Tuesday, August 10, 2010
Document Preparation is the process between the receipt of the physical document and the scanning of it. It is not very skilled work, but it is important for the smooth operation of your scanning. In a Distributed Scanning environment it is common that the person doing the scanning will be the one doing the prep work.
So, what does prep work entail exactly? Well, it depends on how the document is received, but here are some common tasks:
- Opening envelopes
- Removing staples & paper clips
- Smoothing folds and crumples
- Inserting separator sheets
- Entering document details in a register
- Creating a batch
- Capturing batch details
As you can see, there is a great deal of work that needs to be done. Much of it is manual and repetitive, and thus not ideal for highly skilled workers. It is possible that you could have a centralized Document Prep area even if you're working in a Distributed Scanning environment, but it is usually associated with Centralised Scanning.
Since little of the document prep work can be automated easily or cheaply, it is important to record metrics manually in order to be able to assess the health of this department. Some useful metrics include:
- Amount of items received, by type (e.g. mail, drop off etc)
- Weight of items received, by type
- Number of batches
- Size of each batch in documents and pages
By tracking the batch sizes, you can compare these with the batches actually scanned to to help verify the quality of your scans. A common scan problem is that multiple pages might be fed into the scanner simultaneously. By tracking your batches from the beginning, you can compare your expected page count against your actual page count, and flag batches where this might have occurred. If this is caught before the document is captured, it can save you a great many headaches down the line.
A well run document preparation department can improve the functioning of your scanning department dramatically.
Looking for a Document Management System? Signate 2010 is powerful, secure and easy to use.
Ask about our upcoming Mail Room feature that controls the life-cycles of documents from receipt to capture.
Thursday, August 05, 2010
Unlike with Distributed Scanning, Distributed Capture has far less going for it comparatively. In order to capture information about a document accurately, it helps if the capturers have good business knowledge around the document. This mitigates against them being "generalists", capable of handling every type of document consumed by the organisation.
Instead, it makes a great deal of sense that the capturers be subject matter experts and skilled in the capture of certain types of data. For example, in a medical aid, the Claims Department might have it's own capturers dedicated to processing incoming claim forms.
However, (you knew there was going to be a however), there is one major downside of having a distributed capture department: delays. In our Claims department example, it is likely that the department would be swamped close to month-end as members try and get their claims in before the payment deadline. This would mean that you get a backlog, and documents in the backlog aren't fully accessible since their information hasn't been captured.
If that's fine, and you're happy to live with that, great. But often you'll find that the backlog causes more work; clients phone in to find out how their claim is doing, you haven't got to it yet, which makes the client irate and takes time from your already busy staff.
So, for clients where getting the documents accessible quickly is critical, I usually used to advise a "double capture" approach. What this means is that you have a centralised capture department with fairly broad but not deep training. It is their job to capture a very small subset of the document information: the document type, basic client details and perhaps the document date.
The document workflow (your system does have document workflow?) then uses that information to route the document to the correct department, for example Claims, for more in-depth capture and processing. This offers a few advantages:
- Documents are easily findable quicker.
- You can have ONE fax number or postal address, reducing customer confusion, increasing control, and reducing costs
- Your document processing metrics are more accurate and detailed
Obviously you get these advantages at the cost of extra hardware and extra staff, so I would recommend against it unless turnaround time on getting your documents in is critical. If you have a really good document management system, it might read the document itself and make it available. Even though it might not be quite as accurate as a human, it should be good enough to find the documents, which might reduce or eliminate your need for "double capture".
In this, there is no simple answer. Distributed capture is great in most scenarios, but can result in less accurate metrics, customer confusion, and backlogs. Completely centralised capture is usually a very bad idea, although I have seen some companies really make it work; but that usually requires very high degrees of staff training and very expensive software.
My "double capture" option is highly effective and can lead to impressive turnaround times but it can be quite expensive. Using software to do the first part of the "double capture" is a highly effective technique, and scales very nicely indeed. Unfortunately most document management systems which perform this up-front processing are very expensive, or have it as an expensive add-on.
Looking for a Document Management System? Signate 2010 is powerful, secure and easy to use.
