Digitization And Paperless Processing: What You Need To Know

Digitization via Document Imaging and Capture: The Technologies and the StagesThe technologies involved in digitization include:Scanning hardware – input devices to convert the paperdocuments to images. Products range from low-volume desktopscanners to MFPs with scanning capabilities, to high-volumescanners.Capture software – the set of tools used to take an image from ahardware scanner and turn it into a manageable information asset.Features include quality assurance (QA) checks and rescan support,barcode reading, optical character recognition (OCR) processing,and automatic and interactive indexing on scanned documents.Capture software products typically have their own imagerepositories, but also can be integrated with the repository of anECM system (see below).Input sCaptureEnterprise content management software – provides thecapabilities to manage the images for more complex applications,supplementing the functional capabilities of the image repositoriesprovided by capture solutions.Workflow and business process management (BPM) software –technologies that allow the routing of the imaged documents as partof a business process.Page 5ECMAdministrationQA entRepositoryECMBrowser andThick-client UserInterfaceTaxonomy& Metadata

Digitization of paper documents via imaging and capture involves afairly standard flow that uses several different technologies. Themajor stages are as follows:Document preparation – the steps required to prepare documentsfor scanning, such as unfolding, removal of staples and paper clips,sorting by document size, insertion of separator sheets.Scanning and document capture – the use of devices to digitizeinbound paper documents and convert them into an electronicimage. Types of devices that can create image files include scannersand MFPs that combine scan, fax, and copier functionality.Image cleanup – processes performed on the images, such asde-skewing, de-speckling, etc. Such capabilities are sometimesavailable within the capture devices, and they also available fromthird-party capture software vendors.Recognition and indexing – processes that recognize an image andinformation it contains, associating the data or metadata with theimage. Most capture software provides or supports automatedrecognition and indexing techniques such as barcode recognition andOCR. In addition, most capture software products provide theability for operators to perform manual indexing of an image. Somedistributed capture operations include an additional quality assurancestep, in which images are routed to another workstation for finalreview of the image quality and data accuracy.Export to a content management system or other repository –the value of distributed capture is greatly enhanced if images andindex data can be imported into a content management system orother repository. Options include digital sending technology that canelectronically deliver images to multiple destinations (file systems,content repositories, email addresses, etc.), or capture software thatprovides export or “release” modules designed to migrate theinformation directly into a destination content management system.Typically, the activities that incur the highest costs are those whichinvolve the most labor: document preparation, rescanning followingquality assurance, exception handling, and post-recognition (exceptrelease). The actual scanning and the release activities tend to belowest in cost. Depending on the types of documents being scanned,the recognition/indexing stage can be either high or low in cost.A key consideration for any digitization application is the concept ofthroughput. Throughput involves factors such as the speed andimage quality required for an application, as well as the usability ofthe final images and the reliability of the hardware and software.Effective throughput is critical for scenarios involving timedependent documents (e.g. new account origination documents), aswell as for high-volume scenarios (e.g. insurance claims).Further considerations from a throughput standpoint include thefunctional extensibility and scalability of the operation. Will yourorganization likely start small, with applications that are lesscomplex, and in the future expand its digitization capabilities todocuments that present more complex requirements? If so, yourhardware and software should have the functional extensibility toaccommodate the more complex applications. Likewise, factor infuture needs for scalability. You may start off with low-volumeapplications, but should plan ahead for the possibility of highervolumes – either from expansion to higher-volume applications orfrom changes in the business – as the result business growth from amerger or acquisition, for instance.Page 6

Which Capture Model: Distributed vs. Centralized?The question for many organizations is which model makes sense for their business needs: a distributed capture operation, in whichmultiple parties scan documents from multiple locations; or a centralized model, in which documents are scanned at a single, centrallocation. Then there is the middle ground: departmental, or moderately distributed, scanning.Centralized ModelDepartmental ModelIn the centralized model, business operational mail items, meaningitems addressed to departments or generic roles (e.g. “claimsadministrator”) are all opened, prepped, and scanned at a singlelocation (generally the central mailroom or mail center) and then aredelivered electronically through either the capture system or theECM solution. Originals are either filed or destroyed, depending oncompany policy for such items.In this “moderately decentralized” model, individual business unitsact as centralized scanning centers for their own operations. Eachdepartment typically has one or two staff members who areresponsible for receiving inbound mail, and who perform scanning,indexing, and distribution for all members of the department. Thissolution occupies a middle ground between centralized anddistributed capture; as such, it presents higher hardware costsbecause it requires a larger number of scanners across the variousbusiness units. The advantage is greater quality control, asdepartments in general tend to accurately index their own documentsif done at a group level.This shared-services model is typically deployed in large-scale mailcenters handling millions of pieces of mail per year. Fororganizations that receive large volumes of inbound documents,such as insurance companies, it may make sense to outsource themail center scanning operation. In this approach, all inbound mailfor all departments and business lines is delivered to a centrallocation. At the mail center, staff (contractors) open the mail, sort,remove staples, then feed the items into high-volume scanners, andperform quality checks. Generally, actual employees of theorganization take over for the indexing of images, which requiresmore business knowledge.Page 7Distributed ModelIn the distributed model, scanners are deployed to the end users,scanning mail at the point of use rather than centrally. In this model,imaging and indexing is fully distributed among individuals, whoscan items according to their role. This is accomplished withextensive use of personal scanners or dedicated MFPs. This methodoffers the most individual control of documents, but also loses someamount of organization at a high level, because of the number ofdifferent people assigning index keys to documents. In addition, thissolution requires a higher per-employee technology cost because ofthe scanners required.

