How do you calculate a workers success rating?

[9608a775]

I'm interviewing someone for a job I posted. She has a 93% success rate. How does upwork calculate the success rate? I would rather only hire people with a 100% success rate. Why would someone have less than 100%?

Also, some people on the platform who have done lots of work using upwork have nothing listed for their success rate. Why is that? Shouldn't everyone who's done work on the system have a success rate?

Suggestions: It's very difficult to understand your success rating system. I'd rather see a five star review system so that I can quickly see if clients like a worker or not. People should be able to post written reviews and click on one to five stars. It makes me nervous hiring someone from you site when I can't figure out if they're good or not based on what others say. Your success system doesn't explain what 'success' means or how someone gets the 'success percentage' they get.

[versailles]

Please read this: Your Job Success Score on Upwork

[sam-sly]

I think the Job Success Rating is a good comparison metric, but it isn't perfect. If I were hiring, I probably wouldn't rule someone out with a 93% JSS but I would be very cautious about someone with a low JSS.  Here's why...

If you read the link Rene sent you will see some of the JSS factors are directly related to client satisfaction, but some are not as directly related. Known factors include:

- Public feedback from clients

- Private feedback from clients (basically why the job ended and a scale of 1-to-10 whether the client would recommend the freelancer)

- Long-term contracts (I believe active 3 months or longer)

- No feedback contract closures (if a "pattern" of clients doesn't leave feedback the freelancer's score can be dinged. Upwork understandably won't say what constitutes a "pattern" but this can result in less than 100% JSS with happy clients.)

- Inactive open contracts count negatively after a while

The last two factors are a little iffy. They can indicate a negative client experience, but not necessarily. Sometimes clients leave the platform after they approve and pay for the work. If the freelancer leaves the inactive contract open it can decrease the score, but if the freelancer closes the contract he/she risks a score decrease from lack of feedback.  I don't think this would result in a low score, but it might drop a 100% score lower into the 90's depending on how many contracts that freelancer completed.

The public feedback is easier to manipulate than the JSS. Some freelancers use to harass clients over a 4-star public rating so some experienced clients started giving 5/5 to avoid the hassle. My understanding is that is why Upwork introduced the JSS. Also, a freelancer could give a 100% refund to get the negative feedback off their profile. With JSS, the private feedback is still factored into the score. This is why I would be wary of freelancers with all 5/5's and a low JSS (like in the 50's - 70's). 

[abbeybrown]

As someone with a 93% job success score, I completely understand clients would prefer a 100% job success score, but there's plenty of reasons why someone would have a higher job success score than me without being any better. The success score is calculated from three different time frames: six months, a year, and two years. If someone only had one job per month for the six month frame, it would be relatively easy for them to get a 100% score. If they had twelve jobs within the six month time frame, there's a lot higher likelihood that something could go wrong--incompatible client/freelancer, a client that tries to bully a freelancer into doing what they want by threatening a bad score, the creativity of the freelancer and the desires of the client don't match, the client is extremely picky, etc.. 

When I look at potential clients, I know I can't rely on the star system too much, so I read the reviews and I would think this method would work for clients looking for freelancers too. If clients aren't gushing over the freelancer, they probably weren't great. 

[wendy_writes]

Michael G,

As a freelancer writer for a good number of years and as someone who has both directly hired or hired in conjunction with project owners (who I write for) artists, website developers, voice and music talent, animators, video producers, etc. the BEST way to find trustworthy freelancers is to:

1. Scan profiles

2. Review portfolios

3. Skype chat (video not needed) with a few of your 'final contenders' for a job.  This gives both of you insight into not only what the job parameters and expectations are; more importantly, it gives you both insight into personalities, work ethics, etc.  

- This is the optimum way to hold what used to be a face-2-face interview / meet and greet.  Both parties can usually determine if the 'people fit' is good.

- Hiring someone who you feel confident in and who you trust is a two way street. Both sides must be vested in a project to make it work.

- You might want to call it a 6th sense for the right person.  This is worth every second it takes. Many times over.

4. Look at numbers / stats/ etc. Algorithms are a far cry from perfect.

A little bit of insight into the star system > direct quote from a client "I never give 5 stars. NO ONE IS EVER THAT PERFECT".   🙂

[kat303]

Michael - no one knows how the JSS is calculated. When a client receives their work, some then just disappear without closing the contract and leaving feedback. The freelancer is then left with closing the contract and when that happens, because no feedback from the client was left, that impacts a freelancer's JSS.  Or a client starts a job and then half way through disappears again leaving the job open. The freelancer is left with half a job completed and an open contract. again requiring the freelancer after several attempts to contact the client, to close the contract.

