DPI designs quality into the data entry process by rigorously defining the business rules for each field in a record.  Data entry operators cannot enter data that violates any of the business rules associated with a particular field.  DPI also makes extensive use of table look-ups, formulas and error checking programs.  A simple example of a table look-up is to use the zip code to verify and improve the accuracy of the city and state data.  Through the use of accuracy improving programming techniques and by hiring only the most accurate data entry operators, DPI regularly exceeds its 97% accuracy guarantee for single-key data entry using USA labor.  However, when projects demand higher levels of accuracy and budgets allow, DPI can both key and verify data.  This involves a second operator re-keying what the first operator has already keyed.  When the second operator keys a different character from what the first operator keyed, the data entry software prevents further entry until the keystroke is confirmed.  Any or all fields can be keyed and verified.  Customers typically use “key and verify” for fields that are considered to be critical since this level of accuracy can be double the cost of a “key only” field.  The use of “key and verify” increases the DPI accuracy guarantee to 99.9% for USA labor.

To insure the quality meets DPI standards for higher volume offshore projects, (1) software can be written to validate each field in the output file against allowable values and (2) a statistically valid sample of batches is randomly verified by USA based DPI data entry operators.  Accuracy reports are produced and provided as feedback.  Individual records containing a field that does not match the validation criteria or records containing an error uncovered during random verification are sent to the offshore partner in an ongoing effort to positively impact quality.  Since labor costs are much lower the key and verify technique is generally used in offshore situations.  For clean images with typed data offshore accuracy can be 99.9% or higher.  The most problematic information for offshore labor is handwritten proper nouns such as names of people, business names or street addresses.  Since even English speaking offshore labor is generally unfamiliar with proper nouns, they often data enter what the individual letters appear to be.  With poor handwriting it is easy to confuse an “i” with an “l” or an “a” with an “o” or “u”.   Familiarity with proper nouns may allow a USA operator to look at the individual letters in the context of the entire word and potentially deciphering poor handwriting.  Legibility is heavily influenced by the design of the document and the demographics of the individuals filling out the form.  Based upon DPI’ experience verifying behind offshore data entry, 97% -99.9% is a reasonable accuracy rate to expect with handwritten forms containing numerous proper nouns.  To maintain control and insure data entry is completed for each document received from the customer, DPI creates an output file to be used by the offshore data entry facility with the desired record layout and a record corresponding to the index of each scanned image.

In the data entry marketplace accuracy can be measured at the character or field level.  A simple example to show the difference would be the data entry of a telephone number.  If one number was incorrect, the character basis would calculate as 9 correct characters out of 10 or 90% accuracy.  The field basis would calculate as 0 out of 1 fields correct or 0% accuracy.  Field level accuracy is the more stringent standard but some companies will quote character accuracy to make a higher accuracy guarantee.  The truth is most data entry companies do not go to the trouble of actually testing their own accuracy as this is a time consuming process that requires sophisticated programming.  DPI measures our data entry using field level accuracy.