Input for Kiva staff on the "Popularity algorithm"

[Guest]

.

[RichardF]

How would this differ substantially from the "Amount Left" sort?

I believe adding the time factor levels the playing field for the "Loan Amount" and "Date Listed" factors and it turns "Loan Amount - Amount Left" into a true Popularity index in terms of the working definition, "The faster the rate (amount/time) a loan is funded, the more 'popular' it is."

[alan]

So a $1000 loan with 1 hour left and $975 raised wouldn't be as popular as a $100 loan with $25 raised and 28.5 days left.

I agree in principle that what your proposal does is to measure popularity in about as objective a way as possible. There are doubtless other ways to do it with similar claims to objectivity.

I still see two problems: one is that Popularity is a very blunt instrument. The second is the equation of Popular with Compelling.

[RichardF]

My personal preference is to eliminate the Popularity sort and beef up user-defined search and sort capabilities.  But this is Kiva's game and it sets the rules, so I though I would play it for a bit.

[YowieFreak]

The best way to ensure that nothing gets done is to set up a committee to decide what should be done.  And the larger the committee, the longer it is that nothing gets done.

I wonder if that's what the idea is behind asking us to come up with a recommendation?!?!       

[cpbailey]

There is a business saying, "Whatever gets measured, gets done.  Be careful what you measure."  It is apt here.  Basically, take the goal and then use whatever parameters encourage that goal.  What SHOULD the front page loans and first page indicate are most important?  Unless that is decided, it is skipping a vital step.

By the way, Donors Choose has a different dynamic than Kiva.  Donors Choose allows many individuals to post with the expectation that many will not get fully funded.  With Kiva, there are limits in total amount a limited number of organizations can post.  This has somewhat scaled with investments.  Further, Kiva's expanding investments roll over, creating a 15th of the month surge that gets bigger every month.  Donors choose is a new investment each time.  Under Donors Choose, it is wise to put ones that are close to being fully funded at the top of a sort.  Also, a way to help spread the love is to prioritize teachers' first projects on Donors Choose. 

Colette

[Limesarah]

I don't know if this should go here or in another thread, since it's more philosophizing than concrete advice, so feel free to move it...

I've been feeling increasingly uncomfortable with the tone of Kiva management's discussions about popularity ranking.  Sure, Kiva is giving the MFIs interest-free loans, but it's still a major commitment of time and other resources to create and upload Kiva profiles.  It seems unfair to set the system up to benefit the most "popular" loans -- you're taking an organization that you're helping specifically because they're underfunded and underappreciated, and then making them do more work if they want your help.  One of the major points of Kiva's mission seems to be to fund the world's "unpopular" businesspeople...but then management seems intent on creating a popularity hierarchy within that population. 

It's certainly statistically interesting to see which loans are more popular, and if Kiva can give the MFIs some advice on what makes a good writeup, all the better.  But if Kiva is working to help fund disadvantaged businesspeople, why then set up a system to disadvantage them?  I (theoretically, at least) trust that any loan on Kiva has been vetted by a responsible MFI and thus is "worth lending to".  Shouldn't Kiva be trying to highlight those loans that get lost in the shuffle?  I'd love to see a post on the Kiva blog about why people from some cultures don't want to smile for their loan photos.  This would increase intercultural knowledge and maybe help people chose that nonsmiling borrower next time, and seems a lot more respectful than telling the MFIs that smiling photos get funded faster.

[tomviolence]

Unfortunately, it has been definitively shown that the identical loan (A duplicate) funded at two very different rates. This proves it is NOT the write up, photo, smile, color of the cow etc.

Therefore, blaming the MFI for Kiva's inventory issues is wrong.

Yes, a clear write up, nice picture & variety in posting helps, but the whole "it's the MFI's fault laons expire" is a chimera

[YowieFreak]

... color of the cow etc. ....

Maybe the Fellows could teach the MFIs how to photoshop their loan pictures to make multi-coloured cows, etc - I bet that would increase popularity!!

Or perhaps photoshop a brand name onto the cow - that might encourage corporate lending!  
(I wouldn't go so far as encouraging the actual laser etching of the brand name onto the actual cow - that would remind me too much of Kelloggs Cornflakes.)

[JoanW]

The best way to ensure that nothing gets done is to set up a committee to decide what should be done.  And the larger the committee, the longer it is that nothing gets done.

I wonder if that's what the idea is behind asking us to come up with a recommendation?!?!       

Hmmmm...you could be on to something here

Seriously, we have asked Kiva to make a change in how they do things. They have asked us to come up with how exactly we want them to make that change. So far we have 7 pages of opinions, and a few suggestions.  Since Kiva has been clear they will keep the popularity sort - elimination is my favorite answer too - we need to come up with something that meets Kiva's goals and addresses the concerns KFers have expressed.  Maybe a poll to see which of the less than perfect solutions offered is the most favored?  I feel for our Liaison trying to herd this bunch

[Jan & John]

I have a very simple mind most days...

To me Popularity means exactly that... "this loan is going fast, get some quick" !!!

And certainly not manipulated by any kind of behind the scenes magic...

Kiva already shows the percentage raised... as I type the most 'popular' loan is only 6% raised - duh!.

The most popular loan should be the one with 99.99% raised.

If Kiva wants to 'weight' the loans for their own reasoning, then please call it something else.

jan