Request demo
Try it now

Digital Asset Management in Higher Ed Webinar: The complete transcript featuring best practices and lessons learned by Corey Chimko of Cornell University

by ,

We held a webinar in early February about how to plan for, structure, and utilize a digital asset management system in higher education. As we posted before, the webinar featured best practices and lessons learned by Corey Chimko of Cornell University. 

Widen Digital Asset Management in Higher Ed Webinar Speakers
  • Nina Brakel-Schutt, Brand Strategist for Widen Enterprises along with Kara Bubb (project manager), Al Falaschi (video expert)  and Jake Athey (marketing manager)
  • Corey Chimko, Digital Resources Coordinator at Cornell University Photography
DAM in Higher Ed Webinar: The complete transcript. (1:12:12 hr/minutes)
Nina: Welcome everybody and thanks for coming to our webinar this morning, “Making DAM a success for higher education.  We hope that you're all excited to learn how to plan, structure and utilize Digital Asset Management solutions.  
We're going to be exploring this topic through the lens of Cornell University and with the help of Corey Chimko.
Corey Chimko: Hello, everybody.  
Nina: So our special guest is Corey.  He's coming from us today from Qatar, overseas, so thank so much, Corey.
Corey: My pleasure.
Nina: Yeah, he's the global DAM administrator for Cornell University and I'll be your host.  I'm Nina Brakel-Schutt; I am the brand strategist here at Widen.  Also from Widen we have Kara Bubb, she is one of our project managers and implementation specialists.  And we have Al Falaschi, he is our video expert.
Kara: Hello, everyone.
Nina: Oh, yeah, here's Kara.  So if you have questions this morning during the webinar, you can ask them in the question window in the side panel.  And we have Jake Athey, our marketing manager from Widen, he'll be gathering all the questions throughout the webinar and then we'll answer them at the same time at the end. 
In addition, Corey has generously offered to field any specific questions you have after the webinar and we'll be sharing his contact information with you so you can reach out to him directly.  And then lastly, we're also recording this webinar, so if you want to revisit any of the information when it's done, you can do that as well.
We thought it would be nice to share a snapshot of who's attending today's webinar, so when you registered we asked you a couple of questions.  Do you have a DAM system and what do you hope to learn during the webinar today?
So you can see it's pretty evenly divided who had a DAM system and who doesn't.  It was 51% that do, or sorry, that don't, and 49% that do.  And then as far as things that people wanted to learn, roughly it fell into four categories, system planning and setup, best practices for managing assets, structure categories and system administration, and then just a general overview of DAM systems and processes.  
So the content in the webinar has really been tailored to address the things that you wanted to learn, as well as to address just numerous questions we've been fielding from institutions over the last couple of months.
Real quickly before we get into the corporate presentation, we wanted to talk a little about who Widen is and where we're going.  Some people had questions about the future of our roadmap and things like that.  So we have two core services: Digital Asset Management and Premedia.
We are a digital media solutions company and the combination of these two services has given us kind of a unique niche in marketplace.  Our DAM system grew out of needs that we had from Premedia side.  We were creating all these assets and needed a way to centralize, and store and share them.
And so the combination of the two and innovations we've got going on both sides of our business have given us a nice comprehensive understanding of the asset lifecycle more or less from creation through optimization.
As far as our roadmap and where we're going, some of the things that are happening now were in the cloud.  We have a cloud-based only software service solution.  And some of the newer things that are happening, we've got some nice things going on with our API and our web services integration.  
We have a partnership with Digimarc to offer digital watermarking.  We've just launched a partnership with ConceptShare for some collaboration integration, and the Media Collective offers some really nice analytics around usage of your assets in the system in itself, and you can relay that to Google Analytics to get a holistic picture of how your assets are being used.
We've also got PDF exports going on.
And as far as things that are coming next, you should look for some nice enhancements in metadata.  We've got comments and ratings coming about in our latest version, which is 6.3 of the Media Collective in the next couple weeks here.
Collection history, more alerts and notifications, video advancements, mobile, of course, and then glacier storage, which is the longer term storage through Amazon Web Services.
So for today's agenda it's divided into three sections, our presentation.  The first section will address what's top of mind for 2013 in higher education in DAM.  The next section will be devoted to Cornell University and how they're using digital asset management to better their institutions, and then the third section will be tips and advice from Corey and that's where he's going to show some of his personal observations after being a global DAM administrator for some time.
So I mentioned, we've been fielding a number of questions from universities since probably early November of 2012 through the middle of January 2013.  And we thought it might be nice to house all these questions in a presentation for everyone at one time. 
Some common themes have emerged from all of the questions we've been receiving and these four things are really the common themes.  
So project scope...people have a lot of questions around you know, how big is a DAM project and what stakeholder groups do I really need to involve.  How many levels deep does it go within my organization as a higher education institution?  And how many departments should I include...things like that.
