Recent studies have shown that data entry is one of the most hateful and unnecessary tasks, which raises the question why this task remains manual in the age of artificial intelligence, data mining, and technology.
Is there a way to make it less hateful?
My ongoing fieldwork in data-driven startups, called AgeShi (an honest company, but not its real name due to privacy requirements), suggests that technical solutions are not as complex as many think and will never replace human data entry.
For almost two years now, I’ve been studying the evolution of recruiting and career practices.
Get your news from people you know what they’re talking about.
Its initial plan was to develop and use artificial intelligence to sell it to customers as part of its comprehensive services. Meanwhile, AgeShi has asked its analysts to collect and enter that data manually. But when AgeShi ran a pilot project with an AI consulting firm, they discovered that AI would create a maximum of five percent of the data it collects manually and at a much higher cost.
They could not afford artificial intelligence. As a result, they shifted their data collection strategy from artificial intelligence to human intelligence. Realizing that analysts are costly and are particularly reluctant to do this work, AgeShi hired a dedicated data entry operator in an affiliated office to do most of the work.
Technology doesn’t always work.
This type of situation is not uncommon. Technology does not always work as expected and has different implications for different tasks and organizations. However, technology failures only partially explain why manual data entry remains a task, and it may be for some time. To understand more, you need to look at functions in a broad context.
Tasks rarely exist in isolation. They are part of jobs, and those tasks have other functions. They relate to the people who perform them, others who run or work with them, and others in business and organizations.
This broad context and their relationship make it difficult to remove any task altogether, but it also means that a despised job does not always imply wholly despised.
In AgeShi, the data entry function becomes almost inseparable from the data collection process. I can easily find some of the data in the annual reports; for the rest, analysts and then data entry operators had to search the internet for additional data.
Once they found this, they had to enter it into the database.
It makes no sense, and it is virtually impossible to automate the data entry part of this function.
This highlights that not all data entry jobs are created equal. As a result, the approach to data entry is more complex than recent surveys suggest.
Wrong guess
Analysts have hated the work of data entry, but not just because of the nature of the work. Expectations shaped their positions. When hired, many expected them to do what analysts say they would do – such as creating reports from data, writing content, and interacting with clients.
Some of them initially applied for advisory positions that included the above responsibilities. They expected more elegant and perfect work. It is not surprising that they considered data entry boring and considered it less than that.
On the other hand, data entry operators expect what they mean by a job title – data entry; contrary to analysts, some reported that they were surprised by the job and the amount of thought and decision.
Even in the same job in the same organization, there may be differences in what employees do, resulting in different consequences.
For example, one study found that the U.S. Transportation Security Administration agents did a much worse job of investigating than male agents. The result is lower-quality jobs and fewer opportunities.
Enhanced data entry operator
The story in In AgeShi is far from complete. I still see missions and roles evolving there. The recent strategic point in the company is that data collection has become less critical for the company’s goal.
Through this axis, the role of analysts has evolved to interact more with company developers to create products for customers and work harder for production.
However, the functionality of the data entry operator is mainly unchanged. People come and go through it, but some people who do data entry have been promoted to analyst positions, suggesting that a threatening job may be the way to a less lousy job.
