Jump to content

vhskid

Member
  • Posts

    89
  • Joined

  • Last visited

Posts posted by vhskid

  1. One more thing...

    While the impressions / clicks drop might be caused by your score change, they might be just as well unrelated.

    7 days is a very short period and slower months concerning a whole industry is nothing unusual. 

    Another cause might be the Fiverr algorithm working on intervals. 
    I noticed that there are weeks when I'm flooded with requests and that's followed by complete inbox inactivity (regardless if the requests lead to orders or not and my score is constant). I'm guessing that the algorithm is dividing the impressions / leads pool among sellers by showing their gigs more for specific periods and giving them chances alternately.

    • Like 6
    • Thanks 1
  2. 9 hours ago, digifel said:

    Success score is only visible to the seller.

    The algorithm that matches gigs with buyers in their search results looks into the individual gig scores and decides what to show. 

    So yes, buyers don't see your Success Score, but how your gigs are rated affects your impressions. 

    In some form, the system worked like that before. The difference now is that we can see how Fiverr's AI and private reviews score us. 

     

    9 hours ago, digifel said:

    How many more orders do I need to complete to get back to 9? 

    We won't know that. 

    • Like 6
    • Up 1
    • Thanks 1
  3. On 3/31/2024 at 5:22 PM, smashradio said:

    The success score data likely isn't considered personally identifiable information since it's essentially a set of numbers calculated by a system to establish ranking and visibility, not specific data collected about you or provided by you regarding yourself.

    It would be like requesting Google to disclose their proprietary algorithm because the search engine indexed your photo and personal website. The internal values used to calculate that sort of thing isn't "personal data", even if it's applied to the data sets they have on you.

    If all it took was a GDPR request to unveil trade secrets, I don't think companies would dare operate within the EU.

     

    AI analyzes our behavior, interactions, and communication style, then assigns data markers rating every little thing. 

    The algorithm puts us in the data brackets and compares, then matches with (also analyzed) buyers in their search results. 

    All these data patterns and segmentation are used for behavioral profiling.

    If there are enough Success Score data markers tied to the seller's profile, then this data stack could be specific enough to be considered as a digital fingerprint. 

    The data from one profile probably wouldn't be enough for cracking the algorithm and reverse engineering, but I imagine that Fiverr would argue under the angle of the occurrence of trade secrets violation. 

    We shouldn't know how exactly the algorithm works, but we should have access to the list of (at least those) negatively flagged interactions / messages. 

    It would be fair and transparent (even if available only behind the SP paywall), but won't happen because things would get ugly, starting from paralyzing Customer Support by appeal / clarification requests.

     

     

    • Like 1
  4. 32 minutes ago, emmaki said:

    Look at the Help Center page for reviews. Emojis. They updated it over the weekend (the topic of my post there).

    I know and I'm following your topic. 

     

    9 minutes ago, emmaki said:

    (I don't buy, so I have no idea what they're doing on the actual reviews)

    I don't as well, I just have an elephant memory (which can be a blessing and a curse).

     

    10 minutes ago, emmaki said:

    Call it amnesia if you want, but I'll go with confusion. I suppose the end result looks the same 🙂 It's also well outside of the "transition period" when all of these tests were supposed to be finished and we were meant to get used to the review system.

    I would say that with the new level / scoring system everything is in a continuous transition period. By design (keeping everyone on their toes, over a cliff, every second of every day) and due to incompetent product release.

    I, myself not getting used to anything. The stars' comeback included.

    • Like 3
  5. 32 minutes ago, emmaki said:

     

    Guys, have you seen this? Not my post, but @rawque_gulia commented that as a buyer, he had seen the return of stars in place of emojis... 

     



    Hmmm, amnesia?

     

    On 3/8/2024 at 7:00 PM, emmaki said:

    I remember when Fiverr tried to convince us that 3 stars wasn't seen as a "negative rating" by the system. 

