Testing improvements to the rating & review system

[catwriter]

7 minutes ago, emmaki said:

So anyway, dashboard score = your actual rolling feedback score and profile score = overall score. I guess? 

I hope, or I might lose a level... 

Oh, and I just noticed your new gig images. Is that new Pro/TRS fashion now? Advice from your Success Manager?

[emmaki]

1 minute ago, catwriter said:

I hope, or I might lose a level... 

Oh, and I just noticed your new gig images. Is that new Pro/TRS fashion now? Advice from your Success Manager?

Well, my profile picture had to be changed because it wasn't professional enough.

My gig images had to be changed because they weren't professional enough.

My profile description had to be changed because it wasn't professional enough.

My tagline had to be changed because it wasn't professional enough.

I changed my gig descriptions on my own though. 

After all this super-professionalization, Fiverr kindly announced that Pro was opening to more people. I can point to at least one of these Pros as a copycat, simply because a) they once hired me to review their gig and b) they used my (old style) gig images on their gigs. They're a Pro now. With my old-style gig images.

Go figure. Just one of many reasons I am certain that the new Pro was simply a downgrade of the Pro-gram to fill up a Pro marketplace. 

My profile is now an anemic, professional-looking thing with low-effort but "professional" looking gig images. I did take the time to find an actual template that looked good everywhere without being cut off/blurred/whatever.

It won't surprise anyone reading this that none of Fiverr's recommended template sizes work. I forget the dimensions I eventually used - which still had to use a transparent border to represent zones where Fiverr's systems would cut off faces/text/whatever. 

Hope you don't lose a level, cats! Who knows, though - the whole platform is a mess at the moment, but at least its hitting all the right notes on the stock market for now. Which is really the bottom line and I suspect the biggest problem at Fiverr. 
 

[bashir_expert1]

On 12/6/2023 at 11:59 PM, visualstudios said:

I find that the labels (and emojis) on the rating system are not balanced. While "exceptional" is better than the previous "perfect" (as nothing is "perfect"), it's not a linear progression.

To be consistent, it should go:

Very poor -> poor -> average -> good -> very good.

Why is one step down from average "poor", but one step above average "very good"? That system seems designed to make lower grades more frequent. Also it's strange that the negatives are "poor" and "very poor", while the positives are "good" and "exceptional". The language could be more neutral using the suggestion I provided above. "Very good" should be enough for the top end, if a buyer is truly amazed, they can always leave a tip for exceptional service (going "above and beyond").

That's a valid point.

[ahesanurasun1]

All information is helpful. 

[catwriter]

45 minutes ago, emmaki said:

Hope you don't lose a level, cats!

Thanks! Wouldn't be the first time... And with how those monthly evaluations work (plus all the other changes to the system), levels became a joke I can't take seriously. Not to mention that this year was the worst when it comes to finances since I've joined the platform, so losing a level would be the least of my problems.

 

49 minutes ago, emmaki said:

I can point to at least one of these Pros as a copycat, simply because a) they once hired me to review their gig and b) they used my (old style) gig images on their gigs. They're a Pro now. With my old-style gig images.

Go figure. Just one of many reasons I am certain that the new Pro was simply a downgrade of the Pro-gram to fill up a Pro marketplace. 

At least that one copied from a Pro... There's a Pro seller who copied a gig description from me, and changed it just enough to avoid word-for-word plagiarism. They later changed it again, no idea if they decided to copy someone else or came up with something of their own.

[edc_lab]

Good morning @emmaki

  

13 hours ago, emmaki said:

If you want to go down the route of "you don't know me" then I can only agree, but point out that I know two things about you now: you use AI - no big deal - and that you deny using it except for a very mild use case.

Sorry, but you're fighting for your ideals with the wrong person, you're just talking with a guy that fixed his grammar with ChatGPT for a forum post and now you're talking again about people that scamming by producing and selling content in your working field (which is bad, of course).

 

13 hours ago, emmaki said:

Now consider how using AI might impact people's decision to order from you. If AI can do it for "free" - or more importantly, they think they can get it done just as well for free - why should they pay you $5, $10, $20? 

