CloudResearch Connect says non-approval decisions are final and it cannot give further information about them. Reapplying with the same or a different email does not unlock studies, and detected duplicate accounts are removed.
Connect's participant terms accept signups only from individuals located in the United States, United Kingdom, Australia, Canada, Ireland or New Zealand. Participants must be at least 18.
This is an application-location rule, not an explicit residence or citizenship list. Account approval and study matching remain separate: new signups undergo identity and survey review, and researchers can target studies by location, demographics or prior participation. The list does not grant travel or relocation permission.
Documented cash-out limits and conflicting payment guidance
Payment terms
Connect's cashout help lists PayPal transfers with a $5 balance minimum and $100 maximum, processed once daily. It says the full balance moves after confirmation, so that wording must be read alongside the stated maximum; handling a larger balance is not explained. PayPal must be linked first, and switching or unlinking that account is no longer supported. The help page says participants outside the US can only cash out to PayPal, while also saying bank transfers are available to all participants. It lists bank transfers with a $25 minimum, $250 maximum and at most one transfer per day; Amazon gift-card cashout is marked disabled. The participant FAQ also describes bank transfers as available to all participants. In contrast, the participant terms' disbursement section requires withdrawal to a linked PayPal account and says all disbursement is via PayPal, without a country qualification.
These are attributed official-source statements about withdrawing Connect earnings. The help page's non-US bank-access statements remain internally inconsistent, and its bank-transfer guidance and the participant FAQ conflict with the participant terms' PayPal-only disbursement wording. The reviewed sources do not resolve those discrepancies; no legal or source precedence is chosen, and this does not establish that banks are unavailable. Dollar currency is unspecified, so no USD label is inferred. Once-daily processing is not a bank/PayPal receipt guarantee; fees, conversion and live-account success were not established. Retrieval and displayed update dates do not establish a policy-change date. Payment access does not establish residence eligibility. This correction does not renew other payment, worker-experience or reporting claims.
Connect states a $7.50/hour study minimum and encourages a $10/hour baseline for faster recruitment. The displayed hourly figure uses the researcher's estimated completion time.
The sources use $ without a currency code. This is study compensation, not paid waiting time or a salary. Eligible variable-length screenouts receive partial pay; failed attention checks can be unpaid.
Connect's targeting guide lists participants located in the US, Canada, New Zealand, Australia, Ireland and the UK. Researchers can restrict studies by country, demographics or past participation.
These are the guide's named study-location markets. Account approval and an individual's matching studies remain separate; the list is not worldwide eligibility.
In an older US travel discussion, commenters reported taking studies while visiting other states. One reported a rejection from a state-targeted study; another shared a follow-up support message allowing their domestic trip.
Historical anonymous accounts; dates and quoted support emails are unverified. Confirm current travel or relocation rules directly.
Selected public accounts. Experiences are self-reported.
Study availability differs between participants
Undated
In one Reddit discussion, the author and several commenters described lower recent earnings and fewer worthwhile studies. Others reported a strong month and continued payouts.
One self-selected discussion with unverified identities and exact dates. The reports do not establish typical earnings or explain differences by country.
Stagnaro and colleagues placed both Connect samples among the strongest groups for response validity in a nine-sample comparison.
Author manuscript; data collected June 2022–November 2023. Connect samples contained 1,046 and 1,028 participants. Samples were donated or discounted; authors report no compensation. This evaluates survey responses, not worker pay or current AI-task availability.
Sahara AI’s data-task platform for labeling and evaluating AI data. Rewarded tasks use crypto tokens; community tasks may offer future usage-based payments.