CloudresearchReviews of cloudresearch.com

This was a great way for me to make extra cash to make ends meet. But after almost a year of doing surveys, I had a handful of rejections (not all for good reasons mind you) but I had hundreds and hundreds of successful submissions with thank you notes attached. I logged in one day and I had a 1 star rating, saying that I was reported for AI use. I have never used AI once since starting this!! I emailed them saying this and they said in short "it is what it is, and we wont be communicating any further" also "work hard and rebuild your rating" well since they gave me this 1 star over 2 weeks ago, I have only been offered 1 SURVEY!!!! I used to get tons a day! How am I supposed to rebuild a non legitimate 1 star rating when you are no longer sending me surveys because of this 1 star!!!! I have spoke so highly of this website and company to so many people. Recommended it to friends and family, and then this happens? Honestly screw you! Its hard to put food on the table and this helped! And to lie and say I use AI is just flat out absurd!!!

After submitting hundreds of successful surveys the company inexplicably downgraded my star rating from 4.8 to 1 and restricted access to surveys. Explanation doesn't make sense and is extremely dishonest. Don't trust this company, and don't be fooled by the planted five star reviews amongst the sea of one star reviews.

I've been with CloudResearch for a couple years now. Until recently it was an easy way to earn some extra income. A recent ratings change, however, has penalized long-time, honest participants, myself included, effectively removing us from the work pool. My approval rate through their platform is in the thousands of projects with rejections in the single digits. My performance was rated 5 stars in their prior system. Hardworking participants like me have had severe limitations placed on our accounts, with a reduction in available projects to near-zero, and attempts to reach out and inquire have been met with boilerplate automated responses. When the new rating system was first implemented a few days ago, one of the listed behaviors that would result in penalties was "returned projects." When people pointed out that penalizing withdrawals violates a core principle of academic research, that specific notice was quickly and quietly removed from the page. We have no clue if this penalty is still in effect on the backend or if it was a mistake, because Connect hasn't provided a proper explanation as to how their new system works. This is very shady behavior, and their participants and researchers deserve better. We don't even know what we did wrong, if anything. Unless they reverse course, I can't recommend their services. If you're planning academic surveys or projects through Connect, I would advise you to reconsider and choose another platform. With this move, they've erased years of goodwill and punished those providing valuable, high-quality data.

here is my own formula i made to figure this all out. hope it helps! dont even need to care about their "email" BOOGERMANFORGENESIS Weighted Linear Combination (WLC) ,Weighted Additive Scoring Model (WASM) Microsoft Entra, Google Cloud Identity, AWS IAM classify accounts using anomaly detection, risk scoring, and weighted behavioral models. The Actual Math Behind the Score weighted risk/quality score, also known as a weighted linear model plus normalization. Raw Score Formula Q = w_s·S + w_f·F + w_a·A - w_r·R - w_t·T - w_c·C - w_ai·AI Where: Q = raw quality score S = successful submissions (rate) F = positive feedback score (avg rating) A = profile completeness / “About You” freshness R = rejected submissions (rate) T = time‑out submissions (rate) C = flagged conversations (per 100 submissions) AI = AI‑detected responses (rate) w_x = weight for each factor Normalization to 0–1000 raw score converted into visible score: Q_display = round( ((Q - Q_min) / (Q_max - Q_min)) · 1000 ) 449 is this normalized value, not raw internal score. Positive Factors (Scored as Rates) Successful submissions rate S = successful_submissions / total_submissions Positive feedback score F = average_feedback_rating Profile freshness A = 1 if “About You” updated recently A = 0 if stale or incomplete Positive score Q_positive = w_s·S + w_f·F + w_a·A Negative Factors (Penalties) Rejection rate R = rejected_submissions / total_submissions Time‑out rate T = timeouts / total_submissions Flagged conversations C = flags_per_100_submissions AI‑detected responses AI = ai_detected_responses / total_submissions Negative score Q_negative = w_r·R + w_t·T + w_c·C + w_ai·AI Full Raw Score Q = Q_positive - Q_negative Why Old Perfect Accounts Score Extremely High Example using stats: total = 6162 + 10 = 6172 success rate ≈ 6162 / 6172 ≈ 0.9984 rejection rate ≈ 10 / 6172 ≈ 0.0016 Example weights: w_s = 500 w_r = 800 Positive Q_positive ≈ 500 · 0.9984 ≈ 499.2 Negative Q_negative ≈ 800 · 0.0016 ≈ 1.28 Raw score Q ≈ 499.2 - 1.28 ≈ 497.92 Normalized → near the top end of the scale. why perfect accounts look “too good” to anomaly detectors. Access Level Bands Q_display < 300 → Restricted 300 ≤ Q_display < 600 → Limited 600 ≤ Q_display < 800 → Standard 800 ≤ Q_display ≤ 1000 → Priority 449 sits in the Restricted band. Why a Near‑Perfect Worker Can Still Land at 449 Example current pattern: S = 0.97 R = 0.03 T = 0.02 C = 0.01 AI = 0.02 F = 4.5 A = 1 Positive Q_positive = 500(0.97) + 150(4.5) + 50(1) Q_positive ≈ 1210 Negative Q_negative = 400(0.03) + 300(0.02) + 250(0.01) + 600(0.02) Q_negative ≈ 32.5 Raw Q ≈ 1177.5 If Q_min = 800 and Q_max = 2200: Q_display ≈ 270 Small changes in weights push into the 300–500 band → 449. How to Move the Score Up Negative weights are heavy. Small improvements matter. Examples: Drop AI from 0.02 → 0.00 Gain = 600 · 0.02 = 12 raw points Drop R from 0.03 → 0.01 Gain = 400 · 0.02 = 8 raw points Drop T from 0.02 → 0.01 Gain = 300 · 0.01 = 3 raw points Total gain = 23 raw points → easily +70 to +100 normalized. yours truly boogermanforgenesis!