I perfectly understand that, cause other than UI/UX I also do illustrations and today everyone could do good things just by using Bing. But you cannot compare my "grammar behavior" to pepole using AI for saving money. I'm aware of the bad side of AI and how much it can impact our services, but the message cannot be "you're as bad as those people not hiring freelancers to save money through AI cause you're using AI too", it's not the same thing.
In my case I simply wanted to improve my communication on a very delicate topic for feedback, that's it.
As you can see, I'm not using any AI here cause the communication is more informal, but for the incriminate message I decide to have a different tone, nothing more.

 

13 hours ago, emmaki said:

I use ChatGPT and Perplexity all the time and I do not resent AI - I'm just tired of people abusing it and, all too often, denying it when caught red-handed. 

Caught red-handed, what? I'm going to repeat it for the 10th time, but please read it carefully otherwise I need to use AI again, ok? 
I wrote the original message in English and I checked it with AI for fixing grammar mistakes.
Here's an example: I wrote "for release" and the AI fixed it with "rollout phase" which is definitely better (at least in my opinion).
In that case I learned a new thing, but what should I do next time? Can I use that term or not? Is that too GPTese?
The core concept is mine and if you find that post too professional, well, simply it isn't, I literally wrote basic UX concepts.
Sorry, but it's funny reading that you can use AI tools for your scopes, but in my case, I can't use these tools to fix my grammar on a forum post.

 

13 hours ago, emmaki said:

Are you so quick to jump to the "I'm not a (pseudo)-scammer like them" defence now?    

I'm not jumping over anything, you literally said "you're in good company (with these pseudo-scammers)" so I was just making clear the part that bothered me, nothing else.

 

13 hours ago, emmaki said:

It doesn't matter if your English isn't perfect. Believe me, I live abroad and I end up using translation tools all the time in situations where I have no idea what to say and I murder the (ridiculous) grammar on a daily basis.

It's the same feeling for me, but the problem is that I hardly accept making silly mistakes.
The fact you live abroad should be a plus to understand that. There are different cultures all over the world and in my country, Italy, native people messing with the Italian grammar aren't considered very educated and that's why I care about it, I'm influenced by that even when I switch to English.
I know that I will never speak it like a native, but I like English, and for me, making an effort with it it's also a form of respect for the ones that are reading or listening to me.
But, yes, I'm going to work on that aspect and I'll try to be more relaxed on the grammar side.

 

13 hours ago, emmaki said:

(...) a better use of ChatGPT to help you articulate your thoughts would be to get GPT4 and ask it to speak to you in English and practice conversations. It will correct you as you go along. 

(Just use Grammarly basic for correcting your spelling and grammar). 

Thanks for these suggestions, that's the type of comment that I like to read.

 

I hope to have addressed all the aspects, but please, don't compare me again to these fraudulent people which I don't like at all. It's the opposite of what I do, that is offering custom and unique solutions based on real problems, data or feedback.

Have a nice day, bye! 

[edc_lab]

15 hours ago, coerdelion said:

Your technical English is well above average.  Problem is, most people don't speak technical and *really* don't want to think much while they're reading. 

Reading age in most places other than technical / academic sites or papers should be around 14-15 years.  Makes it easy to read and understand, while passing the AI test. 

How do I know?  I put one or two of my essays from university through an AI detector. Was highly amused when they came back as AI generated. 

I graduated in 2008, well before AI was available.  And don't have a time machine

 

Thank you so much for your kind words, it really means a lot for me! 

[edc_lab]

Hello @catwriter

15 hours ago, catwriter said:

As long as your English is understandable, it doesn't have to be perfect. 🙂 

And if you stay on the forum, you will see a plague of AI generated posts that bring nothing of value, and you'll (hopefully) understand the frustration some of us feel.

That's probably the point and why I triggered the discussion, for some of you it was a kind of "here we go again" moment.

When it comes to informal tone I'm fine with it, but for that post I just wanted to give a more professioanl tone and I ended up using too fancy words, my bad!
Thanks for your message too! 

[emmaki]

I wasn't comparing you to anyone dude. I was simply pointing out that there is a lot of misuse of AI everywhere and giving examples of it. 

My entire point was only that you were presenting your own perspective with AI text without disclosing that you had used it. The rest is just window dressing. You already understand the problem on the forum, it seems. 

