An expense report is a document (digital or paper) drawn up by employees of a company to communicate the expenses incurred in the workplace. Many companies, especially large ones, prepare expense reports on a regular basis. The following article describes the expense report preparation process and its main steps. It focuses on the problems of manual processing and the benefits of automated expense report management.

Expense report process

It is necessary to prepare an expense report every time an employee anticipates costs that the company will reimburse. Typically, this process involves: settlement phases for real and lump sum expenses, supporting documents, and more. Correctness and timing of reimbursement are extremely important, so a secure, reliable, and digitalized process is essential to avoid delays and errors.

Expense reports serve as a tool to determine, verify, and account for these advanced costs, so that reimbursement can be pursued. It is an important process for companies and control bodies such as the Revenue Agency as it contributes to the formation of business income.

Examples of expenses typically covered by reimbursement include:

• Food and lodging expenses: catering and accommodation on business trips. You should check the company’s policies before paying lump sums in order to avoid non-refundable costs.
• Travel expenses: for travel by plane, train, ship and so on. Fuel and mileage are refunded according to the ACI tables with your car.
• Other expenses: those not related to the previous ones, such as parking, representation, photocopies of sentences filed by a lawyer, etc.

Expense report process – steps

In general, expense reports consist of the following phases:

• Preparation of the expense report: this phase involves creating the document for reimbursement requests to the firm (client) by the worker. Employees are solely responsible for drafting. The compilation of the document typically requires a series of precise details, which can be summarized as follows:
  • Personal and company data;
  • Expenses list. Travel-related expenses, including amounts and progressive numbers.
  • Date, time and place.
  • Tax proofs such as receipts, invoices, and travel documents. This document proves the actual economic outlay for travel.
  • Comments and notes useful to justify extraordinary costs.
• Supervision: during the supervision phase, the information entered for reimbursement is verified for accuracy. This phase can be performed by internal company resources or by outsourcing. The auditor verifies if the information entered is correct and if there is a receipt with tax validity. In addition, He confirms if the expenses to be reimbursed are in compliance with the company’s regulations and policies.
• Administration: The last step is to transfer the validated data from the expense report to the company’s accounting system. This ensures that the accounting entry is completed and the employee is reimbursed.

To manage the expense report process correctly, it is essential to easily process the multitude of different purchase receipts that can be involved in this process.

Receipts processing

This section examines the types of documents related to the expense report process. Additionally, we will explore the key complications that have limited the automation of the expense report process in different contexts.

Documents involved

In the expense report management process, the documents involved as supporting documents are many and often quite different from each other. These are generally semi-structured documents that often contain some of the information required to prepare the expense report itself. In order to optimize the process correctly, it is essential to be able to automatically classify and obtain useful information. The most common types of documentation used are receipts, invoices, tickets, motorway tolls, etc…

Main difficulties

Each of the documents mentioned above is a semi-structured document. This type of document, depending on the supplier, can significantly change the position of the information of interest and the format. For a large company, another problem related to the centralized management of the expense report is inherent to the language. In fact, purchase receipts are usually linked to trips to different countries with a relative change in language, currency and format. In some cases, the receipts are even handwritten, which further complicates the processing (both manual and automatic). Finally, the receipts are often delivered in different formats such as scans (often degraded), PDFs and photo from mobile phones (usually difficult to read).

Main steps

Identification of the class to which they belong

One of the relevant information to identify when completing the expense report is precisely the category related to the receipt that is attached to the expense report. Generally, the categories considered eligible are in the following list:

• Entertainment expenses (hairdressing, clothing, etc.);
• Transport (plane, train, taxi, vehicle rental);
• Overnight accomodation;
• Breakfast / lunch / business dinner;
• Travel tolls;
• Parking;
• Expenses relating to congresses / conferences and events.

This information is often not explicitly present in the document, but may be deduced by the reading and identification of the class to which it belongs (for example for an overnight stay we will check that the receipt is actually a hotel).

Relevant information to extract

Many other relevant information may be deducted directly from the tax receipt as it is often contained within the tax receipt. Information that is generally extrapolated for accounting, tax and audit purposes is: information about the vendor, the full name of the vendor, VAT / VAT no; information about the country and address; the currency for the payment, the total, the taxes paid, the payment method, any purchase lines, and finally the date, time and place.

