Just What Are Those Low-Hanging Fruit?
Identifying Quick Wins In OCR Projects

In workflow/OCR consulting, early wins are often a hallmark of a successful automation project. They beget momentum, followed by buy-in, followed by wider support and adoption. That’s why–in the initial stages–even small victories can be accurate signs of a successful, high-momentum project to come.

FlexiCapture OCR solutions are all about turning the complex into the simple. They give the greatest long-term ROI on processes that were previously bound by steps and checks and human interventions. But when a project sponsor has gone out on a limb to change the status quo, putting their budget and reputation on the line, it’s wise to climb the hill before conquering the mountain.

This leaves us and our clients with the (sometimes literally) million-dollar question: “How can we identify the ‘low-hanging fruit’ that will make a quick win possible?”

Often, we take the Socratic approach and answer that question with some of our own. As a starting point, these questions have proven consistently helpful over years of OCR and workflow automation consulting.

What’s disproportionately painful?

We want to home in on the 20% of the processes that cause 80% of the headaches, as the rule suggests. Even within that subset, there’s usually one workflow that causes the most groaning and eye-rolling. That is the place to start!

If clients can pinpoint a single process that yields more than its fair share of costs or confusion, then they’ve got a prime candidate for immediate automation. After all, a highly visible pain point means that the solution will be highly visible, as well.

What information carries the biggest legal risk or compliance burden?

Just as current pains are a focus of project planning, we’d be remiss not to look at the likelihood of future pains. This is often very industry-specific, as with HIPAA compliance during/after EMR adoption, confidentiality of credit history in mortgage application bundles, etc.

Whatever the specific regulations, there are typically large and very expensive legal consequences at stake. Whenever FlexiCapture can help address data accuracy and security, a compliance-oriented automation project may be one of the quickest ways to provide value.

If you’ve tried to deploy OCR before, which business processes were involved? What got in the way?

Another opportune starting point is any process that an organization tried previously to automate, but was stymied by inadequate technology or rough implementation. On a technical level, ABBYY customers routinely find that FlexiCapture simply outdoes other applications in sheer accuracy–which is often tested head-to-head on real-life forms.

Combining standard-setting performance with a proven project planning methodology, it’s often possible not merely to resurrect older initiatives, but to see them succeed on an even greater scale than originally planned.


We want to set the stage for a victory from day one, and these questions are just the tip of the iceberg. There are thousands more that help us to help you lay the groundwork for a smooth, effective, and high-ROI automation project.

Collectively, our consultants have spent several decades in OCR and workflow automation consulting, all the while building and honing a time-tested workflow toolkit. So, if your team could use some expert guidance, then we invite you to reach out today.

OCR Solutions to Data Quality Problems

Think about all the potential causes of bad enterprise data. Application problems, database processing issues, flat-out misunderstandings…the list goes on.

But it turns out that data entry itself is the most common culprit by a wide margin. A 2016 study from 451 Research revealed that staff data entry is a problem for the majority of companies surveyed–a sobering 57.5% in all.


Data Quality Improvement Through OCR Data Entry Automation

Source: http://core0.staticworld.net/images/article/2016/02/dqm-2-100643610-large.idge.png

Not invoicing errors or math errors or even customer errors wreak more havoc on enterprise data quality than plain old typos. Furthermore, data entry mistakes create an unsteady foundation on which downstream processes build. This phenomenon–let’s call it “data entropy”–is as simple as it is insidious:

Errors get magnified, not resolved, as workflows move forward.

Most enterprise applications are quite adept at flagging bad arithmetic, non-existent geographies, and the like. It’s easy enough for an application to figure out that “Seattle, CA, ” isn’t a real place or that two line items of $25 don’t actually make a total of $40. But can it know that a new customer recorded as “Tom” is actually “Tim”, or that item #2468 wasn’t supposed to be item #2469?

The fact of the matter is that only algorithmically discernible errors can be handled. That’s why 100% accuracy is unlikely once manually-entered data have flowed that far downstream, no matter how robust the business logic.

These mistakes are contextually plausible, which renders us mere humans especially prone to them. And when we are tired, stressed, or preoccupied with other matters, the likelihood only grows. Rather than implementing more stages of fallible human review, how much better to prevent human error at the source, before misinformation propagates through the organization!

Mitigating Risk By Mitigating Errors

The human variable is the riskiest one in data entry. With OCR software like ABBYY FlexiCapture, that variable is all but removed from the equation. The result is consistent and pre-validated data for the organization and a much lighter workload for its staff, who can be allocated to higher-value tasks.

