For small companies and large enterprises alike, mail continues to be an integral part of the document workflow. In a typical mail center today, 60% to 70% of the labor is spent sorting incoming and interoffice mail. Incoming mail is defined as those mailpieces that are received by any company, and in addition to the postal address, contain company-specific addressee information, such as name, title, department, subdivision and other specific details. Unlike the postal address, this information is not regulated by strict rules, and it may appear in a great variety of formats on the mailpiece. Enterprises are confronted with the problem of sorting, distributing and processing incoming mail on a daily basis. These processes are laborious, expensive and time-consuming, and deal not just with letters and postcards, but with an increasing number of other formats including flats, accountable mail such as FedEx, UPS or USPS Express Mail, parcels and faxes.
It is comparatively easy to manage the flow of incoming mail in small companies: often it is enough to read only the first and last name on an envelope. However, the problem tends to grow with the size of a company. In large corporations, the structure of addresses may include multiple fields that conform to complex rules and various priorities that have to be considered to unambiguously locate an addressee.
The speed and accuracy in which mail is sorted and routed impacts overall business efficiency. It is difficult to find an industry or a company that can afford to ignore the importance of incoming mail processing. Efficient processing is particularly valuable in remittance/payment environments that deal with time-sensitive and/or valuable documents. Delayed deposits cost banks and investment companies customer goodwill and interest. As the speed of business accelerates and the use of overnight deliveries steadily increases, the problem of misplaced parcels, mail, faxes and files is more critical than ever. And in companies that receive product samples, lost parcels can mean lost opportunities, time and dollars.
Research shows that, depending on company size and correspondence volume, at least 2% to 2.5% of all mail is either misplaced or subjected to an in-house delivery delay. This increases to 3% for any company that has more than one site. These percentages may sound small; however, in terms of lost potential business, they highlight a serious problem. For a company receiving up to 100 pieces of mail per day, about 600 are either delayed or misplaced annually. If just 10% of these mailpieces are customer-related, the company will miss 60 opportunities for potential revenue, contracts, payments, customer inquiries and/or satisfaction. In the case of a large company, the negative impact of misplaced or incorrectly sorted mail is even more critical. For example, for companies like E.ON Westfalen
The paradox is that even though organizations understand the importance of timely and accurate delivery of incoming mail, the processes of sorting and tracking are often still done manually. This has been an area considered untouchable by automation, especially for small to medium-sized mail centers. Large mail centers have had the option of acquiring universal sorters, but the investments involved have been too great for smaller centers. Yet, very few large mail centers have found it possible to automate incoming mail, even in the face of rising labor costs. This is especially intriguing since advanced, automated postal systems have been effectively applied by large, national posts to process outgoing mail for decades. The reason for this discrepancy lies in the specifics of the address block on the incoming mailpieces. While the structure of a postal address block is strictly defined, incoming mail structure is diverse and may diverge not only for different companies, but even within the same one. Therefore, it is difficult to find universal software that can reliably and efficiently process the entire spectrum of semi-structured addresses.
The concept of locating and reading mandatory address fields within the address block, which proved to be successful for postal addresses, failed or showed to be inefficient in the case of incoming mail. This application required software based on a different concept to reach high recognition results. Nowadays, there are products exploiting a new concept in reading incoming mail addresses. They are specifically designed to automatically recognize the wide variety of incoming mail and are intended to support high-volume mail processing facilities. Such products can sort any type of inbound company mail up to the depth required by the particular company. They are flexible and can easily adapt to any address structure and field type defining the destination. Advanced functionality allows customers to successfully implement the products for the most challenging tasks of incoming mail sorting, which previously could not be solved.
New technologies are able to locate the correct destination address block on the image of an envelope, flat-size mailpiece or parcel. The software automatically parses and recognizes the fields within this address block. It is designed to deal with different address block structures and automatically reads the variety of fields and formats that may be encountered on a mailpiece. The fields may be located in different lines within the address block, may have different prefixes or have no prefix at all and may be present within an address block or skipped.
The differentiator of this new approach is its ability to look beyond each particular field within the address block. The new products allow fields to be combined in groups, finds groups of fields that are stably present in the address block, sets complex connections between fields within the group, defines priorities that describe the importance of the field in determining a potential recipient of the mail and, finally, builds a chain that leads to the proper mail recipient.
Each company has a specific address structure. In small companies, it may be limited to just the name of a recipient, followed by the company address. If a company structure is more complex, the internal address may include such fields as mail stop, department name and post office box. But real-life mailpieces often go beyond the typical address block structure. There are addresses that may include unusual fields, specific to a particular company (for example, drop code or division code). Big companies may have a hierarchical mail stop or department name structure (multiple nesting mail stops or departments) or a few mail stops, or department structures existing in parallel.
The universal language mechanisms used by incoming mail processing software allow for flexibility in describing and working with a variety of formats and cases. Ready-to-use blocks or predefined fields can provide a user-friendly definition of particular address structures in these systems. Predefined fields are often encountered on mailpieces and imply standard information about field structure, field prefix or suffix and expected field values. The software contains knowledge about such fields, enabling it to locate them on different mailpieces. It also offers mechanisms that allow for flexibility in modifying standard fields, so they meet the requirements for a particular address format. In those situations when an address includes fields that cannot be defined using the predefined field types, the customer can define fields with unique characteristics.
The technology does not have country-specific address structure constraints and can be readily applied to Latin-based addresses. It opens new opportunities to implement this technology for a broader range of applications going beyond incoming mail processing.
An important part of the successful reading of handwritten and machine-printed addresses belongs to contextual information or utilizing a database describing company-specific address structure, fields specifying a destination and a list of potential addressees. The database should contain the most comprehensive data, specifying valid fields and their values for a particular company. Fields on a mailpiece are recognized and compared against records from the database to find a single record that matches the mailpiece. There are also mechanisms that help resolve the ambiguities inherent to this task, which is difficult to formalize. For example, a mechanism that sets priorities helps resolve situations when a database has multiple records with several fields matching the same values. Alias mechanisms help to resolve the ambiguity caused by the fact that values of the same fields within the address may often be written differently. For example, the first name William can be also written as Bill, Will, Willy, etc. The department name "Marketing" may be spelled as "Mktg" or "Market." Despite the various ways they may be written, all these words have the same meaning and, as such, describe the same destination specified in a database in a standard unique form.
All of these features and other powerful mechanisms allow for the successful resolution of incoming addresses. The software can be easily integrated with various types of sorting equipment and adapted to a company's business needs and requirements. Automated incoming mail processing brings significant benefits to organizations that implement it. The technology makes sure that incoming mail gets to the right place faster, reducing document processing time and raising the efficiency of business processes, while improving quality of customer service.
Dr. Tatiana Vazioulina works in product management and marketing at Parascript, LLC. Email Tatyana.firstname.lastname@example.org, call 303-381-3106 or visit www.parascript.com.