Sign in to follow this  
scottbraut

Answers to the "Questions for Scott" thread

Recommended Posts

Hello All,

 

Here are some answers to some of the commonly asked questions found in the original "Questions for Scott" thread. Given the number of close variations, I've gone ahead and separated common questions into general topic areas.

 

Understanding Search

 

We often get asked how our search rankings work. The first important thing to note is that there isn’t one singular search and discovery experience at Shutterstock. In addition to the core search experience at Shutterstock.com, we have Palette, Instant, Spectrum, and People search; mobile apps like our iPad app; as well as different sort orders such as “New,†“Popular,†“Relevant,†and “Random.†We also have Bigstock and Offset. There are many ways for customers to get to your images, including direct visits from Google, and many ways for them to filter the collection for specific file types like vectors or video. We localize our results for specific countries as well - i.e., it makes sense for Japanese customers searching in Japanese to see the most appropriate results for their own language.

 

Our search rankings are complex and based on a number of factors. Keywords, freshness, and popularity (downloads) all factor into search results. We focus on providing customers with the best image that matches their needs and we don’t favor specific contributors or portfolios. Location can be a factor when the search query calls for it (for example, ‘streets of New York’), or when region-specific results are appropriate.

 

We believe that if we always match the best possible media with a customer, over time this will lead to more satisfied customers, and in the end more downloads (and more royalties) for our contributors. That belief has delivered over 400 million paid image licenses to date.

 

Search is powered by a set of services that include machine-learning algorithms: a type of software that is able to improve over time. In a very simple sense, the goal of an algorithm is maximize downloads. To get the right media in front of a customer, the algorithm takes many factors into account, such as how well the keyword matches a search query, or the past success of this particular photo, music clip, or video. For some search queries, customers are looking for newer content or content that has not been used before, and the algorithm will leverage factors such as the age of the media; for other queries, age is less important. There are many, many other factors that the algorithm takes into account, and we are adding more every day. One of the ways we improve our algorithm over time is to add new factors. We are also frequently making small changes in how the algorithm weighs these factors.

 

In practice, some of the changes we make dont always work out for the best, so when we make a change, we need a way of knowing if it’s a good idea. We achieve this through an experimentation process called A/B testing. In an A/B test, a typically small amount of our customer base is put into an experiment in which we’ve changed the algorithm, and we’ll measure how well it performs relative to a control group. There are often many such experiments going on at Shutterstock, and its probable that you’ve been in more than one if you are a regular Shutterstock user. If a change results in more downloads or improvements in other key metrics, it will usually find its way into the production experience. If not, we’ll move on and try something else. You can think of the process a little like evolution – we are constantly getting slightly better through a process of trial and error in very small amounts.

 

It is often the case that a new experiment will change the search positions of the media in our collection. This is understandably a concern for many contributors, but its important to keep a few things in mind. First, experiments are generally pretty small in terms of exposure to the customer user base, and as mentioned, our algorithm is constantly maturing. But more importantly, performance should be viewed over many experiments. In general, we will only ship an algorithm change if it results in more downloads, and thus more income for our contributors. The impact of a single change will vary from one set of images to the next, but over many changes it’s almost always a net win for everyone. In addition to this, every time we make it better to find great content, we also get a little better at attracting new customers, which means more new customers for our entire contributor community to share their media with.

 

The most important thing to understand is that there are dozens of things that you control to improve your earnings, the discoverability of your work and the search position of individual images in your portfolio. While they may at time seem basic, we carefully study the attributes of top-selling images and portfolios, and they can dramatically impact success.

 

Without listing them all here, please read 33 Things That You Control to Increase Earnings and Success.

 

We’ll post more about keywording and on this overall topic soon.

 

About Review - Questions about review and review disputes

 

With over 40 million images on Shutterstock alone, our collection is one of the largest of its kind. Our team reviews millions of images, and in terms of quantity, we get relatively few complaints about Review. That being said, they do happen. When you feel strongly you've had a "bad" review, we ask (and encourage) you to write into submit@shutterstock.com.

 

Our team investigates every single review dispute, no matter who the contributor is, whether we've looked before, or whether we've agreed or disagreed in the past.

 

We commonly find that many disputed reviews reflect a correct review, or at least, it was understandable why the reviewer made their decision. A very small amount are found to be true errors or issues, which are then addressed with the reviewer. If a pattern were found, we would address it. Sometimes, it's just a matter of confusion, or of a rejection reason being too broad, or of meaning something different to each party involved.