Importantly, the OCR engine does not tire or lose its concentration after hours of repetitive, detailed work. It remains as accurate after the millionth form as it was after the first (in fact, more accurate, thanks to automatic “learning” over time).

As a very conservative rule of thumb, we can use FlexiCapture to automate 80% of a complex process in the first 8 weeks for an 80% labor cost reduction.

A typical error rate reduction is 95%-100%, depending on the writing/typing quality among other factors. But for this exercise, let’s stick with conservative assumptions and plan on a paltry 90% accuracy improvement. So, given a 90% lower error rate within the 80% that has been automated, we’d see overall errors drop by about 72% right off the bat.

For an average organization, such a reduction would push data entry to the bottom of the list of data management concerns–or even cross it right off–while doing away with related labor costs for good.

Your Next Steps To Improve Data Quality

Understanding the role of FlexiCapture OCR software in principle is often a far cry from choosing real-life processes to automate. Once settled, there are numerous technical and business requirements to work navigate. At the core of our workflow automation expertise is a proven strategy for fast, successful, and cost-effective FlexiCapture implementation. From initial discussions all the way through go-live, you can take advantage of partnership with a nationwide leader in workflow automation.

Please contact us directly for a complementary workflow analysis, FlexiCapture pricing request, or product demonstration with real-life documents.

Do you work for or alongside an Accounts Payable team? If so, then be sure to join one of our bi-monthly webinars on invoice processing automation!

Recognition Server Review: Bilingual OCR Archival Moves Gov’t Office Forward

Peru’s Oficina de Infraestructura Penitenciaria (OIP), or Office of Penitentiary Infrastructure, is chiefly responsible for prison facility improvements so as to help inmates reintegrate into society. It’s also a shining example of the power of Recognition Server to transform a complex archival process spanning multiple workflows and languages.

The OIP already wanted to become the most modern infrastructure agency in South America by 2021. When the Peruvian government rolled out nationwide digitization goals, the OIP realized that it would take a serious effort to stay ahead of the curve.

Some high-grade production scanners were already in place, but the missing piece was an enterprise OCR solution that could keep pace with their rapid, bilingual input.

Working with a local ABBYY partner, the OIP identified Recognition Server as the most accurate and versatile option for its two major workflows.

As depicted above, one workflow involves scanning from MFPs into a hot folder, from which Recognition Server directly pulls documents, processes them, and writes the results to searchable PDFs. This process requires no verification at all, and runs so transparently that users don’t even realize what OCR “magic” is occurring behind the scenes.

The second workflow is a combination of a) making day-to-day papers searchable in the OIP’s ECM system and b) catching up on a backlog of non-searchable TIFF document images. This workflow doesn’t require validation, either. But unlike the first, it does make use of Recognition Server’s Indexing Station to speed up future document searches.

As a result, the OIP now enjoys digital, searchable versions of hundreds of thousands of new documents every single year. According to the IT Director, “Recognition Server allows us to organize business processes from scratch.” Rather than being constrained by manual data entry or rigid pre-existing processes, the agency can now re-think and optimize every step of its document workflows as a modern, paperless system.

OCR conversion with this level of accuracy and speed results from skillful implementation of ABBYY’s best-in-class enterprise OCR solution. To learn more about ABBYY Recognition Server pricing, find deeper technical information, or set up a live OCR demo, please contact us today.

Got SharePoint®? Then You Need Recognition Server.

Since its introduction about 15 years ago, SharePoint® has established a base of over 160 million users, emerging as one of the most widely used enterprise document management and collaboration platforms on the market.

For natively digital documents, like your spreadsheets and Word files, all is well. But it’s another story when it comes to making paper documents–the bane of offices everywhere–available across the enterprise. Document digitization initiatives are good and necessary, but they are not complete.

Sure, you can store document images in SharePoint easily enough, but it’s extremely inefficient (at best) for users to actually navigate these images. For instance, recall the frustrations you’ve probably encountered trying to hunt through image-only PDF files. Continue reading

Recognition Server Review: OCR Technology Empowers Legal Claims Team

It’s tough even to imagine more paper-intensive work than that of LA-based law firm Kantor & Kantor. As insurance claims specialists, they routinely see individual files numbering well into the thousands of pages. Perhaps more importantly, they must often peruse entire sets of documentation to find tiny yet critical pieces of information–which can determine someone’s entire financial and medical future.In short, much of their valuable time (and clients’ limited budgets) was spent looking for needles in haystacks.

Kantor & Kantor’s management was always technologically forward-thinking, but early OCR efforts fell victim to slow, inaccurate software. Even with the old software in place, it still took five human assistants to manage the “automatic” process! Continue reading