It comes with built-in full-text indexing that you can use to index your documents up-front, making them searchable before a human even looks at them.
Monday, August 02, 2010
How to set up your document scanning is a bit of a tricky one. The two extremes are:
- Distributed - Each area in your business scans it's own documents.
- Centralised - You have a dedicated Mail Room which handles physical documents.
Benefits of Distributed Scanning
- Staff can be moved in and out of scanning roles relatively easily
- Staff are more familiar with the business area
- Departments feel more in control
Benefits of Centralised Scanning
- Staff are more focused on scanning
- Scanners are fewer and faster; major cost savings on hardware, maintenance and licensing
- Sometimes possible to time peak loads in various departments to "average out" load overall
The best solution will depend a lot on your business. However, in my experience, the Centralised Mail Room is the best from a cost perspective. Usually it is the best from a turnaround perspective as well.
Where a distributed model works relatively well is with knowledge workers; however I feel that a well run Mail Room works even better because you're more certain that your document handling procedures are being followed correctly.
I hate to have "one size fits all" advice, but in this case I must say that I find Mail Room a compelling advantage across the board. I've been involved with many companies that have moved from Distributed to Centralised, but almost none that have been moving the other way around. If you've had an opposite experience, please contact me, I'd love to hear more.
Looking for a Document Management System? Signate 2010 is powerful, secure and easy to use.
Wednesday, July 21, 2010
Search is the killer app for the web, and I firmly believe it's the killer app for document management as well. Pretty much all document management systems have a search facility, but most of them make pretty much the same mistakes.
Tying Search to Metadata
Most document management systems allow you to capture metadata, data about your document. This metadata is normally used in your business processes and also to find the document later. As an example, for a document about a person, the metadata may indicate the persons name, date of birth, and identity number. The search is often tied to this, so when you want to search it has a field for the name, and a field for the date of birth, and one for the identity number.
Simple enough, and seems compelling, especially if you can have several types of metadata (also called document types, catalogs, or schemas). Except often the search screens are suboptimal, not allowing you to select a range of dates for the date of birth as an example. Consider if you had several catalogs: Person, Company, Invoice, Application Form, and you wanted to search across all these catalogs (more than one might have a Name field as an example). The approach indicated above falls apart quickly at that point. Oh, some systems try and do this, but the user interface becomes cluttered and confusing.
Searching only on Metadata
Metadata is normally a subset of the data on the form. It is the data that interests you, as a business. However, there might be other information on the form that would help you find the document. Consider an example of a customer phoning in about an old medical claim, they don't have the membership number, nor do the remember exactly when it was for, but they do remember that it was a doctor in Durban. If you didn't capture all the doctors details (and normally his address is not terribly important), you might struggle to find the relevant claim.
This is why it is important to include the data that isn't critical to your business as well. Except it would be prohibitively expensive to do so. Well, this is where OCR (Optical Character Recognition) comes into it's own. It is not perfectly accurate, but that's okay, we're just using it's data to enhance our search results. For electronic documents, such as Word Documents we can also extract and index all the text in the document (called Full Text Indexing).
Too Many Bells and Whistles
I've seen search screens with every device and option under the Sun. The options are so dazzling and so confusing that users dread looking for their documents. Now, I'm not saying those options can't be useful; they certainly can be, but in almost all cases all you need for a search is a simple box for the user to type what text they're looking for. Put all that confusing stuff on the Advanced Search page. Think of Google, how simple it is to find documents quickly and easily, how you can refine your search to narrow down on the specific document you're looking for.
Simplicity, Simplicity, Simplicity
So, consider a system which does not have the shortcomings outlined above. One where there is structured capture for the truly important data, full text indexing and OCR for the rest, and a simple search box which returns results swiftly and easily. Training costs are reduced, because all you need to say is "it works like Google" and you've already got half your staff on board. Customer query turnaround times are reduced, as well as your costs, and all without sacrificing one jot of power or functionality.
So, next time you see a product demo, tell the sales weasel that you want less bells and whistles on the search page, not more. Tell them that you want simplicity and ease of use.
