Reservations can be classified into four canonical (standard) types - All, Active, Cancelled, and No-show.
These data are organized into three dimensions - shop, period, and segment.
Four analysis modes can be used to summarize the data using metrics.
What is a Metric?
A Metric is a statistic or summary measurement of one field (variable) of the data. There are three main categories of metrics: Totals, Rates, and Averages.
A Total is defined as the sum of the values of a particular variable. For example, Total Reservations is the sum of the number of reservations, and Total Pax is the sum of the number of people.
A Rate is defined as the percentage ratio of two sums. For example, Cancellation Rate is the sum of the number of cancellations divided by the sum of the number of reservations and multiplied by 100. Rates exclude walk-ins and deletions and are expressed as a percentage from 0% to 100%.
An Average is defined as a sum divided by a count and can take three forms (Per Period, Per Reservation, and Per Pax) depending on what count is used in the denominator of the average.
- Per Reservation averages use the number of reservations in the denominator. For example, Average Pax per Reservation is the average party size (number of people) divided by the number of reservations, and Average Booking Days in Advance per Reservation is the average number of days prior to the meal that a reservation is booked on a per reservation basis.
- Per Pax averages use the number of people in the denominator. For example, Average Pre-Orders Amount per Pax is the sum of the pre-ordered amount (in currency) divided by the sum of the number of people for reservations that have pre-orders.
- Per Period averages use the number of periods in the denominator. The Per Period averages always require a period that is day, week, month, or year. If the period is week, and the date range has 14 days, then the number of weeks is two (assuming the date range starts at the beginning of a week), and two is in the denominator for the Per Period averages for that query with a weekly period. If the period is month, and the date range is from February 5 to April 18, then the number of months in the denominator of the Per Period average is 3 (February, March, and April) even though the months are not complete.
Note: A Per Reservation or Per Pax average can be grouped by the period to see the average as a time series. However, Per Period averages cannot be grouped by a period because the period is part of the calculation for the metric itself. Therefore, Per Period averages can only be grouped by segment or shop, while Per Reservation and Per Pax averages can be grouped by segment, shop, and period. Furthermore, Per Period averages require a period, but Per Reservation and Per Pax averages do not.
What is a Type?
A Type is a canonical subset of the data. Reservation data have four domains:
- All Reservations: includes all reservations except for deletions
- Active Reservations: includes all reservations that are not cancellations, no-shows, or deletions
- Cancelled Reservations: includes all reservations that are cancellations
- No-Show Reservations: includes all reservations that are no-shows
What is a Dimension?
A Dimension is a primary organizational feature of the data. There are three primary dimensions:
- Shop: the restaurant or shop that generated the data
- Period: the bucketed time period for the data such as day, week, month, or year
- Segment: a particular category by which to classify data. For example, the following segments can be used to classify reservation data: meal, source, channel, day of week, etc
What is Mode of Analysis?
A Mode is a type of simplifying analysis that focuses on only one or two dimensions at a time. There are currently four modes:
- Total Period: aggregates data across all shops and the entire date range to focus the analysis on the segments
- Period Comparison: aggregates data across all shops, but buckets the data into periods within the date range to focus the analysis on periods and segments
- Shop Comparison: aggregates data across the entire date range, but breaks results by shop to focus the analysis on shops and segment
- Adjacent Period Comparison: aggregates data across all shops and the entire date range, but compares results with the period of the same length as the selected date range that came immediately prior and adjacent to it
- Year-on-Year Comparison: aggregates data across all shops and the entire date range, but compares results with the same period in the previous year to focus the analysis on results versus last year and by segment.
What are the Reservation Segment Definitions?