Your example (for release > rollout phase) is just an example of ChatGPT injecting its own stuff into your writing. That's not fixing a spelling or grammatical mistake. It's very bad at that. Here are some of ChatGPT's biggest uses complaining about how the company is slowly being Microsoft'd since the whole Sam Altman debacle: 

https://community.openai.com/t/gpt-4-is-getting-worse-and-worse-every-single-update/508470

How this squares with Fiverr's drive to get a lot of users using AI to deliver high quality work is another thing. This is only going to get worse, IMO as more and more people start using GPT to do a lot of things for them. 

Even OpenAI models are affected by this, which means that many of the AI tools out there (which run on OpenAI APIs) are going to be impacted, if they are not already. Many of those tools are simply nothing more than a huge rip off, so far as as the credit/$ conversion goes. 

You said you didn't want to write a professional post - but that's exactly what ChatGPT understood and delivered. You went along with it because it "sounded better". Do you see the problem now? Just use Grammarly. It's designed to correct your grammar, not reinvent what you're trying to say.

If you must use ChatGPT to improve your language in business, just remember that completely incompetent people and frauds do the same (as do multimillion/billion-dollar companies). Fiverr has a big problem with sellers in this mold. And each and every one of them holds the same idea you do: "this sounds better than what I wrote". But for anyone who is versed in AI-speak, it's a red flag. We're already stuck in the echo-chamber of algorithmic displays ("only show stuff they react to strongly") - give it time, but soon that'll be joined by AI perspectives dressed up as human ones. Which they kind of are, since copyright continues to be an issue. 

You may resent the implication, but you are doing it to yourself - and, unintentionally being a part creating a future that I don't think anyone particularly looks forward to. Avoiding the implication is as simple as not doing it. The end. I am essentially saying what everyone else is saying that you like to hear - just taking a different and far more direct approach and taking the time to explain why I think like I do. It has nothing to do with cultures or languages, and everything to do with complex processes of a global industrial revolution driven by forces we barely understand. 

38 minutes ago, edc_lab said:

"you're as bad as those people not hiring freelancers to save money through AI cause you're using AI too"

Not my message at all. My message is simply "don't use AI to represent your own thoughts". I mean, you're not wrong, but it's the bottom end of the market that loses to AI, because the cheap clients are the ones who use AI. Large SEO agencies don't hire volume SEO writers anymore: they hire cheap people who have learned how to VA and use SEO tools to generate the volume writing. The cheap SEO volume writers who are suffering are those who have not adapted. Most of these people have pretty bad English and very large egos due to their success. Even so, they are succeeding because they don't hide any of this and they are offering something that people want (saving time on boring optimized articles to hit SERPs, in this case). 

There's a reason forums are now on the top of search results when they hardly used to be. People are looking for human perspectives. But now you have people misusing AI on forums following strategies like parasite SEO, which (badly-done) is more or less "let's get ChatGPT to write a verbose comment and slap a link in there".

Do you see what I am getting at? You are not explicitly the problem, since you are simply trying to spellcheck your writing or whatever, but you are a part of a growing problem, whether you realize this or not.

Once again, the issue is that you presented something clearly written by AI as your own perspective. That's it. The rest is window dressing to explain my perspective. We can simply agree to disagree. 

P.S. When I offered a UI/UX gig, my approach ("tell you how much your website sucks") was to slam anything I didn't like and go into agonizingly nitpicky detail about just why I didn't like it. People did pay me to essentially tell them how awful their website was without being particularly polite about it. Offering custom and unique solutions - that someone else could do - was a gig extra.  As was added swearing. Not really a "professional" gig though, so I dumped it. 

[visualstudios]

6 hours ago, emmaki said:

It won't surprise anyone reading this that none of Fiverr's recommended template sizes work. I forget the dimensions I eventually used - which still had to use a transparent border to represent zones where Fiverr's systems would cut off faces/text/whatever. 

I created my own template, after extensive testing. The aspect ratio and resolution for fiverr thumbnails is non standard, so they'll always blur a bit, but at least I can fit everything in frame. There's no way to make the video thumbnail be crisp, it distorts it on upload anyway.

[visualstudios]

6 hours ago, emmaki said:

but at least its hitting all the right notes on the stock market for now.

Is it though? The performance has been... ok at best. It fell 90% from the top, which ok, it was covid mania and stimulus money. But the recuperation has been laggy, it's down YTD, in a year where S&P jumped around 20%. That's just bad. I wouldn't call that "hitting all the right notes on the stock market". 