 

We review ten(s) of millions of images. We have a high motivation and interest in driving the number of errors made down to zero, since errors become support requests and second reviews. Reviews incur a cost to us regardless of whether an image is accepted or rejected, so we have no incentive to reject images intentionally or by mistake, or to review images a second time. We also have little incentive to accept images if they fall beneath a certain quality level, but we reverse review decisions if we feel the case was an error or borderline.

 

It's important to also note that we track various metrics related to reviews and reviewers, so if there’s an abnormal trend, we see it.

 

There's not always a lot to say on this topic because every situation is different - in some cases, a dispute is 100% legitimate and we deal with it. Sometimes, the original review was correct. Without going into each individual review publicly in the forum, there's no single, categorical statement that can be made.

 

There are a number of things that are in our plans to improve, such as the resubmission process, and even some system improvements to better improve consistency across reviewers.

 

If you feel strongly that you've had a bad review, please contact us.

 

On Communication

 

Lately we find ourselves supporting more and more communication channels. To a fault, we’ve made a tradeoff by going very “broad†with our interactions and not as “deep†into any one channel. We’re communicating across social media, workshops, foreign-language sites, blogs, community events, etc., but we’ve often failed to give enough detail in specific threads in the forums. We’ve had the best of intentions, but our "true" goal is to exceed all expectations.

 

We have at least three people active in the various forums and many more than that communicating with contributors daily; I often take responsibility for the MicrostockGroup forums but will also post at SS; Vincent posts at the Shutterstock forums and Anna posts in the Russian-speaking forums. Forums are one of dozen(s) of communication channels we maintain, including international events, email, social media (link 1, link 2, link 3), blogs, multilingual guides (link 1, link 2), workshops, dinners and office research visits, etc. We've added interviews with reviewers and review coordinators to our blog content, and there's a larger team that is engaged with contributors through those channels.

 

Since we're talking to Shutterstock's 60k contributors through all of these channels, there's no one channel that's more important than the rest, with the possible exception of email (see below).

 

All that being said, we're going to spend more time here and we're looking forward to a forum relaunch on new software.

 

Historically -- it's important to note -- email is our only "official" support channel because email tickets can be assigned, tracked, and resolved efficiently at scale. But we do try engage in the forums when there are bugs -- see examples like this and this. Again, email is best, because we can track, quantify, forward / assign and resolve issues appropriately through the email system; forums are much more difficult to use for that purpose.

 

For specific contact addresses, please see this thread.

 

Our team watches the forums regularly for feedback, even if specific individuals do the majority of the posting. I'll be posting here with more regularity.

 

Regarding Royalties

 

We’ve paid out over $200 million to contributors in royalties to date. The current royalty structure has done a good job of balancing contributor earnings with the overall growth of the business. We’re driving significant sales growth at higher price points through our enterprise and "Direct Sales" business, and that is increasing royalties on a dollar basis for an individual sale --- to up to $120 or more in royalties for a single download. Because On Demand and “any products not listed†prices are based on a percentage, as pricing goes up (which it often has), so do royalties. We’ve paid out more in royalties every single quarter since 2005 (representing over 35 quarters of increased growth) while keeping our overall royalties rates and structure relatively stable.

 

Intellectual Property and Piracy

 

Shutterstock actively investigates piracy websites and every claim of piracy made concerning Shutterstock content. Through our investigations and efforts, we have been able to have content removed and some sites disabled. It’s sometimes the case that pirate sites change their URLs or physical locations so quickly that it can be very challenging and expensive to force them into compliance. It’s also often the case that these sites are distributing malware or trying to lure people into paying to download content that doesn’t exist. We appreciate your efforts in bringing these sites to our attention. We can’t provide more details of legal and compliance investigations publicly. If you find an infringement, website or issue of concern, please email infringementclaims@shutterstock.com. We take those complaints very seriously.

 

Bugs and feature requests

 

Please see this thread for basic questions related to bugs.

 

We’re going to post more updates to the forums on specific bugs. Depending on the bug or feature, it's often the case that work is being done, but that might not always be clear to the external observer. The forums themselves, for example, are going to be replaced entirely very soon, which is why the current search bug hasn’t been addressed.

 

The way we handle development, we generally don’t do big, monolithic releases the same way that a software manufacturer would. The team is constantly pushing code live, constantly fixing (or preventing) bugs, and constantly working on new features.

