Matching implementation

To add data and match transactions, follow the implementation steps for Matching. You can also use these steps to add new definitions to the matching process.

Follow these steps in the given order to add data to Transaction Matching:

1. Create transaction files

Transaction matching starts with generating transaction files. These files include columns such as count, type, date, transaction ID, vendor, customer, and amount. There are no limits on the number and type of fields that can be included in a transaction file.  Be certain that any added transaction file contains all the fields necessary to create matches, including date(s), amount(s), and any other identifying data.

2. Create staging and transaction tables

After creating transaction files, you need to create staging tables and transaction tables:

• Staging tables are temporary storage tables where raw data is imported for initial processing.
• Transaction tables are the destination for processed data, where matching rules are applied.

Ensure that you create both staging and transaction tables for each data source.

3. Create datasets 

After preparing your tables, create datasets that contain the data needed for the matching process. These datasets consist of data extracted from ‌staging tables and specifically customized for matching.

4. Create match groups 

‌Match groups represent a specific comparison scenario, such as Bank to General Ledger matching. After creating datasets, create a new match group and set it to active. This match group defines the datasets you want to compare. You can include multiple match groups, if you have different comparison needs.

5. Associate datasets with match groups

After creating the match groups, associate the datasets you created earlier with the suitable match group. This step links the datasets for comparison within the match group.

6. Apply match group dataset filters 

Filters are optional but useful for narrowing down transactions that are to be included in the matching process. You can apply filters at the dataset level within each match group. For example, you may want to filter bank transactions by reference number or account number.

7. Create match rules

Match rules define how you compare transactions from different datasets. Create rules with specific logic for matching transactions.

• One-to-one rules compare individual transactions.
• Many-to-many rules compare groups of transactions, such as batches of payments.

Each rule must have a name, rule type, and confidence level. Set the rule's sequence to ensure the proper order of rule execution.

8. Add match rule details 

Define the specific matching criteria for each rule:

• Amount comparisons ensure amounts are equal or within a tolerance.
• Date comparisons specify the date matching criteria, such as within a specific date range.
• Grouping fields group transactions by criteria, such as transaction ID or date.

You can add multiple criteria and filters to fine-tune the matching process.

9. File import task for data ingestion 

With the rules and groups set, create a file import task to load transaction data into the system:

• Import raw data into the staging tables first.
• Then, use a calculation task to move the data from staging tables into transaction tables.

Ensure you specify the appropriate staging and transaction tables in the process.

10. Run matching rules

After you move the data into transaction tables, you can execute the matching rules. Add a calculation task to run the matching stored procedure, for example, [matching].[spAppg_Match_Bank_to_GL_Matching]. This process applies ‌match rules to compare and reconcile transactions.

11. Manual or automated execution 

Finally, decide whether to:

• Run the workflow manually.
• Schedule the process to run automatically regularly.
• Run the match rules through the Match Rules form.

This allows you to automate or trigger the entire matching process when needed.

By completing these steps, you can effectively establish data input and automate the transaction matching process.