I have been a participant for years. I had a high rating, only two rejections, and thousands of sucessful surveys completed. Now I am at 1 star and restricted. This is unacceptable. You should not treat your survey takers like this. I am not the only person dealing with this mess. Do better!

Ignore the many 1 star reviews, especially those mentioning “ethics” and “legal issues” due to survey returns. I have over 200 returns, for a multitude of reasons, and have a 5 star preferred rating. Most people refuse to believe they’ve done anything wrong because they’ve been on the platform for so long or are delusional about their own work. They’ve either failed quality checks or aren’t being honest with their rejection counts. This is a great site and has been for many years. I’m still getting consistent surveys because I take my time and ensure quality work.

If I could rate it lower than one star I would. After having been a participant for years they arbitrarily instituted a rating system that has disenfranchised many, many long time participants. Effectively they banned us all by caveat. This site is no longer recommended, and should be avoided in favor of other similar sites.

Cloud Research is not transparent, mistreats participants, and can't really be relied upon or trusted. They recently mentioned punishing participants who return surveys (which violates consent and ethics regulations), then hid the infographic that displayed that information when challenged about the legalities of it. I would suggest other's look for work elsewhere. I have been using Amazon MTurk and Cloud Research for years now with no issues (2 rejections on MTruk due to inexperienced researchers errors, 0 on Cloud Research). I never use a VPN, Only use Dark Reader as a browser extension when doing surveys, never use Ai, always close other tabs and programs when instructed to, always enter full-screen when instructed to, have never had a disagreement with a researcher, have always been honest and accurate in my responses, read everything word-for-word, do longitudinal studies to the end, and provide long open-ended responses. Everything I can think of to be a good-quality participant. I used to have a perfect rating on the site 850, I have over 1500+ approvals (0 rejections), now my account is restricted at 384 overnight with no explanation other than "keep doing good work!". This is very untrustworthy for everyone involved. The participants can do everything correct, put in a lot of effort and time, and be told that they should keep giving good answers despite being told simultaneously that everything they do is bad for no specific reason. The researchers have no way of telling who is and isn't a trusted participant. Some people who get rejections regularly for poor quality work are now rated higher than people with a perfect record of following instructions. Any communication from them is either avoidant, manipulative, or blaming others for issues on the system they have complete control over. They will gaslight, lie, and hide whenever anything goes wrong. Even now while writing this review, I have to double-check there is no identifying information, because I fear the company will retaliate against me for providing an honest review here. You shouldn't have to be literally afraid of a company and hide from them as someone who does work for them. This company is creepy, like an abusive ex that you can't avoid because they are your manager, and they abuse their power and position over you.