How to process purchase receipts

Manual approach

Extracting data from purchase receipts can be costly, lengthy and error-prone when done manually. The processing steps require well-trained people who can identify the pertinent information in the invoice and systematically extract it from sometimes even complex layouts. Challenges and issues in manual treatment include:

1. Cost penalties
   While manual document data extraction may work for small businesses with limited processing volume, it becomes costly as the business expands. In many cases, manual data extraction results in hidden costs in addition to the costs of hiring multiple employees to perform the data extraction. Adding multiple employees to the workflow often leads to coordination issues, which in turn can lead to errors, especially in data identification and capture. Validating the data itself is a crucial step that adds to the cost component of treatment. In fact, extracting information without checking steps can lead to error rates of up to 4%. The 1-10-100 data entry rule is well known in data entry back offices: verifying data accuracy at the entry point costs about $ 1, cleaning up errors by double-checking the entire batch of data costs $ 10, and missed errors cost the company $ 100 or more.
2. Time delays
   Manual data extraction can be time-consuming, especially global supply chains, as multiple audits and approvals are required and teams in different countries often need to be coordinated. The integration of the different processing and verification steps can also become complicated because of the different people involved in the hierarchy and the required approval levels.
3. Human fallibility
   The data entry process is repetitive and time-consuming and may be demoralizing. In addition, there is no standard format for purchase receipts. While much of the information listed in the preceding sections is present in all documents, each provider uses a different format with significant spatial variability in the data. Manual data extraction can therefore prove tiresome given these variations. Language variations between the place of issue and the place of delivery may also pose an important challenge in understanding the data.

Traditional OCR Solutions

Processing purchase receipts with traditional OCR techniques and matching patterns/regex is a highly discouraged and costly approach. This is because you need ad hoc sets of rules and templates for each type of document. There are a lot of formats and sellers are potentially in an undefined number a priori. There are often many languages to consider for a solution that needs to function in global processes. This makes the number of rules or templates required very numerous if not infinite and constantly changing depending on the new formats and countries you wish to consider. All this implies a high cost of setting up and maintaining the solution and also often lack of performance. Furthermore, the solution must be maintained and configured by trained resources and with technical training.

Intelligent Document Processing (IDP)

A modern approach based on Deep Learning techniques is the best choice to solve problems of this kind. The ability to use the best Computer Vision techniques for image analysis and NLP for natural language understanding allows you to scale to a large amount of different formats and languages without having to adapt the solution every time (writing new rules or configuration of new templates). There needs to be just enough data to form the system.An approach of this type can also greatly benefit from the human validation step which, in addition to correcting the errors made by the system, can enable the continuous learning of the algorithm, allowing it to improve over time and calibrate itself on the specific process.

Compared to traditional solutions, even adding a new field to be extracted, adding a document category to be classified or wanting to add a new language among those supported does not involve writing code. The collection of new documents will be sufficient and the subsequent retraining of the system can also be followed by a resource without technical skills but who has knowledge of the process. Finally, the most effective solutions of IDP also allow to obtain an unprecedented accuracy of results that far exceeds both manual processing and traditional solutions.

myBiros and benefits

myBiros is an IDP solution that allows the automatic documents processing. Among the main features, there are the extraction of information and the automatic classification of documents. myBiros provides a prebuilt set of ready-to-use APIs with pre-trained models for the most common use cases and the ability to retrain the entire pipeline (both the OCR engine and the document interpretation system) for custom cases.

By exploiting advanced deep learning techniques that analyze multimodal features, it is possible to process all the document types mentioned above with the same solution. Thanks to the use of pre-trained models and data augmentation techniques, it is possible to train the system with a limited number of data.

This allows you to train AI models even to those who do not have extensive amounts of documents. Through the scoring mechanism, the system allows to reduce false positives by enabling the possibility of reviewing low-confidence data, minimizing errors. Interaction with a human user allows you to correct system errors by continuing to train him so as not to repeat the mistakes made in the past (Human-in-the-loop and continuous learning). Finally, the high scalability of the cloud-based architecture allows you to process highly variable amounts of documents without having to previously allocate expensive resources.

The features mentioned so far allow myBiros to perform optimally on the documents involved in the processing of the expense report: receipts, invoices, parking tickets, motorway tolls. If you are curious to know how myBiros works in order to simplify the processing of any purchase receipt, contact us and try our demo. We are ready to help you!