[emmaki]

So, I'm not a stock market person, I just go by whatever the few articles I read say. Which is not the best way to judge, I get it (since most of them were barely-edited Fiverr PRs). Still, the overall impression I got was that Q3 was good. IDK about the year and the rest of the stock market. 

The Fiverr pandemic COVID WFH mania was an anomaly, for sure. I wouldn't be using it to judge anything other than a crazy WFH bubble that popped as soon as the Authorities™ realized that getting people back into the economic productivity of the commute, renting office space, and buying lunch at a chain place was a greater need than saving the world from the Toxic Coughing Miasma.

So now quite a lot of those bored office workers have mostly left Fiverr back to the joys of their office because otherwise, they're being economically irresponsible. We're just back to normal Fiverr: a place with some great talent, a lot of average talent, and an unknown amount of... people, let's call them, who are using tools or other people to do all the work, with the full blessing of a platform that is deliberately targeting more upscale clients as well as new target markets. I know I sound like a broken record on this point, but it doesn't square up. 

So, with all that (opinion) in mind, wouldn't it be better to compare performance against whatever Fiverr's stock market performance was just before Deadly Toxic Miasma? Fiverr's problems are the same as they were before, just now with added AI and an increased focus on making more money (the real root of my notion that Fiverr is hitting all the right notes - as far as I can see, that take just keeps going up...). Here is my expert analysis of the situation: 

As you can see, Toxic Miasma bubble aside, Fiverr's plodding along. Which I would say is pretty good for a company whose business model is directly threatened by the emergence of AI. 

Also, I'd like to just casually toss the war in the Ukraine into this analysis, since that's still a popular thing to do. Isn't it?

[catwriter]

34 minutes ago, emmaki said:

Also, I'd like to just casually toss the war in the Ukraine into this analysis, since that's still a popular thing to do. Isn't it?

Don't forget Israel and Gaza...

[emmaki]

4 minutes ago, catwriter said:

Don't forget Israel and Gaza...

I feel that's probably a subject best avoided here, tbh, although it is doubtless affecting staff in HQ (as well as others affected by the ongoing horrors). The CEO's Twitter - sorry, "X" - is quite interesting. And that's all I have to say about that....

[leannelrivers]

On 12/19/2023 at 1:54 PM, vibronx said:

I'm suddenly very worried about the order I have in queue from a new buyer.

It's a great way to scare people into paying $39 a month for the privilege of using the Request to Order feature. 

[donnovan86]

31 minutes ago, leannelrivers said:

It's a great way to scare people into paying $39 a month for the privilege of using the Request to Order feature. 

Even then, you can't possibly know if a vetted buyer might enjoy the result or leave your expected number of stars. But I will agree, request to order does sound very good, but it depends on the niche. For writers in particular, it's not alwasy a great option. It does go well for those that take fewer, yet larger orders. I have request to order, but I only used it on one gig and it did scare off a lot of bad buyers, but also pretty much everyone else as well. So at least from my experience, it depends on the niche and what you offer. Because in some cases, request to order might do more harm than good. Yet as it was said, you need to find a way to stop those unwanted buyers and this might help.

[visualstudios]

1 hour ago, emmaki said:

As you can see, Toxic Miasma bubble aside, Fiverr's plodding along.

Meh.

 

[reflection2imag]

Great, very Informative. Thanks

[ebookandarticle]

On 12/14/2023 at 3:50 AM, pixel_editor197 said:

I am disappointed with this Fiverr update review system. 

Recently 2,3 clients submitted feedback with good comments but I received 2.7 and 3.7 ratings. 
When I asked from clients about my work they were amazed and told me they submitted 5 5-star rating, but unfortunately, it's just the case with the new Fiverr review system. 

When I asked from client, if you thought it was wrong or mistakenly submitted this rating you can update it by contacting the support team. My client also asked how they can update the rating I told him there is only one way to contact the support team. 

When the client asked from support team about an update rating they just sent a warning of TOS. I read Fiverr's terms and conditions and I know it's against Fiverr's policy to ask for a good rating and change for rating but in my case client wants to know how I can update it. I just informed the process of how you update it. I also told him rating is really important to me and very helpful if you could update it. 



I'm still afraid about asking and sharing this information on how Fiverr react to this review. 