A JSS in the 90's isn't that bad. Take a look and see how many jobs the freelancer has with no feedback and take that into consideration (as well as actual feedback)

[laneski-maureen]

This has happened to me multiple times. I work with someone for months, then my part is done, but the client says they might use me later. They do not close the contract. They maybe even stop checking the platform messages or stop working for that company. So then, I'm left without a review or with an open contract. I had a few huge misses early on, and my success rate was in the high 70s/low 80s. It was in low 90s too. I was paying every month for the premium version for about 2 years. The algorithm changed and my score plunged. I think to 50s. I was told it was because I didn't have multiple contracts with long-yerm clients. We had just added milestones and adjusted rates depending on the work. Then, I cancelled my premium membership and also stopped working on UpWork for a while because I had so many new online tutoring students in 2020. I recently logged back in and my rate is now 35%! Holy moly. So, I am probably just going to leave. I have been networking and getting work outside of UpWork. Who will hire someone with such a low rate? I feel like leaving or starting over makes way more sense than trying to get hired with such a terrible JSS.

[mtngigi]

"Why would someone have less than 100%?"

Because you can't please all the people all the time. We all get clients whom know matter what we do to complete a job successfully, will not/can not, be satisfied. A score lower that 100% does not mean you will get a less-than-stellar freelancer.

In fact most of the really great freelancers on this site have gone through periods where their scores go below !00%. By no means is this a reflection on their work. One unhappy client can bring our success scores down for a while. It is what it is.

Most times it's just simply that things didn't work out and has nothing to do with the quality of work or the freelancer.

[richard_wein]

It's unfortunate that it's called a Job Success Score, as that implies that a freelancer with a score of 90% has had 1 in 10 jobs end in failure. But that's unlikely to be the case. Nobody knows quite what the algorithm is, but it seems that a freelancer's score can suffer a significant adverse effect if a client gives a score of 8 out of 10 in answer to the question "How likely would you be to recommend this freelancer" (not the exact wording). Yet the client may consider that a reasonably good score.

[purplepony]

Michael, are you familiar with "things aren't always what they seem" and similar?  This can be the case with the Job Success Score.  If you're interested in reading numerous viewpoints regarding the subject, there's much to read regarding the topic within this Community.  Regardless---best of luck to you presently and in the future.   

[cdstallcup]

"Why would someone have less than 100%?"

As a freelancer, I had a score in the upper 90s for months. Then one client doubled the scope of the project after we contracted, and threatened to give me a bad rating if I didn't agree to do the work. Once I've contracted, I can't control the rating the client gives me. I said no, and true to his promise, he gave me a bad rating.

On a second job, I was given a different manager to work with every week. Each had different expectations. In the end, whichever manager ended the contract also gave me a low rating. That job was just a mess, and going to end badly - I saw it coming and asked for clearer communications, but couldn't control that outcome.

Just as clients sometimes get freelancers who don't meet your expectations, we get clients who mess up our JSS through no fault of our own. Regarding how JSS is calculated, I'll say I'm one of the best writers I know, but at times my writing ability and reliability don't influence my JSS as much as my ability to intuit which clients are solid and honest.

I'd suggest you go with that 93% rating. My guess is that's a pretty good freelancer. I wish you and the freelancer you choose the best ...

Denise

[resultsassoc]

Wendy C and Richard W have great answers.

The Job Success Score is biased toward recent activity. A freelancer with three long-term hourly contracts may not have received any feedback in the last six montrhs. JSS<100%. A freelancer may have ended a contract for reasons regarding client behavior, and refunded all moneys paid. The now-upset client can leave no public feedback, but her private negative feedback dings JSS.

I have never asked a client to leave feedback. No feedback dings JSS. A few clients have been very satisfied, paid bonuses, but never provided feedback. Anolher ding. I'm sure my JSS is lower than whale poop.

As a client, I do most hiring elsewhere, but have spent more than $5K on elance/Upwork. I read feedback. A 5.0 with no narrative is meaningless. A 4.2 with narrative is far better. A 1.0 on a $50 job has little meaning. The only thing I seek in feedback is repeated decent scores (4+) with alll fives but a three or even two in one specific area. If twelve clients independently found Mohammad Lee to be great, but slow, I won't consider him for an urgent job. If it isn't urgent, or involves something such as weekly consultations in the freelancer's area of expertise, I'm fine.

