the sidebar of the roster.
Statistics such as Total Hours & (Recommended Hours) and Sales Forecast & (Wage % Revenue) compare your overall roster cost against predicted revenue and the recommended number of hours.
Configure which stats are shown in the sidebar by clicking on the panel to open Roster Stats > Configure Key Stats.
You can also view the Total Hours/Recommended Hours stat at the top of each day:
You can configure which stats are shown under each day by clicking on the date and configuring the Pinned Stats:
Roster: Daily View
Viewing the roster in Daily View will give you a clear view of your predicted sales and recommended staff in a single day.
In this view, by default, you'll see a graph showing the predicted sales across the day, the recommended number of staff (if configured in Roster Ratios), and the actual number of rostered staff for each 15-minute segment in the day.
Additionally, you can view the total recommended vs actual number of rostered staff hours for each team on the schedule below the graph, when grouped by team:
The Day View also allows you to switch the graph to a table for further customisation of the number of recommended staff.
For each of the teams configured in the data stream, click the settings icon to open the Cognitive Creator Settings and adjust recommended staff counts, using metrics such as:
Minimum/maximum number of staff at once
Minimum/Maximum shift length
Time it takes to open/close (i.e., roster additional staff before/after sales are predicted)
For each team, see the recommended number of staff at specific times in the day, with under-rostered marked in red, and over-rostered marked in blue.
Modifying Predicted Sales in the Roster
In some cases, predicted sales may not accurately reflect expected trading conditions due to factors such as weather, major sporting events, or outlier days like public holidays. To account for this, you can modify predicted sales directly in the roster by increasing or decreasing them by a set percentage, or by selecting alternative historical dates to use for the sales average.
In the roster header, open Actions > Demand Predictions.
From there, choose the date that you would like to adjust the predicted sales for.
If you know sales are going to be generally higher or lower on a specific day, you can make use of the Modifier to adjust the predicted sales amount:
Either enter a specific amount in New Total, or adjust the percentage modifier.
Ensure you click Save and Close to apply your changes to the roster.
Live Insights
You can also use the Live Insights widget in the dashboard to see your actual and predicted sales data against the rostered and actual staff counts for the day:
Learn more about the Live Insights widget in this help article.