    I guess that was another little lie. 

    image.png.5e69967d0dba3cc240a34422d2bf277e.png

     

     

    On 3/8/2024 at 6:26 PM, visualstudios said:

    Ok, so they removed the emojis. However:

    The scale is still unbalanced. 

    Very poor - Poor - Average - Very good - Exceptional

    This is not a fair scale.

    Other interesting things of note:

    3 is considered a "low rating" - in other words, average is bad.

    4 stars ("very good") can only have negative qualifiers - you can't rate 4 and praise anything.

    Screenshot 2024-03-08 at 17.24.12.png

    Screenshot 2024-03-08 at 17.21.06.png

    Screenshot 2024-03-08 at 17.21.10.png

     

    • Like 2
  6. 2 hours ago, emmaki said:

    Yeah they sent an excel file with a bunch of boring stuff in it. Just ask CS.

    I'm expecting them to deflect and send incomplete data or some vague summary as other companies do. 

    Sending another one, a more thorough request sometimes does the trick, but specific legal wording addressing specific parts of data is required. 

    Knowing Fiverr it would be a dozen back-and-forth before getting some results and / or a long waiting time. 

    I don't have time and energy for this hide-and-seek but maybe someone with relevant legal experience will forge the path to getting some insights into gig scores / metrics data. 

     

     

    • Like 3
  7. 49 minutes ago, priyank_mod said:

    Since we have a rolling time window of 2 years - may be a bad review/cancellation from Feb-March'22 went out of the consideration set and is not influencing the score that strongly anymore. 

     

    Again - the 2-year window concerns the Rating score (stars), not the Sucess Score:

     

    image.png.4fdcc7929f64f69271436529b61843df.png

    So the 2-year-old public and private reviews go into ️ Rating.
    We were never informed about the time scope for the Success Score.

    I think that there is a very widely spread misinformation and / or misinterpretation among the sellers in this forum regarding this. 

    This score, that score. This rating, that rating.

     

     

    • Like 6
    • Up 2
  8. 1 hour ago, charlsmcfarlane said:

    If it’s taking into account a specific time frame, for example the last 24 months, then changes can happen at any time. Let’s say you had a negative private review 24 months ago. As soon as you hit 24 months and one day, that negative review isn’t impacting you any more, so your score would seem to inexplicably increase.

    I have no idea if this is how it works, I’m just guessing at what could cause movement like this with no orders.

    The 2-year window concerns the Rating score (stars), not the Sucess Score. 

    But yes, we'll never know how this is all interconnected in the black box.

    For my SS change, I'm betting either an effect of the scores being dependent on other sellers' performance or "fine-tuning" the algo by Fiverr's product team.

     

     

    • Like 3
    • Up 1
  9. 17 minutes ago, donnovan86 said:

    Clearly a sign that things are being changed in the background. I didn't have any SS change, however for one of my gigs the client satisfaction went from negative impact to severe negative impact. And that gig had 2 orders the past 6 months. Last one was reviewed before the new level system and success score. So I have no idea what brought that change :))

     

    I would assume that this is an effect of the scores being dependent on other sellers' performance but it might be "fine-tuning" by Fiverr's product team as well.

     

    tumblr_pjhfnkLcFM1qgzs6bo1_500.gif

    • Like 3
    • Up 1
  10. 58 minutes ago, katherinasim said:

    Does anyone see any changes in your Success Score Stats? In March I have completed 7 orders, 6 of them with 5 star rating (1 autocompleted), 2 of them with the "Fiverr's choice" badge, and still this gig has a score of "6" and strong negative in "client satisfaction" 😅 I have no idea how to improve it

    My SS went 1 up recently, but I haven't had any new orders / reviews since January.