The first thing I look for in a response is questions. For any project with any importance, that doesn't involve a commodity service, nearly every job post - including mine - fails to provide all the information I need to determine if the client is someone with whom I want to work, what the project's expectations are, and what the real issue is. If I need a business consultant, and get a response from Jack Welch saying "I'm available," that's enough. If I get a decent response from Michael G, without any questions, I'm likely to reject it.

The second thing I look for in a response is price or rate. If I need the services of a great writer, I'm more likely to hire Wendy C, with a relatively high rate, than Clark Kent, with a bottom rate. If he doesn't think he's worth much, I'm not going to argue. Too low a rate is a red flag. I've never turned down a response based on a high rate; I've rejected hundreds who apparently didn't think they were worth a whole lot.

The thing about price is well-illustrated in business plan writiing. A third-world English-fluent freelancer turning out cookie-cutter business plans using software might charge $250. The best software-produced business plan has an upper value of zero. I, and a handful of others, might charge $5,000.

You can get that brain surgery done for $38K, or $9.95.

[tlsanders]

Michael, average star ratings were removed because they tend to be artificially inflated. You can still see the ratings and comments past clients left a freelancer, but the JSS also includes things you'd otherwise miss, such as private feedback from a client who didn't want to "be mean" publicly or a history of making refunds on jobs gone wrong to avoid public feedback.

For that reason, if there's a significant disconnect between star ratings and JSS, you have reason to be concerned.

That said, there are many things that can lower JSS that aren't necessarily performance-related. For example, JSS measures (in a much more complex way) the percentage of jobs successfully completed. So, if a client contracts with a freelancer, then loses his budget and closes the contract without any money changing hands, that's a "not successfully completed." Similarly, when too many clients abandon completed contracts without closing the contract and leaving feedback, that starts to chip away at JSS.

Some clients also just don't tend toward the over-inflated rating system that has created a social norm of giving 5 stars if the freelancer didn't kill the client's dog in the process of doing the job. Thus, a freelancer may get a 4.5 rating rather than 5, from a client who leaves an excellent comment and rehires the freelancer for another job, but that's a negative in the perfect 5.0 world.

For illustration, my JSS is 99%. In the past two years, I've completed 105 jobs on Upwork. 88 of those clients left 5-star feedback, 13 never left any feedback, and four left feedback ranging from 4.1-4.85 stars. At least 75% of my clients have re-hired me at least once.

[financemark]

Its simple really - here is the formula:

Upwork's JSS Formula finally revealed.

[financemark]

Michael,

The score considers some projects as "unsuccessfully completed", when such is not the case.  Also, seems to be very difficult to maintain a 100% JSS over time - and as a freelancer completes more work.

I think all good freelancers here wish to do our very best for our clients, but it is not "always" possible here any more than it is elsewhere.  Speaking for GOOD Freelancers here:

-Sometimes a freelancer takes on a project that we shouldn't.   

-Projects can explode into something not disclosed by client.

-Clients sometimes seek out cheap/free work. 

I had a 100% until a recent client experience. 

Project started off and progressed well. But after initial deliverable was successfully accepted, the client insisted that I keep updating my fnished work to the ever-changing needs of the many partners in his firm.  After successfully completing two of these requests for free, I asked for 1/2 my normal rate for the third, forth and fifth request - which was only funded at the last minute and with complaint.  On his sixth request, I demanded my full rate.  He then attempted to gain free work by saying that my deliverables had errors (they didn't).  He said that he would dispute the charges and leave me negative feedback if I didnt continue helping him for "free".  I chose to take the hit rather than continue to work for this person.  Thankfully, I only worked after funds were in escrow and kept great notes - his request for refund was denied, but he gave me a 2.2 out of 5.0. 

This is an example of an unethical client.

Take some time to read the Freelancer Community Questions and you'll see that there are a number of unethical Clients who use this service.  The very best freelancers usually spot these and don't go there, but it happens. 

[tlsanders]

---

@Mark K wrote:

Michael,

The score considers some projects as "unsuccessfully completed", when such is not the case. 

That's not exactly accurate. It's simply that Upwork only counts known favorables as favorable. It's kind of like voting. If you get ten votes for you and three against you and 20 people don't vote at all, those 20 missing votes aren't counted against you...but, you don't get credit for them because possibly they might have voted for you if they'd voted.

Also, seems to be very difficult to maintain a 100% JSS over time - and as a freelancer completes more work.

I'm curious as to whether others have experienced this. I was here for six months or more with JSS in the 94-99 range before I went to 100, but then stayed there for about a year before a contract dropped me to 99 (as I was expecting), and I've held there ever since.