 

To understand a little more about the prioritization process, please see this thread.

 

New Forums, Guidelines

 

We're going to be relaunching the forums on a new software platform as we invest in nurturing communities and contributor education.

 

The Shutterstock Forums are meant to be a helpful sounding board for all members of the Shutterstock community -- new and veteran -- on submissions, review, and content. We highly encourage your productive comments, and do take your concerns into account, so that we can make working with Shutterstock better for you (although we are not able to always respond to each forum post).

 

In the spirit of keeping the discussion constructive and helpful, we'd like to remind you that the Terms of Use for Shutterstock Forums does not allow abusive, slanderous, hateful or threatening posts. This type of activity will not be tolerated, and forum access may be revoked for offenders.

 

Here are some additional guidelines to follow:

 

- If you have a specific concern about your submissions, account, or content, please contact us immediately at submit@shutterstock.com with all relevant information. We can more quickly assist you if you send the information directly to us.

 

- Please be respectful of other contributors, community leaders, moderators and staff. Avoid insults, ad hominem and personal attacks, as they are not constructive or helpful to the Shutterstock community. Remember, every existing member is a potential role model for the next new member to join the forums.

 

- If you summarize or post correspondence between yourself and a Shutterstock representative, please do not alter or truncate responses you receive from Shutterstock. This can be misleading to other contributors, and misinformation is counter-productive.

 

As always, feel free to write in to submit@shutterstock.com if you have any questions.

 

Happy posting. :)

 

---------------------------

 

While posted under my name, this post contains input or contributions from multiple members of the team at Shutterstock. To meet them directly, please keep an eye on the Contributor Blog.

 

The conversation is far from over (it's really just begun), but thanks for your feedback and contributions. If you ever have a chance to meet us in-person at an event in your city or town, please don't hesitate to do so!

 

 

Best Regards,

 

 

Scott

VP of Content

Shutterstock

Share this post


Link to post
Share on other sites

First up, Thanks Scott for the reply. Quick question around the Search.

 

" We achieve this through an experimentation process called A/B testing. In an A/B test, a typically small amount of our customer base is put into an experiment in which we’ve changed the algorithm, and we’ll measure how well it performs relative to a control group. There are often many such experiments going on at Shutterstock, and its probable that you’ve been in more than one if you are a regular Shutterstock user."

 

This is the same as what we refer to as Champion/Challenger analysis.

 

While you check the impact on overall sales of each change, do you also check the impact on the contributor base? For instance, an improvement for one contributor may switch off the port of another altogether.

 

Now I appreciate that it is impossible to look at each contributor individually but it would be simple enough to categorise contributors. For instance,

 

1. Time at SS, No uploads per month, current earning threshold, specialised or broad variety of images in port.

 

This is just a rough and ready suggestion but you get the idea.

 

If you create even a couple of dozen of these categories you could then check how each group is impacted during each experiment in the A/B analysis as well as the overall impact on downloads. This may help identify when a change has had a bigger impact on one group over the others.

 

I would assume (given volumes of data) that you will be using SAS to do this work, or something similar. If so, once each contributor has a category code, it would be very simple to create a Proc Tabulate Table & Scatter Graph to see the impact on each contributor group.

 

If contributor impact is not being analysed during each experiment, this may explain when (even if DL's improve) why one sector of SS contributors is being hit harder than others?

 

 

Edit.

Maybe what would be a good idea is to take one Contributor Account that has been negatively impacted via changes over the last year or so (big drop in sales) and then do a Deep Dive analysis and report back the reasons for the decline.... i.e. specifying which reasons out of the 33 etc. This would give them and the rest of us some idea as to why it has effected some accounts more than others.

Share this post


Link to post
Share on other sites

 

Intellectual Property and Piracy

 

Shutterstock actively investigates piracy websites and every claim of piracy made concerning Shutterstock content. Through our investigations and efforts, we have been able to have content removed and some sites disabled. It’s sometimes the case that pirate sites change their URLs or physical locations so quickly that it can be very challenging and expensive to force them into compliance. It’s also often the case that these sites are distributing malware or trying to lure people into paying to download content that doesn’t exist. We appreciate your efforts in bringing these sites to our attention. We can’t provide more details of legal and compliance investigations publicly. If you find an infringement, website or issue of concern, please email infringementclaims@shutterstock.com. We take those complaints very seriously.

 

 

Thanks for the efforts, however this does not come close to addressing some questions from the other thread...