    • Like 3
    • Confused 1
  11. 9 hours ago, whildebrand said:

    f) There is even a clause in Fiverr's Terms of Service preventing you from advertising your freelance services on Google Ads for any similar services you offer on Fiverr. So you may not "freelance" outside of Fiverr in the proper sense of the term, similar to what any employer would demand of you. You can look it up in Fiverr's Terms of Service.

     

    Can show that point in TOS? 

     

    I think you misread the statement below that prohibits using Google Ads for promoting URLs to your Gigs and Fiverr profile...

     

    image.png.5a5bcd0a36af20af0d857352a84c866c.png

     

    It's not a non-compete clause. You can freelance outside of Fiverr.

    • Like 2
    • Up 2
  12. 1 hour ago, donnovan86 said:

    Fiverr uses AI to try and avoid hiring people for these tasks. And I doubt they even care who is an agency and who's not. Clearly as long as an account makes money for them, it's fine 🙂

    I can tell you for a fact that there are writers who on YouTube and even on their LinkedIn say they outsource, some even have their own outsourcing business you can see clearly.. yet on Fiverr they say "I will write", I will do, etc. And obviously you are compared to those because they are a seller just like you.

    I think that Fiverr is very invested in good buyer-seller matchmaking (subjective, yes) and making deals. Also, the "hidden" agency accounts might skew the algorithm too much, even for Fiverr. 

    Significantly above-the-average delivery frequency and income size indicators (for the category / industry) could be used for automatic or semi-automatic detection of these agencies.

     

    1 hour ago, donnovan86 said:

    I know Fiverr has the Agency feature coming soon

    I'm talking about 2 different things, but the 2nd could only happen if the 1st works:

    • Internal detection of these agencies posing as freelancers and internally treating them as agencies by scoring and gig-serving algorithms
    • Official Agency feature and forcing the "posers" to convert accounts

    The Agency feature has been present for a few years now and it's re-released again or refreshed(?). I don't know what they are changing with this. 

     

    1 hour ago, donnovan86 said:

    do you really think these people will admit they outsource and apply for an agency spot?

    I think that Fiverr doesn't care if these people want to admit that or not, and can force them to convert their accounts, so the agency's searching buyers have a better selection and less confusion.

    The servers constantly grinding the algorithms could also generate less cost.

     

     

    • Like 2
    • Up 3
  13. 1 hour ago, donnovan86 said:

    Obviously everyone is treated as an "account"

    What is obvious here? Regular accounts led by agencies could have been detected and marked internally as agencies and treated as such when it comes to comparing with other accounts.

     

    1 hour ago, donnovan86 said:

    But it's clear that they compare accounts between each other, and if you're an agency or you outsource your work, obviously you obliterate the competition when it comes to volume.

    I have the same concern, but at the end of the day, we don't really know what exactly the sellers' comparison means.

     

    This is a transcript fragment from one of the webinars:
     

    Quote

    And when we're referring, about, to, performance, we're talking about our satisfaction, responsiveness, cancellation rate, on time delivery, and also now, the way you perform compared to your, competition. And like Mikey mentioned, the competition is not all of the sellers that are operating in the marketplace. We segment you according to the people that are similar to you to make it more fair, and more fair competition.

     

    So this "fair competition" would be to flag these "hidden" agencies (at least internally) and not compare them with regular freelancers. 

    But we can't really tell what's going on there.

     

     

    • Like 3
    • Up 1
  14. 16 minutes ago, williambryan392 said:

    And you don't know my level status either or my success score- you don't know where I stand. You're making an assumption based on my views of the new system.

    So is your TRS in actual danger because of the new levels?

     

    16 minutes ago, williambryan392 said:

    Yes they might, or they might not. That's been something I've constantly said. The fear doesn't help. We have to wait and see what the actual impact is.

    It's easy to say "We shouldn't give into fear".

    People have different life situations, different stresses that accumulate, different mental health states. 

     

     

    • Like 4
  15. 15 minutes ago, williambryan392 said:
    35 minutes ago, vhskid said:

    I'm saying that you'd probably be more critical about the new system if it would hurt your business or if it looked like it would be hurt.

