Interesting.  About a year ago I deleted a bunch of pics and vids that had been uploaded when I first started and with Shutterstock, but felt weren't quality enough to sell after learning the industry a little better.  My sales immediately dropped for over a month, but then really started to pick up pretty good.  Maybe it's time for a little spring cleaning...

"The galleries with "only the best" performed DOUBLE than the one containing the best and the rest, in proportion." what does "in proportion" mean. Its absolute income that counts

So... best images sell better... fine. 

So you have 2 accounts selling the same "best" images?!

Hallo!
I have some friends that sell very, very well. They worked as one photographer using 1 account.
When they started to sell for 5000$ a month they started to have problems with our country taxes so they decided to split, open one account per person and level up the whole of their portfolio.

They had the opportunity to try some "what if" experiment: they left every kind of sh*t in the main portfolio and started to upload ONLY the cream in the new accounts or to return and delete what they considered "not perfect".

The galleries with "only the best" performed DOUBLE than the one containing the best and the rest, in proportion.

So if you were thinking "is it a good idea to delete all my cr*p and leave only the best images?" ... well: yes, it is.

Why?

I don't know: I guess it's because of the way the customer see your portfolio: they see a good image, but not useful to their needs: they open your portfolio and see ONLY THE BEST, without cr*p: this probably is impressive and sells! :-)

I'm not able to do this NOW: but when I'll double my portfolio I'll do this for sure: this may halve the time for the customers to find what they want without seeing all my so-so experiments!

My 2 cents, for you guys

So you have the same images in two different accounts??
Sure that you will sell more.
Why don't you try to open 126 accounts under different names but with the same images???

Hallo!
I have some friends that sell very, very well. They worked as one photographer using 1 account.
When they started to sell for 5000$ a month they started to have problems with our country taxes so they decided to split, open one account per person and level up the whole of their portfolio.

They had the opportunity to try some "what if" experiment: they left every kind of sh*t in the main portfolio and started to upload ONLY the cream in the new accounts or to return and delete what they considered "not perfect".

The galleries with "only the best" performed DOUBLE than the one containing the best and the rest, in proportion.

So if you were thinking "is it a good idea to delete all my cr*p and leave only the best images?" ... well: yes, it is.

Why?

I don't know: I guess it's because of the way the customer see your portfolio: they see a good image, but not useful to their needs: they open your portfolio and see ONLY THE BEST, without cr*p: this probably is impressive and sells! :-)

I'm not able to do this NOW: but when I'll double my portfolio I'll do this for sure: this may halve the time for the customers to find what they want without seeing all my so-so experiments!

My 2 cents, for you guys

Not sure about deleting, but I believe in self-curating the uploads.

Uploading dozens of similars, with minor color or crop variations will definitely harm the sales by spreading them over too many choices, instead of allowing "the best" to climb the popularity ranks.

"The galleries with "only the best" performed DOUBLE than the one containing the best and the rest, in proportion." what does "in proportion" mean. Its absolute income that counts

er... not when you try a model.

They had a user, suppose USER001, this is up and working: it has, NOW, 40000 files.
In the meanwhile they opened USER002 and USER003.

002 and 003 have 10000 pics each.

Obviously you can't do an absolute comparison. But you could do this in proportion.

And in proportion, USER001 that in absolute income is performing well, in proportion is not performing as good as 002 and 003 do.

That means "in proportion": that when user002 and 003 will have the same amount of files, they will perform BETTER than 001.

So... best images sell better... fine. 

I understand that this sounds pretty obvious, but I mean that a portfolio with "only the cream" will sell better than the same portfolio with other files not-so-good in addition, for example the same good photos and the same amount of photos of a lower quality (commercial interest).

Double portfolio doesn't mean double sales. So, if you were thinking to delete some cr*p and someone else is telling "oh I have had a file there for 10 years and tomorrow I sold it!!" ...

... er... ok, you sold it. But HOW MUCH did you sell the whole portfolio ?

This is what they observed: the whole portfolio performed better when there is no lower quality photo inside.

So you have 2 accounts selling the same "best" images?!

I'm not talking about myself: I'm able to work and upload only the non-crap. I try simply to do my best, but I talk with colleagues and they perform VERY well.

They have THREE portfolios, obviously not with the same photos (they know too well what this could cause). They work a lot and divided the shootings distributing them between these portfolios: one of these porfolios is older, so, naturally this has lots more of images. But in proportion it contains lots of not-perfect images.

Is there no overlap of images between ports - so you are just saying a port with good images outperforms a port with bad images?

Or are you saying a port with 1000 good images and 1000 bad images has less sales than a port with only those same 1000 good images? All uploaded at the same time or uploaded at different times?

Are you saying that per image sales are higher or total number of sales?

as far as I know, OP = Original Post - the one that started the thread.

