The Data Entry Scam: Why the Odds of Winning are Heavily Stacked Against You

How a ‘legal fraud’ and defined processes will ensure that you hardly make any money.

* Of course, it depends on the contract.

When I recently heard about data-entry jobs, I was first surprised that this even exists today in this day and age. Wouldn’t it all be automated or at least refined through automation? Apparently not. And there is a reason: it’s cheaper to get people in third-world countries with the lure of big money to get it done for you while scheming the contract in such a way as to give them next to nothing.

Yes, you could say that the people accepting the contract should know better. And that would be fair. For those who might not know, this is my analysis of what is probably happening and my attempt at letting them know better. So, this is some math and statistics as much as an outsider can guess.

Various Pieces of the Racket

First, we need to understand how the contracts are probably typically drawn. Disclaimer: I’ve not personally seen a single data-entry contract. But those who have seen one or have signed one, if these points ring a bell or if there is something in error, please let me know.

Tiered Pricing

A typical contract will have a tiered pricing based on received quality and also quantity. For the sake of our example, assume a graph such as the one below. This is seemingly all good. It seems to be based on merit and quality of the data that is entered. You get nothing if quality is less than 80%, and then there are large increases as quality approaches 100%. All you have to do is make sure that quality is great. Hmmm, not really.

Quantity, oh and also Time

It’s not enough that you hit a certain quality, it’s absolutely necessary that you do it within a certain time and also complete a certain quantity. But why would somebody do that? If their real intention was to ensure that the data is entered properly, why would they stress on limited time and fairly large quantities? We’re just on to point two, and you’re already getting suspicious right? Probably because they really don’t want you to do well might be your thought. Maybe. Let’s move ahead.

Cost per Seat

This is a minor point but it adds up in the next one. You have to pay the agency a certain cost to acquire the data to be typed in, or per seat/software that the data will be entered in. Let’s say it’s somewhere between Rs. 25,000 and Rs. 50,000 per seat. With that, you are already down in terms of overall cost before you even started. Want to hire 5 people to make a business of it — you’ve already spent a few lakhs of rupees.

Locked Out Software

The data has to usually be entered in a specific software. And that is completely locked out except for certain hours — as in, you won’t be able to login or enter data except for a few hours during the day. Again you wonder, why!

Not only that, when their software comes up, some of them completely kill all other programs running on the system. Extreme measures. Hmmm, again … oh well. Copy-paste is also entirely disabled. This combined with the fact that you’ve paid for a certain number of seats on which alone you can edit or type this data. With this setup you can’t even get the data typed out offline or during off hours and then copied in later. Sorry, no can do.

Effectively then, you have to enter a large amount of data that they decide, within a limited amount of days that they impose, within a few copies of the paid software which is only available for a limited amount of time, and they need it to be at high quality. Why all these restrictions? Could it be that they are putting in every kind of restriction so that you cannot do well? Oh, come on, what kind of exploitative, sick people would do that kind of stuff?

The Lure of Big Pay (and Urban Legends of People Who Made It)

This is one of the biggest carrots that keeps this business continuously churning. The payout for 100% quality is huge. If you made it. IF. And there is talk of certain companies that did it and are now rolling in money. Could they actually give us 5 references of legitimate companies that we can verify this with? I verily doubt it.

On the other hand, one person I followed up with over a phone call said that he’d shut down the business. It clearly wasn’t working. You won’t get these references from the agency, but really, these are the ones you should be following up on.

Dramatically marketing the upside is how the agency gets you to sign on. Like a Las Vegas casino. They have huge posters of that one guy who won a lot of money. But do you know that the casino’s slot machines and other gambling opportunities are setup in such a way that the odds are always in its own favor. Basically, you couldn’t really win. The odd winner will slip through and they’ll show that up with big banners and hoardings everywhere. But most of the remaining are playing to lose. Because the game is schemed in such a way that you never win.

Statistics and Typing Accuracy

We need to make one more digression to the study of typing speed and accuracy to build up to another point. Search the net and you’ll find some research on typing speed and accuracy. Agreed that some of this research was done for typewriters, but seemingly, the results aren’t dramatically different for computer keywords. Various research suggests accuracy of about 91% to 94%, i.e. if you typed 100 words, you are likely to make between 9 to 6 typos.