    Who on earth said it hadn't!? Massive assumption on your part. If you want the facts. February was my lowest revenue month for over 2 years, about 75% lower than my average month.

    You're just assuming that everything is fine with me because I'm not slamming the new system.

    The arguments can go both ways. Successful sellers can say they don't like it, and sellers that are getting hammered by it can say they do. That's allowed...

    I meant your level / TRS status here. Since yes, I don't know your revenue stats. 

    While revenue drop might depend on different things - not only changes introduced by Fiverr - the more or less drastic levels' demotions are big indicators that things might really go bad for someone during this year.

    • Like 3
  16. 22 minutes ago, williambryan392 said:
    29 minutes ago, vhskid said:

    Calm down, Karen.

    Is that really necessary/constructive/kind?

     

    1 hour ago, williambryan392 said:

    I find that quite obtuse.

    Was this necessary/constructive/kind?

     

    23 minutes ago, williambryan392 said:
    30 minutes ago, vhskid said:

    who sees the system's faults with a broader, deeper, and more careful perspective.

    suggests by definition that I do not, that I am in fact narrower, shallower, and have a more careless perspective.

    I don't think this is fair [...]

    Agree to disagree.

     

    25 minutes ago, williambryan392 said:

    [...] or true.

    There is no objective true / false when it comes to subjective images of ourselves.

    • Like 3
  17. 2 minutes ago, donnovan86 said:

    Then we have you as a seller being compared to other sellers and their performance, even if that "seller" is in fact an agency. How can I compete with another writer that outsources 30 articles a day, and I write everything myself? Come on..

    I wonder about it myself - if these profiles are internally marked as agencies and not compared to regular freelancers or not.

    The distinction should be obligatory and public. There is the "Fiverr agencies" account option, so these not-so-freelancer profiles should be converted.

    • Like 2
    • Up 5
  18. 1 hour ago, levinewman said:
    2 hours ago, williambryan392 said:

    Private reviews were never a secret. This information has always been available in the help center and also talked about regularly on the forum. Also if you ever purchased something on Fiverr you saw it when reviewing.

    Private reviews always impacted a sellers position (or 'rank' as some people insist on calling it) on the Fiverr marketplace. Sellers were also always compared to other sellers.

    This data has always been used by Fiverr to decide where to position you. The key difference now is that you see more of the data.

    I understand why people are annoyed by the changes and  annoyed by the communication of the changes, but to be surprised that your positioning is relative to other sellers and that your performance is compared to other sellers to decide that positioning should not be surprising.

     

    Expand  

    This analogy never works and let me tell you why. 

    If you have a spouse and they tell you every day you're amazing, but they secretly tell their therapist you're a total piece of junk as a spouse, you aren't seeing the full picture. 

    Here's the rub: neither of them ever tell you that you are doing things wrong.

    But now, that therapist quickly reveals you've been messing up the entire time and you have 30 days to fix it all before you're forced to get divorced. 

    That. That is how this entire thing worked. 

    Fiverr didn't ever fix the existing problems until they said, "Oh yeah, I'm leaving you, you have 30 days to fix a problem that's unfixable in 30 days. YAY!" 

    Stop carrying water for a poorly designed system. You're carrying water for Fiverr when it makes no logical sense to have this system designed in this specific way, and then expect people to completely change their abilities within such a tight window. 

    It's simply bad implementation, bad management, bad communication, and bad leadership. 

     

    It's logical (disappointing but not surprising) that sellers who weren't negatively affected by the new system (e.g. still sit on their intact TRS status), won't criticize it as much as people whose profiles / gigs were hurt. 

    But it's nice to see "unaffected TRS" @levinewman (at least status-wise) who sees the system's faults with a broader, deeper, and more careful perspective.

     

     

    • Like 5
    • Up 1
×
×
  • Create New...