There are many steps to choosing the right vendor for electronic document production and there are methodologies to choose a vendor beyond price.  The one issue that is probably the most difficult, however, is to objectively compare, price.  Each vendor has its own pricing strategy and it is difficult to determine which proposal is inclusive of all services and addresses the issues important to you. 

The Center for Computer Forensics attempts to simplify the proposal by breaking the fees down by each stage of the production.  Since it isn't necessary to use every single step it allows you to compare each project on its needs. 

For you to control the process, offer the vendors the same assumptions on the variables.  For example, provide an amount of data in gigabytes then offer the assumptive number of pages per gigabyte.  Another example is for spreadsheets, offer the average number of pages for each excel.  If you expect a number of PDF files and you want any PDF file OCRed to create a text file, provide the total number of PDFs with an assumed percentage needing an associated text file created.   Anything that can be left to the imagination, provide an assumption. 

Following is the Center for Computer Forensics' fee structure

Data harvest / Data preservation - We charge by the hour to go into the field.

Data load - No fee charged for data loads.

De-duplication - We charge by the gigabytes loaded into this phase, uncompressed

Keyword search - We charge by the gigabytes loaded into this phase, uncompressed.

Load set production - We charge by the gigabytes that has to be loaded into this phase, uncompressed.  There is a small fee for bates, watermarks, etc. we need to be aware of  your requirements prior to creating the proposal

Keep in mind that each step in the process gets progressively more expensive so the more effective the de-duplication and keyword search, the greater the savings in the final process.

There are other processes such as electronic documents without an associated text file, we charge by the page to create the text file.  If associated text files need to be removed we charge by the page remove these as well.  There are a few other small fee services that should be considered and included in the original assumptions. 

If you have the time, create a sheet for each vendor to fill in with their fees for each step of the process and you will be able to understand the entire process.  This sheet should include a non-disclosure statement as vendors don't want to publish their pricing.

This process may not be perfect but it will take some of the orange out the apple bin. 

*    *    *    *    *    *    *    *    *    *    *    *    *    *    *    * 

There are a number of terms above that need to be clarified so you can understand when to use an EDD service and how to manage the production. 

The purpose of these EDD services and the e-discovery vendor is the reduction of documents (fewer documents to review) and the preparation of the data into a form that streamlines the review process. Following are general descriptions of terms: 

Data Reduction Processes

De-duplication - This process, as the name implies, is the reduction of duplicate utilizing different automated processes depending on the task.  There are several methods depending on what you requirements of the case.

Electronic documents (All electronic files except e-mail) - This process uses an algorithm to assign a unique number to the document and then compares the numbers to all the documents to identify a duplicate document.  The possibility of two unique documents having the same number is very remote and the courts have accepted this process as valid.  The purpose of assigning the number is to convert the document into a language the computer understands. 

E-mail - The data within an e-mail includes time and date stamps and other variable data when the e-mail was sent or arrived on an e-mail server.  For this reason, the algorithm method doesn't work as every single e-mail is unique.  The solution is field de-duplication that allows you to apply the algorithm to each field of the e-mail.  In general the fields include To, From, Cc, Bcc, Subject and Text Body.  There are pitfalls to this method and you need to work with your vendor to determine which fields should be used to ensure you get the responsive documents you need for the production. 

Given the pitfalls of e-mail de-duplication, the process is not a pure as the electronic document de-duplication so you will see duplicate e-mails. 

Keyword search - Once de-duplication is completed, a keyword search is executed to extract responsive documents from the database.  

Data preparation or load set creation - Most clients use a case management software program such as Summation, Concordance, Ringtail or an on-line review tool such as i-Conect or Catalyst to review the produced documents.  The data preparation step is the last step in the process and given these products we create a load set that allows for the client's review.

The most common load set is a TIFF of each document with an associated text file.  A TIFF is a picture format so you have a picture of the document and the associated text file makes the TIFF searchable.  The advantage of using TIFFs include fast movement from one image to the next, annotation of the image, protection of metadata on the native file and no requirement to own the native document.