 

 

Trust in Shutterstock is waning very quickly.

 

Scott, I would like to see an answer to this one.

 

http://submit.shutterstock.com/forum/viewtopic.php?t=141075

 

 

 

From the link...

 

 

There are 120GB+ packed (about 50GB in size) of shutterstock images on torrets. They're there, they're seeded so cannot be removed and damage cannot be undone.

 

What SS needs to do is work out HOW someone got hold of a huge chunk of their collection to leak.

 

How did someone get a hold of a huge chunk (120GB+ packed)of SS collection?

Share this post


Link to post
Share on other sites

Scott, you've put a lot of work into writing that reply, so thanks!

 

Duncan makes some good points, even though some of the technical terms are a wee bit above my head! Personally, I find that a slow period in terms of downloads is usually followed by a period of catch-up, so overall numbers tend to average out over time. That's not quite what others are reporting, though.

Share this post


Link to post
Share on other sites

Thank you Scott I appreciate your time but I do have a question or two.

 

You did explain many things but you did not touch on these areas?

 

1. What about a newsletter being sent out to all contributors? This can go a long way to solve the communication issues in an easy and effective way. This way you give us all the information instead of us having to go out and hunt it down.

 

2. Royalties - We have not had a raise in years. True you have come up with larger sales such as SODs etc but those for many are few and far between. You say you are paying out more royalties than ever. That may be true but that could be due to you selling more images because of more images in the database and more contributors.

 

When we reach out highest limit that is it. No incentive to do better.

 

Review process - Have you ever given thought of a different review process that rewards the established successful contributor? Many of us do this for a living, we know photography or illustrations in some cases perhaps better than the reviewer. It is in our best interest to submit good sellable work so we make more money.

 

Why not have a review process that acknowledges that and if the contributor is established and submits good solid work that the review goes into a spot check mode? There are agencies that do that. It can speed up the review process because the reviewer is not spending the same amount of time on images submitted by people who have proven themselves.

 

One point you did not address that most of us mentioned was the sudden and dramatic downturn in sales that many of us saw starting about a year ago.

 

As I said there are always going to be sales patterns that ebb and flow within certain parameters. But when you see a sudden change then something has changed internally.

 

What about the dead zones many of us mentioned? We have had many who report a complete stoppage of sales for hours and hours on end. Then they suddenly start again. This has never happened before but then wham it started. Just like with sales it is not normal and indicates something internally has changed.

 

I have worked with numbers much of my adult life as a statistician for the police department I worked for. There are others here who have been in numbers all of their lives and we know the things that have been happening are not normal.

 

It seems like so much has changed since SS went public. Its like the suits and accountants came in and took over the show and do nothing but look at the bottom line and forget about those who helped get the company to where it is now.

Share this post


Link to post
Share on other sites
Thank you Scott I appreciate your time but I do have a question or two.

 

You did explain many things but you did not touch on these areas?

 

1. What about a newsletter being sent out to all contributors? This can go a long way to solve the communication issues in an easy and effective way. This way you give us all the information instead of us having to go out and hunt it down.

 

2. Royalties - We have not had a raise in years. True you have come up with larger sales such as SODs etc but those for many are few and far between. You say you are paying out more royalties than ever. That may be true but that could be due to you selling more images because of more images in the database and more contributors.

 

When we reach out highest limit that is it. No incentive to do better.

 

Review process - Have you ever given thought of a different review process that rewards the established successful contributor? Many of us do this for a living, we know photography or illustrations in some cases perhaps better than the reviewer. It is in our best interest to submit good sellable work so we make more money.

 

Why not have a review process that acknowledges that and if the contributor is established and submits good solid work that the review goes into a spot check mode? There are agencies that do that. It can speed up the review process because the reviewer is not spending the same amount of time on images submitted by people who have proven themselves.

 

One point you did not address that most of us mentioned was the sudden and dramatic downturn in sales that many of us saw starting about a year ago.

 

As I said there are always going to be sales patterns that ebb and flow within certain parameters. But when you see a sudden change then something has changed internally.

 

What about the dead zones many of us mentioned? We have had many who report a complete stoppage of sales for hours and hours on end. Then they suddenly start again. This has never happened before but then wham it started. Just like with sales it is not normal and indicates something internally has changed.

 

I have worked with numbers much of my adult life as a statistician for the police department I worked for. There are others here who have been in numbers all of their lives and we know the things that have been happening are not normal.