Looking for a Document Management System? Signate 2010 is powerful, secure and easy to use.
Oh, and it incorporates a simple search, full text indexing, and OCR.
So, you've heard that document management systems pay for themselves? Is it true? Yes, absolutely, in most cases.
What you need to do is consider where your current document management processes are costing you, and how much you could expect to save by implementing an electronic system.
Storage Costs
Your physical documents must be stored somewhere. Normally you'd store a fair amount, if not all, at your business premises. Unfortunately, business office space is not cheap. Measure off the square footage allocated to document storage, and work out it's monthly rental.
Organisation Costs
You must take the time to file your documents correctly, normally you'd have dedicated staff for this, or your staff would do it as part of their normal duties. Add the amount of the staff salaries dedicated to filing documents correctly (and refiling them after they've been used).
Copying Costs
Physical documents are very useful, but only one person can use them at a time. To get around this, people copy documents. This costs money, and can also lead to wasted effort and missed communication when someone attaches a note to their copy, but another copy is passed along the process.
Offsite Costs
If you store some of your documents offsite, good for you! That means you recognise that some documents are less required than others, have systems in place to identify them, and have been proactive in turning this knowledge to cutting costs. However, whilst the offsite storage is cheaper than at your primary business premises, the costs are still significant, and calling for documents you need can be time consuming.
Sharing Difficulties
Whilst not a direct cost, the fact that paper documents cannot be shared impacts your turnaround times, which can have a negative impact on your business image, and increase your customer care costs. Only one person can work on the document at a time, reducing your ability to streamline your processes.
Total
Total Monthly Costs = Storage + Offsite + Organisation + Copying + Sharing
Turnaround Impacts: Organisation, Offsite, Copying, Sharing
Document Management Systems
A well-implemented document management system should be able to significantly reduce or eliminate storage, offsite and organisation costs. Copying should be greatly reduced, and sharing should be vastly improved.
So, now you have your costs, give or take a bit. You should be able to save about 50% of those costs by implementing a document management system. In other words, to pay for itself, your proposed document management system should cost you less over 2 years than one full years overall document costs.
If you're not sure how much your costs are, you can use this Document Costs Calculator to get an approximate amount.
Looking for a Document Management System? Signate 2010 is powerful, secure and easy to use.
Tuesday, July 20, 2010
Capturing data about your documents (commonly referred to as metadata) is critically important for your business. You want the data captured to be accurate, yet you also want the process to be quick in order to ensure that you can maximise your staff's productivity. Especially in specialised industries such as medical, financial, and insurance amoung others, it is important that the capture staff be skilled, in order to flag problems and capture accurately.
This means they're expensive, and it means that if your software is wasting their time, it's doing so continually, putting a slow puncture on productivity.
The areas where software can impede productivity fall into the following main categories:
- Crashes and Hangs - Faults in the software or hardware cause the system to be unavailable. This results in massive productivity loss as it usually affects all your staff, and often for a significant period.
- Loading - The software is slow in loading the documents to view, and/or the form to use to capture.
- Validation - The software is slow to validate the fields that need to be captured.
Crashes and Hangs are difficult to eliminate entirely, but choosing a reputable vendor with a passion for quality is normally your best bet in this regard. A good testing cycle and reliable network and computer systems are also critical.
Loading delays are simple to remove, although far too few vendors make the effort to do so. Pre-fetching documents to be captured, pre-loading the capture forms, and keeping recently used information are common techniques to minimize this type of delay. A vendor which removes a mere second from loading delays can save you about 30 hours productivity per staff member per year.
Validation delays are even more difficult to remove correctly. One common technique is not to do validation at all; that this is even considered is appalling, that it is common is terrifying. Another common technique is to avoid validation until the user submits the form. This results in the user having to wait whilst validations are performed, wasting productivity. The best solution is to validate the data as the user enters it, but it is important that this validation happen without blocking the user from continuing entering data into other fields. This requires concurrency, which is tricky for the vendor to implement correctly, especially when many fields are involved and the users capture data quickly. As a result, this technique is commonly reserved for dedicated and expensive capture software products.
