With technology, computing devices, data, bandwidth, all around us, it might seem that we are living in a sort of futuristic utopia where human beings have to just show up and be served by machines that communicate with each other. But we know better, don’t we?
Data entry continues to occupy a prominent place in the support functions that power the digital world. While converting legacy manual data to digital formats may be one reason why manual data entry continues, we are also creating new technologies with their own need for manual data entry.
Take Artificial Intelligence (AI) and Machine Learning (ML) for example. While they are expected to play a key role in the world of the future, the very process of creating the model requires a huge amount of manual work like data annotation and labelling that needs to be done primarily manually.
Since data entry is an input that produces an output, it needs to be measured and managed to ensure that value is being created and not destroyed through that process. It offers us perhaps one of the purest and simplest measures in the form of speed at which data is entered.
In a world where it continues to play an important role, just what can we expect by way of data entry speed?
And is that all we need to look for? Or are there other parameters that also assume importance, such as accuracy of the work?
Speed at the time of typewriters
Computer keyboards have evolved from typewriter keyboards, or keys. For the uninitiated, typewriters were devices that were used to create ‘formal’ documents as opposed to informal or handwritten ones. Each letter of the alphabet was assigned to one key and by depressing it would leave an impression on the paper where it was required. As these were standard characters, once imprinted on paper, would mean the same thing to all readers, as opposed to handwritten documentation that could leave room for doubt.
These were purely manual devices that did not have any wiring or software code embedded into them. Also, accuracy was of utmost importance as errors could not merely be deleted as in word processing programs today. They would leave an impression on the paper.
In the days of typewriters, in terms of speed, a count of forty words was considered to be reasonable and, of course, the more you did beyond that, the better you were. At fifty you started becoming a professional. The measure used was words per minute (WPM) which was calculated by the operator typing out a text, the number of words in which was known in advance, and dividing that by the number of minutes taken to complete it.
This was overlaid with the accuracy produced to create an index for the quality of the output, or of the typist, with both speed and accuracy taken into account. The weightage given to speed and accuracy would vary based on the purpose.
Computer keyboard and impact on data entry speed
Though computer keyboards have evolved from the typewriter keys, some changes have also taken place. Computers being computing devices, the balance of power between letters and numbers has shifted towards a lot more of the key-pressing being of numbers as compared to typewriters. This is perhaps because numbers convert more readily to information that can be worked upon by computers. Many databases that are created through data entry also have numerical information, such as date of birth, scores, length, width, position, etc. These requirements have resulted in a numeric keypad being added to the keyboard, in addition to the one usually at the top.
Additionally, since these are electronic devices, information can also be gathered electronically. For example, it is now possible to track how many times a particular letter was pressed while creating a document or database, without manually counting.
Keyboards are no longer used mainly for the purpose of creating formal documents that can be read and understood by all in a similar manner. They are equally frequently used for creating databases and tables that collect and display information and lend themselves to calculations and computing. In these ‘documents,’ it is more difficult to define what a word is. Is a complete date a word? Or are the date, month and year three words in a date?
These changes have resulted in the evolution of a character per minute (CPM) or characters per hour (CPH) as a more standard measure of speed instead of the words per minute used earlier. Some users prefer to use ‘keystrokes’ instead of ‘character.’ This measure has an advantage over WPM, its predecessor, apart from being automatically generated. The advantage is that it reflects each character as it is pressed, hence it can be considered to be the most micro level measure that might be used in this exercise. It also makes up for the characters per word that might differ from one text to another, and includes both letters and numerals in its assessment.
What is a good data entry speed?
While there is no single answer to this question, there are some general expectations that have emerged.
It is now accepted that numeric entry is faster than text entry as it needs to deal only with 10 characters or keys while there are many more on the standard keyboard. This means that a task that is weighted in favor of more numeric data, should be handled faster than a one that has less numeric information.
Since most data entry tasks are a composite of numbers and text, 7000 CPH is now generally considered to be the minimum expectation with 8000 CPH onwards beginning to be in the acceptable range.
Let us compare this to the expectation from typists.
An average word in English has a length of 4.7 characters, as explained by
Wolf Garbe. Adding on a character for the space between words makes it 5.7. A typist typing 50 words in a minute would type 285 characters in a minute or about 17000 CPH, much higher than the expectation from data entry operators.