The following table provides a summary comparison of these models.OptionsAdvantagesLoss of business-unit controlBusiness-unit issues (political, control issues)More enterprise consistencyLarger space requirement in centralizedareaPotential legal and compliance issuesLeverages technical expertise and resourcepool of centralized functionMore business-unit control (workflowcontained in one environment)Leverages specialized content expertisewithin each unitDepartmentalCustomized processes for each businessunit according to their rulesSpace needs are more scattered; smallerscanners distributed into multiple areasMost individual control; desktop captureallows for most customization at theindividual levelEnables remote scanning at distributedwork sitesDistributedBarriersMore enterprise controlFocused training with a smaller teamCentralizedLimitationsMay not meet quality requirementsbecause of lack of specialized contentexpertisePotential inconsistent application ofprocesses, indexing, etc., fromdepartment to department; less enterprisecontrol and potential QA issuesMay require more equipment (lowerefficiencies of usage)Duplication of (limited) technicalexpertise among business unitsMay lack sufficient space for a scanning areaMore staff required – imaging specialists,indexingRequires staff training at departmental levelEquipment – using machines for dedicatedscanning blocks their usage for copying,printing, and faxingRisk of insufficient technical expertise todistribute across all areasConsistency, training issues; may havemore challenges with consistent indexingSame barriers as Departmental, but to a largerdegreePotential issues in attaining acceptablelevels of participation and qualityImpacts work by assigning new tasks on top ofcurrent dutiesEmployees will lack advanced technicalexpertise; to be successful, this modellikely requires some departmental orcentralized resourcesTypically fails to adequately address the dual(but conflicting) requirements ofparticipation and quality (usually of indexing)Page 8

The following tables list the segments that represent the various permutations of these factors, with examples of common applications for each.SegmentCharacteristicsBusiness drivers: reducing paper and manual processing of mail, shipping, or dDistributed contributors, each capturing low volumesMinimal indexing or other processing and minimal incorporation into downstreambusiness processesBusiness driver: paper reduction within workgroup or departmental deploymentMinimal indexing or incorporation into downstream business processesRequires incorporation of low- and mid-volume capture devices and demonstratedreliable software capacity to handle (albeit modest) volumesBusiness driver: quickly incorporate critical information into complex high-valuebusiness applicationsCapture to file systems, email folders, or arepository for simple search and retrievalEmployee expense receipt and report processingCorrespondence management, meeting agendamanagementCapture to file systems, email folders, or arepository for simple search and retrieval forcompliance, customer serviceInvoice processing for Accounts Payable (AP)Requires significant processing and incorporation into downstream business processesTransportation and distribution (shippingdocuments)Typically requires manual indexing by experts, visual QA, filtering to designatedworkflows, and notification or confirmation upon delivery of images to downstreamsystemsCustomer service, customer enrollment, claimsprocessing, mortgage loan processing at branchofficesBusiness driver: quickly incorporate critical information into complex high-valuebusiness applicationsInvoice processing and related documentprocessing for APRequires significant processingRequires low- and mid-volume capture devices and demonstrated reliable softwarecapacity to handle (relatively modest) volumesMay involve release into an ECM system repositorySo which model is right for your application’s needs? The key factors to consider are as follows:Scale of deployment (in terms of the number of contributors)Application complexity (simple vs. complex)Location of scanning capabilities (distributed vs. centralized)Page 9Examples