 

It seems like so much has changed since SS went public. Its like the suits and accountants came in and took over the show and do nothing but look at the bottom line and forget about those who helped get the company to where it is now.

 

Agree 100% with everything Dave says here.

Share this post


Link to post
Share on other sites

I read your reply Scott and thanks for taking the time to do it but I found no real answer to my two main questions.

 

1. Are we ever going to get a raise and/or higher sales tiers to aim for.

 

And

 

2. Do the reviewers work with the adequate equipment in order to give a fare and true review (ie calibrated monitors)

 

I know you feel you answered these questions in an understandable manner but I feel the answers need to a bit more blunt and not so wrapped up in (but we do this and we've done that) speak.

 

It's simple!

 

Are we ever going to get a raise anytime soon? (yes/no)

 

Does the reviewers equipment go through a reviewing process by Shutterstock? (yes/no)

 

The price of everyday items are not what they were 5 or 6 years ago and it's only a handful of people in the grand scale of things that are getting the benefit of the $120 SOD's so how can Shutterstock expect the average contributor to continue investing their money in new equipment to keep up with shutterstocks rising standards (if the current reviewing process is anything to go by)?

 

We need proper, straight answers. Or at least I do!

 

Tony

Share this post


Link to post
Share on other sites
.....

The price of everyday items are not what they were 5 or 6 years ago and it's only a handful of people in the grand scale of things that are getting the benefit of the $120 SOD's so how can Shutterstock expect the average contributor to continue investing their money in new equipment to keep up with shutterstocks rising standards (if the current reviewing process is anything to go by)?

 

We need proper, straight answers. Or at least I do!

 

Tony

 

Exactly!

 

(If I tried to convince my suppliers I should only be paying them the same as I was 6 years ago, despite their rising costs, I wouldn't get very far!)

Share this post


Link to post
Share on other sites

thanks scott. i sure hope no one told you this would be easy. however, please don't let the dog sleep with your reply. there are wolves lurking, and they are hungry for answers.

between the lack of incentive, and the possibility of being ripped off - and no real concrete foundation from which we can put our future to build more for you on, i'm seeing a slow down on the reviewers side of this. that's what i'm walking away with.

Share this post


Link to post
Share on other sites

Hi All,

 

Thanks for the questions. On royalties and payouts, there are really two sides of the answer to that question.

 

First, in the "SOD" and "other products" categories, the top royalties **have** increased from a few dollars to $35, $75, $120 or more per download. If you think about Fortune 500 companies, large publishers and ad agencies on a worldwide basis - we're growing in those segments and increasingly bringing those clients onto enterprise plans. They may previously have had creative directors or art directors on a 25-A-Day plan, and now those same users are on an enterprise agreement generating much higher payouts for contributors. On a per download basis, there have been big jumps at the high end in the payouts you're receiving.

That part of our business was growing, as is our overall customer base (at over 1 million customers).

 

There isn't a plan (today) to add another tier above the 38c for 25-A-Day subscription downloads. I couldn't speculate on the future.

 

On equipment - yes, we check for equipment choices. I can't go into every detail of our workflows, but equipment and color accuracy are things that we manage as part of both everyday operations and the hiring process. We work with the reviewers daily and also "review their reviews"; if an issue is ever discovered, we work on that with the team or the reviewer.

 

Best,

 

 

Scott

Share this post


Link to post
Share on other sites

Also guys,

 

An apology in advance - there are more of you than there are of me / us, so I'll try to answer as many of these as possible. Since every post contains multiple questions, if I don't get to every question right away, hopefully they'll get answered as the thread develops or over time.

 

Best,

 

Scott

Share this post


Link to post
Share on other sites
1. What about a newsletter being sent out to all contributors? This can go a long way to solve the communication issues in an easy and effective way. This way you give us all the information instead of us having to go out and hunt it down.

 

Hi David,

 

There is a newsletter, which generally summarizes posts to our blog and other major communications. I'll check on the most recent email version and where that stands - we've been creating a lot of content for the blog, workshops and social media, as well as the downloadable guides (the Protect Your Content guide is over 80 pages in multiple languages). Since our contributor base is fairly diverse and global, we spend a lot of time translating materials into different languages, and we're in the process of translating the site itself into more languages.

 

I'll talk to the team, but if anything, I see communication overall on a vast upswing, even if we haven't always been deep in individual threads in the forums. For example, most of our Support requests went from multi-day turnarounds in prior years to same-day turnarounds during the week. We've got a new version of the Knowledgebase that we're going to release, and new software for the forums. Things are getting better!