Booking Days in Advance is a field derived from the difference (in days) between the start of the meal and the time the reservations was created. It represents the number of days in advance that a reservation was booked. It can take the following categorical values:
- After (if the reservation was created after the start of the meal. Usually for walk-ins)
- 0-1/2d (created between 0 and 12 hours before the start of the meal)
- 1/2-1d (created between 12 and 24 hours before the start of the meal)
- 1-2d (created between 24 and 48 hours before the start of the meal)
- 2-3d (created between 48 and 72 hours before the start of the meal)
- 3-5d (created between 3 and 5 days before the start of the meal)
- 5-7d (created between 5 and 7 days before the start of the meal)
- 7-14d (created between 7 and 14 days before the start of the meal)
- 14-31d (created between 14 and 31 days before the start of the meal)
- 31d+ (created more than 31 days before the start of the meal)
Cancelled Status is a derived field that can take the following values: No-Show, After (if the reservation was cancelled after the start of the meal - usually for no-shows), 0-1d (cancelled within 0 to 24 hours of start of meal), 1-2d (cancelled within 24 to 48 hours of the meal), 2-3d (cancelled within 48-72 hours before the start of the meal), 3-7d (cancelled within 72 hours to 1 week before the start of the meal), and 7d+ (cancelled more than one week before the start of the meal).
Cancel Reason is the enumerated reason for cancelling a reservation (from a fixed list of values). It is therefore only meaningful for cancellations.
Day of Week is the day of the week for the start of the meal or the time the reservation was created (based on user selection) and can only take the following values: Monday, Tuesday, Wednesday, Thursday, Friday, Saturday, and Sunday.
Day Type is the type of day based on either the day of the meal or the day the reservation was created (depending on user selection). It can only take the following values: Weekday, Weekend, Holiday, and Pre-Holiday.
Time of Day is the hour of the day for the start of the meal or the time the reservation was created (depending on user selection). It takes values like 7:00, 12:00, and 23:00 (11pm).
Meal is the meal in which the reservation starts. It can only take the following values: Breakfast, Lunch, Tea, Dinner, and Night.
Channel is the referrer by which the diner made a reservation at the restaurant. This is typically a third-party marketing company (OTA), social network, hotel concierge, staff intro, etc. Channel is primarily manually determined by staff but can also take values automatically.
Method is similar to Source in that it’s the way/channel by which the reservation was made, but the values are automatically determined by the system. It can only take five values: OTA, TC, Miseban, Walk-in, and Manual. The value is TC if the reservation came from TableCheck, Miseban if the reservation came from voice concierge, OTA if the reservation came from a third-party marketing site, walk-in if the reservation was flagged by staff as walk-in or the diner arrived within 15 minutes of the meal, and manual otherwise.
Pax is the total number of guests in the reservation including adults, children, and babies. It is also referred to as the pax and can only take whole number values of 1 or more.
Purpose is the purpose of the reservation as manually entered by staff. Typical values include Birthday, Business, Family Celebration, Group Date, etc.
Provider is similar to Channel in that it is the referrer by which the diner made a reservation, but it is automatically determined by the system and cannot be manually overridden by users. Typically values include TableCheck, OpenTable, Tabelog, Miseban, etc.
Source is the way/channel by which the reservation was made. It is typically set manually, but there are cases where it is automatically set. Typical values include walk-in, phone, web, in-person, and email.
Status is the status of the reservation that is both manually and automatically determined. Typical values include accepted, no-show, cancelled, arrived, paid, etc.
There are two types of visit history metrics:
- Absolute Visit History counts the number of times the primary customer of the reservation has ever visited the restaurant.
- Relative Visit History counts the number of times within the date range that the primary customer of the reservation visited the restaurant.
These metrics can only take the following values: 1v, 2v, 3v, etc. Visit histories correctly account for customer records that have been previously merged. The first active reservation by a customer is assigned a visit history of 1v even if the meal is at some point in the future. Visit history is counted differently depending on whether the metric is active or not. For active metrics visits are counted only for active reservations (cancellations and no-shows are excluded from the count). However, for any other metrics cancellations and no-shows are counted as visits, and therefore, if a cancelled or no-show reservation was the first time the customer made a reservation, but the customer has no prior active reservations, the visit will count as v1.