So, on average, let’s say that a person typing in a form makes a few mistakes. Assuming he’s really good, let’s say he’s made only 5% mistakes. But you, being a conscientious businessman who really cares about the customer and wants to provide high quality, have invested more resources for quality checks and hired extra people to review what each person has typed. With all that extra invested money and effort, you’ve brought down the errors further to say, just 2%. Or you want to give them an even higher benefit of doubt? Sure, go ahead. You’re never going to come out a winner anyway. So now in our highly quality checked world, we’ve made only 1 character mistake for 100 words. Or 99% accuracy on words, and on a character basis, higher than that also.

Surely you’re thinking now that you can laugh away to the bank because you’re delivering 99% quality? Ha ha. No, that’s not you laughing. It’s that crook who’s drawn up this contract.

The Real Quality Based Payout

And here is the most treacherous part of this whole business: The quality assessment for payout is not at all the quality you give them. Read the contract again. Here’s the kind of stuff you’re likely to find…

If there is one mistake in the form, one error is counted.

Or some other wording like that. Sounds and looks fair, right? Read the contract again. As I mentioned before, I haven’t seen a full contract myself but when people told me something like this first, I took it at face value. After all, that seems to be the decent, normal thing to do. But you aren’t dealing with those kinds of people, are you?

What they mean with statement like this is: if they find even one typo in one document/form, then the entire piece is rejected. The few people I tried to explain this to refused to accept this easily. At least look at the quality reports they are sending you and assess it for yourself.

Let’s take an example of one form that has to be filled in. A form with 100 characters. And that one form is your entire submission for the month. i.e. 100% of the output you are going to deliver. Let’s say that the form has 2 characters typed wrong. What is the % errors? Two answers:

  • your belief might be the reasonable one based on characters: 2/100 = 2% errors.
  • according to them, based on forms: 1/1 = 100% errors.

What? So what they do is, if there is even one error in the entire form, they reject the whole thing. And they calculate the quality based on overall forms, where even one typo calls for rejecting the entire form.

In effect for you, it’s either consistently 100% quality or nothing. Now think about the typing accuracy study. Despite all the efforts and investments you make, errors will be there. Consciously aggravate it by asking people to type a lot during a short time and they’ve guaranteed that errors will actually increase. Therefore, there is pretty much no chance of you ever receiving a high quality report. Really people, rethink hard about signing that contract.

If there are statements about how they are doing a random sample to do the quality check, it’s probably a sham. Given the average number of typos, if they haven’t found a number of errors that suits them, they just have to look elsewhere in another ‘random sample’ and they’ll find the errors they want. This whole thing is opaquely setup so that you always get the short end of the stick.

But … then … why?! And what do they gain?

These exploitative slave-drivers are playing you. If you followed the example above, you can see that you are delivering high quality (say 99%) but you will be paid for much lower than that (say 84%). And the payout is totally different in these two cases — even 1000% — from Rs. 20 at 99% to Rs. 2 at 84%.

If you bought something from an online store and you deem it’s quality or fit to not suit you, then the online store asks you to return the original. You don’t get to keep the first one you complained about. But unlike physical goods, these guys have the digital data you typed and sent them. They don’t give you a way to assess the ‘quality’ locally before submitting it. They now hold data entered at very high quality while they pay you peanuts. You suckers! They are going to turn it around to their clients and get paid for the very high quality slab.

Also, they’re probably two-timing you. Cheats in every way.

Do you know how companies like Google and Amazon ensure that you have access to their application servers 99.99% of the time. There are many technologies involved for sure, but one of the main ideas is redundancy. They have multiple machines holding and serving the same data. If one machine went down because of a power outage, another one will pick up the slack.

My guess is that the agencies giving out the data-entry contracts are giving the same data to multiple small players. The chances that two or three sets of people typing the same data make the same mistakes is extremely low. So now all they have to do is wait for each of you two submit it, then compare the mistakes, fix them alone and further improve their quality — they’ll get very very close to 100% quality. Meanwhile of course, they are rejecting you by the form, then assessing you as delivering a little more than 80% quality, and throwing a few coins in your direction.

There are No Second Chances

Let’s say that you really want to increase the payout. Ask the agency whether you can resubmit the data with the corrections made. If their intention was ever to pay you decently, they would agree. Try it, go ahead.

You are likely to hear a “No!” because the whole scam is based on ensuring you make mistakes and then they pay you next to nothing.

I want to know more

If you people have experience with such businesses, I’d like to hear from you. Tweet me at @sathishvj or comment below. Let’s catch up and you can pour your woes to me. And let’s get this information out in the open so that they can’t exploit people like they are doing now.