 

Best,

 

Scott

Share this post


Link to post
Share on other sites
1. What about a newsletter being sent out to all contributors? This can go a long way to solve the communication issues in an easy and effective way. This way you give us all the information instead of us having to go out and hunt it down.

 

Hi David,

 

There is a newsletter, which generally summarizes posts to our blog and other major communications. I'll check on the most recent email version and where that stands - we've been creating a lot of content for the blog, workshops and social media, as well as the downloadable guides (the Protect Your Content guide is over 80 pages in multiple languages). Since our contributor base is fairly diverse and global, we spend a lot of time translating materials into different languages, and we're in the process of translating the site itself into more languages.

 

I'll talk to the team, but if anything, I see communication overall on a vast upswing, even if we haven't always been deep in individual threads in the forums. For example, most of our Support requests went from multi-day turnarounds in prior years to same-day turnarounds during the week. Things are getting better!

 

Best,

 

Scott

 

I have never received an emailed newsletter as far as I can remember.

Share this post


Link to post
Share on other sites

First, thanks for taking the time. While I'm still befuddled by the most popular rankings within our own portfolios, I agree more strongly with Dave:

 

One point you did not address that most of us mentioned was the sudden and dramatic downturn in sales that many of us saw starting about a year ago.

 

As I said there are always going to be sales patterns that ebb and flow within certain parameters. But when you see a sudden change then something has changed internally.

 

What about the dead zones many of us mentioned? We have had many who report a complete stoppage of sales for hours and hours on end. Then they suddenly start again. This has never happened before but then wham it started. Just like with sales it is not normal and indicates something internally has changed.

 

 

And maybe I missed this, but are we ever going to be able to see top selling images or top keyword searches again?

Share this post


Link to post
Share on other sites

Thanks for opening the line of communication, Scott. I would like to re-address an issue that has been discussed numerous times here in the forum (which I'm sure your team is aware of ...since they have been monitoring)

The MOST POPULAR search mode within our own galleries has been totally nonsensical for months on end. For instance, my all time bestselling images that have sold thousands of times appear more than half way through my portfolio while there are numerous images that have not performed well and have had dls in the single digits showing up high in my port. There doesn't seem to be any rhyme or reason to the way they are displayed once you get past the first page or so. (and I know I'm not the only one with this type of experience) The images that are showing up on my first page are what I would consider my more recent popular images and I can understand how they advanced and have no problem with that but after that, it's just quite baffling.

 

I think it's a misrepresentation to buyers who may find an image at SS, then seek out an individual portfolio and not be presented with the contributor's best work. If the search mode is entitled 'Most Popular' then why wouldn't images that have performed well and sold often for many years show up above an image that has sold one or two times over a period of several years?

 

If the search is to stay this way, then would it be possible for us to have a Most Downloaded or 'All time favorites' mode or something to that effect (just within our own galleries?) It seems that we should have a little more control over our individual presentation to customers. I realize that we have the SETS option, and I have utilized that but I doubt that many buyers actually find that tab and go searching there. It's also quite unfriendly to contributors, since the images can't be rearranged within the set and new ones added show up at the bottom of the set...unless you start over.

 

If this is not a glitch or bug which I assume it isn't since it's been questioned for so long and never addressed, and if there are no plans in the works to make changes within our portfolios, would SS please consider looking into some way to allow us to present our best work to customers when they stop by. Buyers have limited time and we all know that no one is going to browse page after page of images to seek out the exceptional ones that may be buried on page 9...

 

Again, this is only an issue within our own galleries and not in the general site search.

 

thanks so much for taking the time to address our questions

Share this post


Link to post
Share on other sites

also... Why is it that when we go to our Image Gallery from the contributor page, the images are large and create a nice display while when a customer finds our portfolio by clicking on the contributor name...the thumbnails are small and less than impressive?

Share this post


Link to post
Share on other sites

While you check the impact on overall sales of each change, do you also check the impact on the contributor base? For instance, an improvement for one contributor may switch off the port of another altogether.

 

Now I appreciate that it is impossible to look at each contributor individually but it would be simple enough to categorise contributors.

 

Hello,

 

We regularly track and analyze a number of "contributor success" metrics.