I am guessing for most of us it is rare for buyers to actually even look at our ports, so what matters is the search result for our images. At one time SS seemed to treat every image separately for search - unlike say DT which took the artist into account. Perhaps that is no longer the case, which would be a shame.

As far as I am concerned everybody should delete 60-80% of their weakest images! It will improve my sales.

Whether it will improve your sales? There are too many variables involved to predict anything.

Correlation is not causation.

On the other hand, if I get up from the bed with my right leg and the day name doesn't contain any "U" or "R", it seems to increase my sales on that day quite significantly!
Correlation or causation?

Man, I'm not saying this is "The Sacred Truth".
Just something I could see with my eyes in portfolios that are far better than mine.
I don't want to say who they are, but consider the same quality of "pressmanster" in Shutterstock: I'm absolutely not at THAT level of quality: I can't afford to not upload my "worst" shoots. They are "ok" for me.

But the ones of these friends I'm talking about are at that quality level: they can afford to delete everything that sells bad. Because they observed that this SEEMS to cause a worst performance of the remaining portfolio.

I think that it's an interesting thing to know, rather than omit it. You choose: but now you have some facts.

Is there no overlap of images between ports - so you are just saying a port with good images outperforms a port with bad images?

No overlap.
And no, I'm obviously not saying this thing, because there was no clue in saying this.

I'm saying that a portfolio with 100% perfect images (for your standard, if you're experienced) outperforms the DOUBLE amount of files with the same amount of same quality of images. This means that HALF the amount is not perfect. Just so so, or bad, even if it passed an agency QC. But when you start you think that everything you do have to pass a QC... then you get better and you think "mh... this is not so good... but I worked and it's neither a duplicate nor a slight different POV ... it's work and I spent time, effort and some money for this: let's upload it anyway".

Well: I'm saying that finally you'd better not upload it, or, of you have to, come back 5 years after and delete it, when you doubled your portfolio with better quality images.

Or are you saying a port with 1000 good images and 1000 bad images has less sales than a port with only those same 1000 good images? All uploaded at the same time or uploaded at different times?

It's not possible to do the exact comparison, because the first portfolio (001) is old.
The second (002) and the third (003) aren't: they are pretty new, say a couple of years. In this period of time they shoot to upload in these three different portfolio in this way: they used the same LOCATIONS, the same MODELS and they are a group of photographers.

They NEVER uploaded the same images: this isn't fair and this is dangerous (account closing) if you made a living of microstock.

But you can understand: the only thing that has REALLY changed is the "selection" of "what level of goodness to upload".
Portfolio 001: everything, we're on average good guys and do everything good, we have tons of images and also the old ones cr*p sometimes sells something.
Portfolio 002: the new files, we've got better than some years ago, let's upload everything we shot becase we don't shoot cr*p anymore since years ago.
Portfolio 003: the news files, but hey, wait, this is not perfect. Is good, but not perfect. Don't upload it.

About this third portfolio they often argue because there is lots of effort behind every shooting session. But they did this to stop having taxes problems and they decided that in 3 years every portfolio will be owned by a single person, not only officially, but in reality.

This means that the photographer managing the portfolio 003 is taking a risk not uploading every-good image: they are VERY good. But they are able to shot BETTER than the good things (for my level of quality) : so this man decided to be more selective.

This means cutting the bad files in the past? No, this doesn't MEAN this as a cause-effect scientific process, but this could HELP to make this kind of decision: these friends reasoned about this and I'm trying to talk with you all about this because this is the classic kind of question that sometimes someone asks about, and no one has information or experiments about it.

I have had the chance to see this: these friends don't want to participate in forums or share publicly something: they talk with me and never hide.

Are you saying that per image sales are higher or total number of sales?

er... this is an average, but in this case if per-image average is better, total numer of sales have to be better
of course, as I said, IN PROPORTION.

If there was a direct proportionality and an exact comparison, of course this would mean that you can do what you want.
Instead made this proportion, you could observe that one way of working SEEMS to perform well.

as far as I know, OP = Original Post - the one that started the thread.

Thanks!
aw that's me  

I am guessing for most of us it is rare for buyers to actually even look at our ports, so what matters is the search result for our images.

This is a good point to reason about! In fact this is the ONLY point to reason about. And what I'm saying (for what I've seen, and this is NOT my work: if I worked well as they do I think I'll never delete a single image!!!   ) is that ... well, it seems that this isn't true: it seems that when you work well someone opens your port because they saw a good quality with a theme they were not interested in, to look if you have something interesting: another friend of mine - graphic designer - have some links of authors because he knows that sometimes he could start to be inspired by images and not to search something he has in mind before.

At one time SS seemed to treat every image separately for search - unlike say DT which took the artist into account. Perhaps that is no longer the case, which would be a shame.

I only hope to have given a new perspective and new datas to think about, not only theories without nothing.