dCharacteristicsBusiness driver: reducing paper and manual processing of mail, shipping, or faxRequires minimal indexing or other processing and minimal incorporation intodownstream business processesKey characteristic is volume: a large number of capture sites, with a high volume ofdocuments to be (minimally) processed and routedBusiness driver: production departmental or enterprise capture, focused on reducingpaper and manual processing of mail, shipping, or faxRequires minimal indexing or other processing and minimal incorporation intodownstream business processesSimilar to Small/ Simple, but on larger (typicallylarge enterprise) scale: capture to file systems, email folders, or a repository for simple search andretrieval for compliance and customer serviceSimilar to Small/ Simple, but on larger (typicallylarge enterprise) scale: capture to file systems, email folders, or a repository for simple search andretrieval for compliance and customer serviceKey characteristic is volume: a large number of capture sites, with a high volume ofdocuments to be (minimally) processed and routed; requires reliable performance inhigh-volume hardware and softwareBusiness driver: speed of incorporating critical information contained in documentsinto complex, high-value business s significant indexing or other processing and incorporation into downstreambusiness processes; typically requires manual indexing by experts, visual QA, filtering todesignated workflows, and notification or confirmation upon delivery of images todownstream systemsMany applications superficially similar to the smallcomplex applications, including customer service,customer enrollment, claims processing, andmortgage loan processing, but differ significantlyin both scale and complexity of applicationKey characteristics are high total volume of documents of several different types andcomplexity of capture, and differential processing, and routingBusiness driver: speed of incorporating critical information contained in documentsinto complex, high-value business applicationsLargeComplexCentralizedRequires significant indexing or other processing and incorporation into downstreambusiness processes; typically requires manual indexing by experts, visual QA, filtering todesignated workflows, and notification or confirmation upon delivery of images todownstream systemsMost mailroom processing, where documents tobe processed originate from external senders:applications, enrollments, claims, mortgage loanprocessingKey characteristics are high total volume of documents of several different types andcomplexity of capture, and differential processing, and routingCritical factors include the value of the information for the business process and theoverall value of the business process itselfUsually involves release into ECM repository, often with capture subsystemPage 10

E-forms for Upfront Data CaptureAs the cost of manually processing forms continues toincrease, more organizations are considering making themove from paper forms to e-forms. Then there are theother business benefits: e-forms get information into a businessprocess faster, more efficiently, and with greater accuracy thanmanual data entry from paper forms. At a time when customers –both internal and external – are demanding instant access toinformation, the business driver for e-forms couldn’t be clearer.E-forms applications range from the relatively simple to the verycomplex. In general, however, they can be divided into three levelsof complexity:1. Applications that electronically deliver forms, usually via theweb, for printing and completion. Within this category are twofurther levels: applications that provide the ability to print, fillby hand, and submit the form by mail or fax; and applicationsthat provide the ability to fill the form electronically, validate,print, and then submit the form by mail or fax.2. Applications that display forms for completion online, in orderto replace online operators and paper forms; ideally, the formsare displayed via the web, providing the ability to pre-fill,validate, save the file locally, digitally sign, and submit the formelectronically.3. Applications that manage and route forms and data, which addintelligent, “process-able” data to static images. Theseapplications primarily address the processing costs of forms,using a dynamic pre-fill form or wizard interface to fill the form,and allowing the user to validate, save the file locally, signdigitally, and submit the form electronically.Page 11Many organizations have already implemented the first level ofe-forms, by posting the forms as PDFs on their web sites orintranets, or on a shared network drive. For these organizations, thechallenge is to find ways of achieving greater efficiency in theprocessing of the completed form. One approach is to establish acentralized location for receiving the mailed or faxed forms, and usescanning and OCR technology to capture the data from thecompleted forms. An alternative is to decentralize the capture of thedata from the forms by scanning at the point of receipt – forexample, a local office or field agent, or by capturing the data at thepoint of origin, either through fax or scanner or an MFP. Thesedecentralized approaches provide the added advantage of eliminatingshipping costs and getting the completed forms into the workflowthat much faster.Organizations that seek to implement more complex e-formsapplications (Levels 2 or 3, above) will require products with deeperelectronic forms capabilities. Several of the major platform vendorsnow provide e-forms capabilities; in addition, the leading vendors inthe ECM market space offer e-forms capabilities. As an organizationconsiders which e-forms solution to implement, Doculabs believesthat the primary considerations should be 1) the functional/technicaladequacy of the solution for your particular application or enterprise,2) the business requirements you expect your e-forms solution tomeet, and 3) vendor viability.

Critical Success Factors for Digitization InitiativesDoculabs has advised many clients concerning best practices for digitization initiatives, helping them define their requirements for bothimaging/capture solutions and e-forms solutions. The following are some best practices to be aware of, before and during the rolloutprocess, to mitigate risks and manage the changes that result from applying digitization to business processes.Take a holistic approach to digitization. Many organizationslook at their business processes too atomically, seeking toconvert paper to digital format via imaging rather than looking atopportunities to eliminate paper and capture data through theuse of e-forms. Too many digitization strategies don’t includesteps toward elimination of paper by electronic capture throughall potential methods. This is most critical for organizationsseeking to implement the technologies for struct

benefits come from distributed capture: digitizing the paper documents at the location where they originate and getting the information from the documents into the workflow for processing as quickly as possible. Until recently, cost was a significant obstacle to any digitization initiativ