 

I don't think it's right to think of a test or change as "shutting off" a portfolio - that's not really the way it works. There are multiple entry points to a portfolio of images. Tests are rolled out, often to a very small portion of the customer base, and when those tests are found to increase customer downloads (i.e., customers are more interested and enthusiastic about what they're finding), the algorithm tweaks are adopted more broadly.

 

When we look into complaints about the performance of individual portfolios, it's often the case that the issue seems to stem from small portfolios, not a lot of diversity, not much upload volume or robust keywords, etc. The "33 Things" post may seem very broad and at times basic, but it's very real.

 

One educational game that we sometimes play in workshops is to pull three images with almost exactly the same subject matter, similar upload dates, etc. We have the audience vote on which image gets the most downloads and why. Based on styling and composition alone, there can be vast differences in image performance. The good news is that contributors are in control of those 33 things.

 

Contributors should voice when they think things are awry, but many of our top earners focus on those 33 things and regularly creating and testing new content, diversifying their portfolios, careful keywording, filling content gaps, etc., and their portfolios perform better as a result.

 

Part of our growth as a marketplace comes from being very open to the idea that we don't always know what the customer wants; the market is very dynamic, etc., and you constantly have to be testing and learning, analyzing success and failure, and making adjustments. Across 1 million customers in 150 countries, you have to willing to acknowledge that at any given moment, there's lots more to know. But that's also where opportunity sits.

 

We're here to help you figure it out.

Share this post


Link to post
Share on other sites
also... Why is it that when we go to our Image Gallery from the contributor page, the images are large and create a nice display while when a customer finds our portfolio by clicking on the contributor name...the thumbnails are small and less than impressive?

 

Scott - Thank you for the responses.

I too would like to see this addressed. Customers should have the best display possible of our images.

Share this post


Link to post
Share on other sites
First, thanks for taking the time. While I'm still befuddled by the most popular rankings within our own portfolios, I agree more strongly with Dave:

 

One point you did not address that most of us mentioned was the sudden and dramatic downturn in sales that many of us saw starting about a year ago.

 

As I said there are always going to be sales patterns that ebb and flow within certain parameters. But when you see a sudden change then something has changed internally.

 

What about the dead zones many of us mentioned? We have had many who report a complete stoppage of sales for hours and hours on end. Then they suddenly start again. This has never happened before but then wham it started. Just like with sales it is not normal and indicates something internally has changed.

 

 

And maybe I missed this, but are we ever going to be able to see top selling images or top keyword searches again?

+3

 

These questions were not addressed and we should have had a raise years ago. Our expenses have gone up and that issue needs to be resolved.

 

The dead zone is once again being ignored.

 

Many questions were not addressed at all and that leaves me feeling like shutterstock does not care about its submitters.

Share this post


Link to post
Share on other sites

While you check the impact on overall sales of each change, do you also check the impact on the contributor base? For instance, an improvement for one contributor may switch off the port of another altogether.

 

Now I appreciate that it is impossible to look at each contributor individually but it would be simple enough to categorise contributors.

 

Hello,

 

We regularly track and analyze a number of "contributor success" metrics.

 

I don't think it's right to think of a test or change as "shutting off" a portfolio - that's not really the way it works. There are multiple entry points to a portfolio of images. Tests are rolled out, often to a very small portion of the customer base, and when those tests are found to increase customer downloads (i.e., customers are more interested and enthusiastic about what they're finding), the algorithm tweaks are adopted more broadly.

 

When we look into complaints about the performance of individual portfolios, it's often the case that the issue seems to stem from small portfolios, not a lot of diversity, not much upload volume or robust keywords, etc. The "33 Things" post may seem very broad and at times basic, but it's very real.

 

One educational game that we sometimes play in workshops is to pull three images with almost exactly the same subject matter, similar upload dates, etc. We have the audience vote on which image gets the most downloads and why. Based on styling and composition alone, there can be vast differences in image performance. The good news is that contributors are in control of those 33 things.

 

Contributors should voice when they think things are awry, but many of our top earners focus on those 33 things and regularly creating and testing new content, diversifying their portfolios, careful keywording, filling content gaps, etc., and their portfolios perform better as a result.

 

Part of our growth as a marketplace comes from being very open to the idea that we don't always know what the customer wants; the market is very dynamic, etc., and you constantly have to be testing and learning, analyzing success and failure, and making adjustments. Across 1 million customers in 150 countries, you have to willing to acknowledge that at any given moment, there's lots more to know. But that's also where opportunity sits.

 

We're here to help you figure it out.

