Visma.net ERP
Transaction matching settings
In Visma.net ERP, you can tailor the settings of the automatic matching process to suit your company's needs.
The automatic matching process uses the available information about imported transactions for searching matching documents and for calculating the relevance rate, which is a measure of how closely a document matches the selected transaction.
As you are working in the Process bank transactions (CA306000) window, you can specify the matching parameters that apply to the current user session in the Settings for matching dialog box (which opens when you select a cash account and click Settings for matching in the window toolbar).
You can modify matching settings in this dialog box and experiment with the settings to find the best values for relevance calculation in your organisation.
After that, you can specify the most appropriate values as the default ones on the Bank statement settings tab of the Cash management preferences (CA101000) window.
During the matching process, the system uses the first two sections of the matching settings,Disbursement date matching and Receipt date matching, to filter documents in the system by date.
The documents whose dates fall into the specified range are considered candidates for calculating the match relevance rate.
Then the system uses the next two sections of the matching settings, Weights for relevance calculation and Date range for relevance calculation, to calculate the match relevance date for the documents filtered by date.
For each matching document, the system displays the calculated rate in the Match relevance column on the Match to payments or Match to invoices tab of the right pane.
The match relevance rate is between 0 and 1.
The best match is the cash transaction with the highest rate or a significantly high rate (see the details in Best match selection rules below).
The cash transactions with a rate of 0 do not appear on the tabs.
In this topic, you will read about how the system calculates the match relevance rate and how it searches for the best match.
Match relevance rate calculation
The match relevance rate is calculated based on the match settings.
In the match settings, you specify the relative weight of the three factors from which the likelihood is calculated:
- Reference number
- Date of the document or transaction
- Payee information
In the match settings, for each of these factors, you specify the percent or weight of the factor in the match relevance rate.
Before the system calculates the rate, it applies to the document date factor an additional weighting, which you can specify as follows:
- In the Payment clearing average delay field, you specify the number of days after the document date in the system that the payment is usually delayed before it appears in the bank statement.
- In the Estimated deviation (days) field, you specify the deviation, which defines the date range that includes almost all (99.73%) possible dates during which the transaction in the system that could match the transaction in the bank statement.
A date that is outside of the range is very unlikely to be the date of the bank transaction.
The match relevance calculation formula is shown below:
For example, suppose that you use the following match settings:
- Ref. no. weight: 70.00
- Doc. date weight: 20.00
- Doc. payee weight: 10.00
- Payment clearing average delay: 0.00
- Estimated deviation (days): 5.00
You have imported the bank transaction with the reference number 1231, the receipt amount 9000€, and the date 1/1/2020.
After you have run the automatic matching process, a possible match has been found by the reference number and transaction date.
The reference number of the possible match in the system exactly equals the reference number of the bank transaction.
The date of the possible match in the system is 12/31/2019, and this date falls within the date range of 99.73% probable dates of the bank transaction.
The probable date range is 12/31/2019 through 1/5/2020, where the end date of the range is calculated as 12/31/2019 + 5 days = 1/5/2020.
The bank transaction date is one day past the transaction date in the system.
Because the date is shifted, the additional weighting function for the date is equal to 0.98, which is the normalised value of the Gaussian distribution with the mean of 0 (Payment clearing average delay) and the standard deviation of 5 (Estimated deviation).
As the result, the Match relevance rate for the found transaction is 0.7 * 1 + 0.2 * 0.98 * 1 + 0.1 * 0 = 0.896, where the factors are the reference number, document date, and payee, respectively.
The calculation of the additional date weighting factor is illustrated on the diagram below:
Best match selection rules
Based on the calculated match relevance, the system selects the best match according to the following rules:
- The best match is the transaction with the highest Match relevance rate that is greater than 0.75.
- If there are no transactions with a Match relevance rate that is greater than 0.75, the best match is the transaction that is more relevant than the other found transactions for 0.2 or more of the Match relevance rate.
For example: If two transactions were found, one with relevance of 0.25 and another with relevance of 0.5, the transaction with relevance 0.5 will be matched. - If only one transaction is found, it is the best match if its Match relevance rate is 0.2 or greater.
- Otherwise, there is no best match for the transaction in the bank statement.
To do this, in the Process bank transactions (CA306000) window, click Unmatch all in the table toolbar, and then click Match automatically to re-run the process of automatic matching.