Working backwards, 8000 CPH would work out to under 25 WPM. In a way it makes sense as the average person types at a rate of about 25 WPM and data entry operators are drawn from a wide variety of backgrounds and skill-sets. Perhaps also makes us appreciate the professionalism and training of typists who were reaching and exceeding the 50 WPM mark. It has been an open secret in all organizations that have significant
data input requirements, that enhancing the typing skills of their staff can deliver significant efficiency benefits.
However, this only sets out general standards and expectations. Each assignment is different. There is a difference in the cost of inaccuracy. There is a different text and numeral balance. So, work it out for yourselves. Besides, when the volume is large and there are many people doing the same job, relativity in output between different people also throws up information that can be used to set our own benchmarks.
The Accuracy Balance
Data entry speed does not exist in isolation. There is always another side to the data entry picture, that of accuracy. It can be nobody’s case to make errors in data entry while striving for speed. You may have heard of the US Government losing over a million dollars in revenue in the late 19th century on account of a misplaced comma. In today’s money it would be at least twenty times more. Or many websites being blocked as a result of a data entry error at a major search engine over ten years back. For people responsible for doing or managing data entry, it might feel like a see-saw on occasions. Press down on speed and quality becomes lighter. Focus on quality and accuracy become lighter. But, that is life and that is business. One has to manage competing priorities.
It makes sense for businesses to be conscious of the need for accuracy. Of course, at what point does the fine balance lie, remains for each business to decide itself. However, here are some ways through which accuracy in data entry can be promoted:
Automation
This is an obvious solution. What can be automated will be automated. It is an ongoing effort. Not being subject to the unpredictability of human behavior, once automated, data entry becomes more predictable and accurate, while increasing data entry speed.
Two screens
This makes for greater screen space being available to the operator. In addition, it avoids frequent switching between screens as you have an alternate screen to bring up source and reference documents on.
Check and review process
Depending on the criticality of the process and the cost of making errors, a process control in the form of an external review by another person can locate and eliminate errors. Depending on criticality, it could be a 100% review or only a sample check. Based on trends identified in the sample check a more detailed analysis could be undertaken of areas that could be more at risk.
System checks
Modern software applications are versatile. If processes require manual data entry, they can be set up in many different ways to mitigate the errors, like validation rules. For example, where it is possible to define the expected entry in a field, any entry that does not meet that expectation can be rejected. For example, if a mobile number needs to be 10 digits, any entry of 9 or 11 digits will be rejected.
Systemic MIS and control reports
Since the equipment is electronic, it can be leveraged to produce information on performance and identify errors in the input in the form of outliers. If 99 out of 100 date fields have been input as DDMMYY and one as MMDDYY, the exception can be unearthed and rectified, if required. If 98 addresses are of San Diego and two of San Antonio, they can be caught in the exception report and checked.
Supportive environment
An environment conducive to work goes a long way in producing accuracy. If the operator can be allowed to focus on the job, instead of worrying about non-work issues like the office being too hot, or network speed being too slow, or cramped workstation, or too much background noise and movement, she is likely to produce better results. Given a choice between doing well and doing badly, all human beings would like to give a good account of themselves.
Data entry speed – what next?
The funny thing is that speed and quality may not be mutually exclusive. Many of the suggestions offered with a view to improving accuracy of data entry may end up improving data entry speed as well.
A supportive environment will promote good performance from all perspectives. Automation, once it kicks in, will make speed a meaningless measure. Systemic checks and controls, while ostensibly increasing accuracy and controlling errors, will, hopefully, provide confidence to the operator allowing him to do her job in an uninhibited manner.
These are, of course, methods that are external to the operator, the key person in the whole situation. We have seen earlier the quality and speed at which trained typists were able to operate on typewriters. Effort should be made to impart similar training to people who are doing the manual data entry task. This can release the greatest value. The question will be – does the operator see a long-term career in this line of work to invest time and money for improvement?
Leveraging the capability and experience of an outsourced organization like oWorkers remains a great option for companies that have a need for
data entry services.
With their status as preferred employers in all communities they work in, they receive a consistent flow of interested applicants for data entry and other jobs. This affords them a choice and they can select people based on their skills for different roles. This also enables them to offer handling peaks and troughs in client volumes, another substantial saving for clients. They can hire up to 100 resources in a 2-day period.
With their technology clients and other technology companies they have forged a relationship that enables them to use the latest technologies for their delivery. This helps clients access the latest technologies as well. They are GDPR compliant and ISO certified and operate from secure facilities in three major geographies. This creates business continuity planning options for clients who can opt for a multi-site delivery strategy.
Led by a team that has over 20 years of hands-on experience in the industry, oWorkers has been serving clients from around the world.
