Hey Upwork Management,
Users have been complaining of a ton of spam posts recently and no solution has been offered except for transferring the burden to freelancers to flag these job posts. As a stop gap solution until you guys figure this out, our agency is offering to do the work of flagging for you at a reasonable rate. This seems to be only way to improve user experience because nothing has been done for quite some time now. So if anyone from your team wishes to reach out, send me a message.
They can even develop a simple system that reads the post description, and they can identify scams and spam, and fradulate activities, i.e: if the client posts that says "contact me on WhatsApp up using: **Edited for Community Guidelines**".
this will work because i know, unless they didn't want to do it
Whatever Upwork is doing is not working because the number of scam and spam posts still remain at an increible level. If it was working, they would not be remaining at an incredible level.
Peter - I agree, not working effectively as it should. If these spam/scam posts are as Nikola suggests: "constantly changing methods of scam accounts" then I sense the spam filtering algorithms are either not being updated if real-time (likely as this is expensive) or the settings are too liberal: allowing more spam/scam in at risk of deleting legitimate posts.
Remember: the Clients have the $, the freelancers dont. This plays a huge role in how companies filter spam. For example: banks constantly measure and re-measure the costs of loose vs tight fraud filters by analyzing the potential losses from too much fraud (false negatives) vs the potential losses from lost-customers who appear fraudelent but aren't. (false positives). Only UW insidres know the true reason for the current trend -- we never will see anything other than a PR-centric response like that above.
Understood. But what gets me is that the EXACT same scam posts appear over and over. At least filter them enough to make the scammers have to change them up a bit to evade the filters. Most of them were blocked before last summer, so I don't knw why they stopped being blocked.
Thank you for reaching out to us and sharing your feedback and concerns. Our teams are working to address the situation around scam and spam jobs, including higher than usual volumes of scam posts, constantly changing methods of scam accounts and gaps in user education around Upwork TOS and safety.
Additionally, I assure you that we'll continue sharing users' feedback with the team and providing updates to the Community about team's initiatives.
Feel free to check an update in this post for more information about the team's efforts around addressing these issues.
I cant see your profile to know for certain, but UW uses Data Science methods for spam filtering. My sense is that UW uses Natural Language Processing (NLP) models to identify spam - and have likely trained these models with proprietary data. These are trick because strict settings can eliminate legitimate posts which sound spammy but arent. The company uses technology - these NLP spam filters, apparently not very well - not humans to do this work for them.
Yes, freelancers' time is money, just like Upwork emplyoees who get paid to manage the site, unlike us who merely use the site. Their JOB is to do that, not ours.