I think you should tell yourself fiverr support about it, rather directly tell your client. The reason of warning behind on that! New update needs time to adjust, hopefully Fiverr will fix all issues soon, if their have multiple issues for new updates! patience will be rewarded in that case! I think 

[edc_lab]

Hello Forum, hello @emmaki

I decided to browse this forum for an in-depth check and yes, I got what you mean. Yes, that kind of use is actually an abuse of AI, it's also pretty easy to spot it without the use of any detector tool, even for me. I agree, if you go in a forum, you should read real thoughts, but I have agreed on that since the beginning.

As you can see I'm not using AI to discuss with you, because I think I can decently deliver my messages (well, I hope).
But as I said earlier, for this specific topic, for that specific post, I thought to produce a well-written message because I imagined it could be used by Fiverr team as a feedback for improving their new rating system.
But I didn't ask ChatGPT to write for me some UX Design concepts, but I simply used ChatGPT as Grammarly.
Well, I think Grammarly is more user-friendly in that sense, so I just made the review harder by using ChatGPT as Grammarly.
Indeed, my entire message was written in English, then I asked to check spelling and verb tenses without touching the other words. After that, I reviewed my text and I edited all "weak words" and the repeated words with synonymous. You can't even imagine how many times I edited these sentences, but actually, you know better than me how many nuances a word can have, and I spent a lot of time researching the best terms.
As you can see each concept is formed by two or three small sentences, nothing too complex apparently, but it took a lot of time for me to form the structure in that way, the idea was to make a "professional" grammar structure with a "non-professional"/in-depth analysis (basic UX concepts).
I didn't ask "rephrase this sentence professionaly" and then I said "wow it looks better, let's use it", this would have been a terrible approach.
For instance, do you know why I spent so much time on alternatives such as the "for release" part?
Because it didn't sound exactly what I wanted to express and the AI suggested "roll out". I checked the term to know the technical difference between release and roll out and then I said "that's the word I was looking for!". That's what I did for the entire message.
I decided to spend considerable time on that feedback because I wanted to offer a simple but efficient one,  I saw that post like a form to be compiled, but I don't know how to explain it better.
I'm not a lazy person who asks AI, then copy and paste content, you can see by yourself the commitment I have even on these informal lengthy posts, I'm not scared to write. But yes, probably an aspect that I need to work on is to be more flexible on grammar mistakes and accept them more easily.

About sites like Copyleaks: I checked my post and it says that also "have a nice day and see you soon" was written with AI so I checked some reviews on the Chrome Web Store and the highlighted one was a one-star review of a guy saying that he wrote a text on his own and the tool gave him a score of 84% written by AI.
So, as of today, it looks like even the best tools can spot "false positives".
To be honest, it saddens me a bit thinking that if I put that effort on a text then I sound like ChatGPT, but I will do my best to improve on that, that's for sure.

But let's move on! I hope everything is clear now and I wish to enjoy the next forum's discussions with all of you guys.

See you soon! 

[beendi]

Love it . Thank you for this 

[vibronx]

14 minutes ago, beendi said:

Love it . Thank you for this 

What do you love about it?

[catwriter]

2 hours ago, edc_lab said:

I imagined it could be used by Fiverr team as a feedback for improving their new rating system.

I wouldn't hold my breath.

[emmaki]

Fiverr's not big on feedback, I'm afraid, Ed. Kesha (Fiverr staff member) stopped the pretence of "passing on messages" (I always picture a virtual trash bin slowly filling up) and liking a scattering of posts on this thread quite a while ago. Admittedly, the last few pages are not exactly on topic except for some people who like everything they see, filled as they are with a Panglossian joy at the world around them. 

At best, someone will from Fiverr come back and say "we listened!" and show a couple of changes like different emojis and maybe a word change. Or maybe they won't. Or maybe they say they will, but then won't.

It's probable that this is going ahead, but the real feedback will come from the metrics and whatever Fiverr specifically wants to see as a result of this test. Which is less 5 star reviews, apparently. The only thing that I can see is really worth scrutiny is whether buyers are actually getting their reviews changed at the last minute, but that will be shrouded in official mystery until the end of time (which may be soon, so maybe we won't have to wait so long after all). 

And stop fretting over your English, it's fine! Better than ChatGPT. It's you. That's the whole point - and you seem to have good character, unlike so many (others) who abuse ChatGPT. 

 

[emmaki]

I just saw this video on TikTok. It reminded me of this post. It's comparing how startups implement new things v big tech. It may or may not be interesting, depending on your stomach for such things. No dancing, I'm afraid.