We often get asked about comparisons of data made against different sources.
For example;
- graphs
- data exports
- within BRM itself (on the Reservation List Panel)
The problem is the data often doesn't correlate (at first glance). Here we examine why that is:
- Month-overlap - when reservations start in one month and end in another.
- What to include - for example, cancelled bookings or not.
- Technical constraints - BRM has a limit on how many reservations it can return within BRM (300).
- Changes - if you change the dates of a reservation after you've taken a statistical snapshot.
- Occasionally a bug - e.g. 30 day months not behaving consistently.
Let's take a real-world example
(taken from an actual BRM store, but anonymized)
Reservations for the month of June.
First of all, let's look at the actual data;
| BRM - June | BRM - June (+cancelled) | BRM - custom 1-30th June | Graphs | Export | |
| no. rentals | 153* | 167* | 145 / 158 | 144 | 158 |
| total rental income | 7,065.00* | 7,129.00* | 6,475.00 / 6539.00 | 6,397.00 | 6,539.00 |
* It seems that in BRM when you search for June(30 days) it includes reservations starting on 1st July. (bug)
[issue ID BRM-4982 we are working on a fix]
Then let's dig in to understand the various sources of data;
| BRM | BRM (+cancelled) | Graphs | Export | |
| include cancelled?** | optional (default no) | yes | no | yes |
| dates | reservation STARTS on June 1st > 30th | |||
** When a reservation is cancelled it has its price set to zero.
Reservations for the month of March
(smaller data set is easier to compare)
| BRM - March | BRM - March (+cancelled) | Graphs | Export | |
| no. rentals | 43 | 56 | 43 | 56 |
| total rental income | 2,369.00 | 2,422.00*** | 2,369.00 | 2,422.00 |
Reservations for the month of July
(now a bigger data set)
| BRM - July | BRM - July (+cancelled) | Graphs | Export | |
| no. rentals | too many **** | too many | 271 | 309 (30 cancelled) |
| total rental income | too many | too many | 19,729.00 | 22,117.00 |
*** It appears that when refunds occur the price (which has been refunded) is re-instated post refund. (even though the reservation's price is set to zero on cancellation.
There are two further dimensions to consider, which add to the complexity.
- When is the data 'frozen'? - it is not. If you modified a reservation 2 months later (price or date) - then it would be altered.
- **** Physical limits on pulling back Reservation Data into BRM - 300 reservations.
Transactions
Then there is another complication. What about comparing reservation data with transactional data?
This is, of course, a very different thing!
Transactions may occur at different dates - even after the reservation has 'ended'.
Reservations may also be paid for through multiple transactions.
Reservations may not be fully paid - or may be overpaid.
All of these things mean you may get a very different answer when comparing or reconciling payments against
See Also