Perform HITs on Amazon Mechanical Turk

Did you know that Amazon has a site on which you can be paid for performing microtasks right in your own home? I'm an avid Amazon user and I didn't learn about this site until about a year ago. While it won't make you rich anytime soon, it is a fun way to make a few extra dollars. 

Amazon Mechanical Turk is a site with Amazon where a Requester posts a HIT (Human Intelligence Task) which a Worker performs for a reward.  HITs are usually microtasks. I've seen HITs paying as little as one penny as well as ones paying as much as $70. I've personally completed ones that paid $18.  There are thousands of HITs available every day. The highest paying HITs (as well as a few others) usually require the worker to take qualification tests. HITs are also timed. The worker only has so much time to either submit or return a HIT before it expires and is made available to other Turkers.

Payment is made to the Workers' Amazon Payments accounts and may be withdrawn as direct deposits ($1 minimum) to their bank accounts or in the form of electronic Amazon Gift Card ($1 minimum).  Furthermore, the funds in the Amazon Payments account may be used to purchase most items on Amazon, with a few exceptions.  

There is a 48-hour waiting period when signing up for a new MTurk account.  If you don't already have an Amazon Payments account, the waiting period is a good time to set that up. There may be a verification process. Additionally, there is an active-10-day holding period on payments for new Workers.

Mechanical Turk does not have a referral system, so the only way to make money is by doing HITs and having them approved. 

UPDATE: I started using Mechanical Turk on August 14, 2013, and have (as of 9/29/2014) earned $288.26 with about $20 worth of HITs pending approval.  The bulk of my earnings are from participating in surveys and transcribing audio files.

Surveys and Studies

There are always new surveys and studies to take at least once a week. They vary greatly in time commitment and payment amounts. Some pay bonuses. Some offer giveaway entries. Most have qualification questions to determine eligibility. Quite a few also have attention check throughout so it is important to read carefully give truthful responses. Requesters sometimes ask high-quality workers to complete their invitation-only surveys or studies. Lately I've mostly done surveys that pay at least 50 cents, unless a potential bonus made the task worthwhile or the topic was especially interesting to me.

A Few Recommended Requesters

For transcription HITs:

• I would definitely recommend CrowdSurf. They offer short 5 to 35 second transcription (and review) HITs that usually pay between 5 and 20 cents depending on the length and urgency of the HIT. Since these HITs aren't particularly time-consuming (each can be completed in under 10 minutes), the ability to do a bulk of these can add up to some nice pocket change. The best part about these HITs is bonuses. They pay weekly and weekend bonuses based on bulk. The bonus terms change periodically, but you can read about the latest terms here. The general guidelines for transcribing with CrowdSurf can be found here. The work is reviewed and scored for quality control and if your work is especially poor quality or fraudulent, your qualifications are taken away. Transcribers also must agree to a NDA (Non-Disclosure Agreement) due to the potentially confidential information contained on the audio files. Finally, since the tasks are brief, the turnaround rate between submission, acceptance, and payment tends to be relatively short when compared to other HITs on MTurk. 

• SpeechInk has a variety of transcription and captioning HITs including both verbatim and non-verbatim jobs. Some of their jobs also require time stamps. Their HITs categories also require relatively in-depth qualification tests in comparison to CrowdSurf. Their HITs are longer, ranging from minutes to hours, and the payments vary greatly from about 40 cents to over a hundred dollars depending on the HIT type, length, and difficulty. SpeechInk is essentially the highest paying requester of MTurk, so if you are willing to take their tests and are willing to spend about half a day transcribing 1-2 hours of audio for $30-$50, SpeechInk should be your go-to requester. Like CrowdSurf, SpeechInk reviews for quality control and has a NDA. 

For mischellenous HITs: 

• CopyText pays a few cents to copy text from receipts. The base of pay is only a penny for 3 lines of text, but they pay small bonuses based on the number of lines accurately typed per HIT. They have a tiered system where the most HITs are temporarily reserved for their high-level workers. The receipt text is in French. The specific guidelines for these HITs can be found here.

• Grace Kim offers bulk HITs. I try to check for them every few days when I'm bored with survey or transcription HITs. For these HITs you have to watch short surveillance videos (from a few seconds to almost 2 minutes long) and click buttons whenever someone enters or exits. You are shown each video twice. On one occasion you click when they go left, and the other time you click the button when they go right. I really feel like a voyeur when doing these tasks. The base pay is usually 5-10 cents plus bonuses, which seem to be rewarded based on accuracy.

• YCharts requires a qualification test, but there isn't an actual test. You basically request qualification and are granted it. Another great thing is that all submitted HITs are accepted. YCharts HITs require workets to find different webpages (Wikipedia page, Events page, Investor Relations page, etc.) for different companies. They occasionally post bulk requests and typically pay 2-3 cents for locating a webpage. They also offer bonuses and achievements. YCharts uses majority rule in deciding the accuracy of submissions. Each task is given to at least 2-3 workers. Your qualification rating with YCharts increases or decreases based on whether at least one other worker agreed with your choice for the webpage. It goes up when there is a consensus and down when you are the outlier.