So now let's consider how much money can be saved from improving loading and validation delays. Assume a system with a 3 second load time, 2 seconds validation time for all the fields. With 12 capture staff processing about 400 documents a day each, this works out to about 350 hours lost per year, or 1.3 person years.
As you can see, those small increments can add up to a lot, and they get worse and worse the more staff you add. High speed capture is a compelling solution, but unfortunately most such products are targeted at businesses where the data capture is time sensitive for other reasons; and thus are vastly more expensive than these simple productivity gains could justify.
Looking for a Document Management System? Signate 2010 is powerful, secure and easy to use.
Oh, and it incorporates high-speed capture as part of the standard package.
It's a wonderful conceit isn't it? Scan in some paper form sent to you, and have the computer automatically read it and process it.
Unfortunately, that's all it is, a conceit. The reality is that reading handwritten text is hard. Typewritten text on the other hand is largely a solved problem, with up to 99% accuracy rates. Handwriting, well, not so much. Oh, it's possible to jimmy pretty good numbers out of the software, and with enough training it can get pretty accurate. Another way of improving accuracy rates is by being able to set a lexicon; a list of allowed words.
So, given that you can limit the words that people use, and given that you can force them to print (cursive is often still a problem), then yes, you can get some pretty impressive accuracy: 95% per word in some cases. However, this is usually not good enough for business; consider that a 95% per word accuracy translates to the software getting more than 3 of the words in this paragraph alone wrong.
For making decisions involving money, or people's health, that's just not an acceptable level of accuracy. So, it seems we are still wedded to manual capture where humans type out what other humans have written in. Or are we? Well, sure, you can't use OCR for the truly critical data on the scanned form, but you can use it for less critical data.
You may not want to use the OCR data for making business decisions, but you can use it as an aid to finding your documents. In many cases, your computerised business processes need only a subset of data on a form, with humans assessing the remainder. In such a case the process is simple: capture the critical data manually, index the rest using OCR, and use both sets of data to find your documents.
This gives you the best of all worlds: the accuracy that only human capture can currently provide, limiting your human processing to the bare minimum, and being able to find the document based on all the data on the form.
Looking for a Document Management System? Signate 2010 is powerful, secure and easy to use.
Monday, July 19, 2010
One of the most common idioms in document management systems is the folder, a hierarchy which documents are placed in to control their organisation. Such a hierarchy is incredibly powerful as it captures parent-child relationships as well as providing easy administration. For the software vendor, folders are easy to implement, and easy to understand.
Unfortunately, this is where the advantages of folders pretty much end. Folders are brittle, in that you must continually maintain them, reorganising them as documents arrive and projects change. For relatively small sets of documents, such as policy documents, specifications and so on, this administrative burden is not excessive.
For larger sets, the work involved becomes immense. Consider the Internet as an example; in the old days, before search engines became ubiquitous, web pages were organised into directories, many of which still exist today. You browsed the directory, looking for the pages that interested you.
Many pages did not fall into simple categories however, and the difficulty involved in keeping the directories current; adding new documents and pruning old ones became exorbitant. Then came the search engines, which crawled the Internet, and allowed you to search for content based on keywords. This swiftly became the dominant form of navigation on the web.
Document management systems today are still wedded to the folders idiom, despite the obvious advantages of search for a few main reasons:
- Unlike web pages, normal documents don't link to each other, making crawling all but impossible
- Folders are familiar for most customers
- Search is more difficult to prove to be accurate
- Search engines are difficult, very difficult to get right
Most of these reasons are good for the vendor, but not good for the customer using the software. This is a great pity, since search is such a compelling way of accessing documents.
Looking for a Document Management System? Signate 2010 is powerful, secure and easy to use.
Saturday, July 10, 2010
My name is Sean, and I'm one of the architects on the Signate Document Management product.
My purpose with this blog is to try and provide insight into document management in general. I'll try and keep the focus fairly non-technical, and non-product specific. I hope this resource will be useful to people or companies planning on implementing a document management strategy.
Looking for a Document Management System? Signate 2010 is powerful, secure and easy to use.