 

Thanks Scott. I would imagine a lot of, but not all of the 33, could be variables within the scorecard / algorithm? While others will not but could directly influence the customer to pick one image over another (Retouching - Post production).

 

I think what would be useful if contributors knew which are in the search system and which are not. Of course, I wouldn't expect you to advise of the weighting each has in the system (some may have a very small impact while others significant etc) as you wouldn't want people to play the system or reveal to much to competitors. But to split that list into 2 groups would prove even more useful when planning / building a portfolio as we would have a better understanding of what's important to the customer.

 

For instance, if the search is adjusted to maximise DL's, then having an understanding of what variables are in use to maximise those DL's can only mean contributors become more aware of whats important to the buyer. Of course, this could only be done at a high level or otherwise you reveal to your competitors the formula for success.

 

I know it's a fine line about getting the amount of information revealed to contributors but it does feel like SS could gain more by revealing a little more than it does at the moment. The removal of some of the analysis tools in under "Insights" hasn't really helped either.

 

With regards to "Shutting Off", I don't believe this is connected to the search at all but is more likely to do with SS ranking. Of course, the search will indirectly affect this as any ranking system would be based on the success/demand for a contributors images. If they don't show up in a search, then the less successful they will become and a downward spiral can result (if no action is taken).

 

So, the question is not, "do SS shut down ports" but, "does SS ranking fluctuate the amount of time an individual port will be exposed for". If the time is increased and there is little sign of an increase in sales, they go down, if there is a jump in sales, ranking goes up. In this situation the amount of exposure would also be affected by the volume of contributors.

 

I wouldn't expect SS to say how they rank contributors but I would've thought they could confirm how the ranking affects the Contributor i.e. the question in Bold

 

Cheers.

Share this post


Link to post
Share on other sites

 

The dead zone is once again being ignored.

 

Hi Susan,

 

Thanks. I've seen the threads / comments on "dead zones," but I'd also like to understand more.

 

When we look into these things, we tend to [often] find that the thing being described is a proxy for a completely different thing.

 

For example, if you have seasonal subjects in your portfolio, the demand for those subjects will obviously go up and down over time, and there's a lot of competition for those subjects. For "Christmas ornament" for a search term, there are now 473,000 results, because it has become such a popular topic over the years. And demand for those subjects really peaks in the fall.

 

That being said - compare that to the search term "Instagram" Christmas ornament. I can't tell you how those would perform, but customers are increasingly coming to us asking for images that have an "Instagram" look. There are 81 images that meet that search term vs. 473,000 for the broader concept. That's just one example of a place to test the waters. I know many folks have been uploading "illustrative editorial" images, for example, which was a hole in the collection until we changed our policies.

 

If you're seeing something that you think is a weird in terms of sales behavior, there are many different variables that are involved. But there's definitely lots of room to optimize discoverability and fill content gaps.

 

I'm interested to hear more.

 

Best,

 

Scott

Share this post


Link to post
Share on other sites

Thanks for that big chunk of info up there. I'll comment on your initial post later - got a shooting coming up.

 

"Instagram" Christmas ornament.

 

However this raises an instant question: Wouldn't putting "instagram" into your keywords be a trademark violation and thus correctly result in a rejection?

Share this post


Link to post
Share on other sites

 

So, the question is not, "do SS shut down ports" but, "does SS ranking fluctuate the amount of time an individual port will be exposed for". If the time is increased and there is little sign of an increase in sales, they go down, if there is a jump in sales, ranking goes up. In this situation the amount of exposure would also be affected by the volume of contributors.

 

We don't do anything in search at the portfolio level. Our search engine is optimized for customer behavior and interactions as they relate to individual images and metadata such as keywords.

 

With 1 million customers and 40 million images, we have a lot of data about customer behavior, but then we also have a lot of data about data (if that makes sense). So, for example, we can know what images are being downloaded from what keywords, what keywords most commonly relate to each other across millions of images and pieces of metadata, or what the most likely intent is if a customer misspelled a search term based on prior misspellings of the same type - even at the language level.

 

There's a lot going on. I could probably call out more of the factors - but there are many. If you optimize for some of those 30-ish things, you'd be in a good position (the more that apply, the better). The ones that can drive customer behavior (styling and aesthetics, unique content, etc.) ultimately become the ones that drive search position, because customer behavior is an important part of search ranking. A good quantity of high-quality, highly relevant keywords is also a priority.

 

Best,

 

Scott

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this