Access.  People want to understand who should be using their assets and how, and then what are the best practices around that? 
For video versus static images, a lot of questions around the raw data of video versus the edited data.  You know, what you should do for storing those and handling digital assets.  And then if you just have a number of static images that you're trying to use and manage your library how you do that.
And lastly, search and metadata.  We know there's higher institutions or higher education as a model is very complex and you have a lot of stakeholder groups that you have to search.  So being able to find that right assets at the right time is pretty critical.
So now I'm going to shift into having Corey speak to you directly about some things and Corey, maybe you could talk to us about some of the key considerations when it comes to project scope for digital asset management in higher education.
Corey: Sure, absolutely, thank you very much.  Thanks everybody for tuning in today.  I'm in Doha, Qatar for the time being, so it's great to be with everybody from so far away.  So yeah, we're going to talk about some higher ed DAM basics today...key considerations and things you have to think about when you're setting up a system in higher ed.  
And it's one of those things where luckily there are DAM systems out there now that are extremely versatile and basically let you set things up any way you want, so it's really important to do some planning in advance to understand the size and scope of both your institution as well as the collections that are in your institution.
You can see here, I've got a number of considerations that we do put down as points here.  
Higher ed collections tend to be huge and decentralized.  This is a big one.  I work for the photography department at Cornell and we're affiliated with the marketing section, division of the university.  And so the vast majority of what I manage are assets for marketing consideration as opposed to research collections, or museum collections or library collections or anything like that.
We do have some of those types of collection in our DAM.  And gain, most DAMs will allow you to manage all types of collections, but today we're probably going to be focusing more on the marketing side of things.
So for us, what we actually do is we shoot photographs and video for any unit on campus that comes to us and asks us for them.  So basically in our DAM system we have to have a way of categorizing essentially every department that exists at the university, as well as managing permissions for those people.  So as you can imagine, it's pretty big.
You'll see later when we have a look at DAM that there are actually over 600 at this point, departments in our, in our client list.  And you know, I would probably say that we're nowhere near actually having that complete at this point because there are a lot of people that we haven't shopped for.  So it can get pretty large.
And because of that you have to really understand and keep things simple right from the beginning.  So the key then is to start small.  We obviously, we started with only our photography collections from our department.  And then once we had those in the DAM and we understood you know, how to upload the metadata efficient, how to categorize efficiently, it wasn't until we had a good handle on that that we actually advertised the fact that we had this DAM system to the rest of the university and allowed people who were managing other collections around campus to use our DAM system to manage their collections as well.  
And then that way you know, when they have questions we could actually answer them because we had solved those issues for ourselves beforehand.  And again, in a higher ed environment you're probably, you're a lumper and not a splitter.  Basically, what that means is you're going to want to keep your category structures and your metadata list as succinct and simple as possible.
Again, because of the level of complexity in higher ed, the less time you spend assigning your metadata and categorizing your assets, the more material you're going to be able to get to.
And just to give you an idea, we view about between 1,200-1,300 shoots a year in our studio.  And if you break that down by workday and you're looking at 67 shoots a day that I have to get through, upload, tag, keyword, etc., so it's really important for me to have a system that's extremely efficient in order to get through that volume of data.  Can go to the next slide.
Nina: Thank you, yeah, Corey, that was huge, thank you.
Corey: Okay.  Yeah, you're welcome.  All right, department vs. university-wide, this is another issue before you want to set your DAM up you're going to want to have a good handle on whether this is going to be a university-wide system where you're going to have a lot of open access for just about anybody in the university, or whether you're going to have things kind of silo'd where only certain departments are going to have access to their own assets and then nobody else is going to be able to view their collections.
In reality it's probably going to be a mixture of the two.  We, as I said, we were part of the marketing and communications department for the university, so we're at the university administration level, which basically means the people me in my division want to have access to assets across the entire university.
But there are also marketing and communications departments within the individual college schools that are separate from us who really only have an interest in assets that pertain to their own college or school.
So you kind of have to have that versatility to allow some staff members to access everything well, keeping certain assets from other people who don't want to see them, because the people, you might think that you're being very beneficial by giving all of these assets to everyone, but at the end of the day those people see those extra assets as chat to be waded through and they don't want to have to deal with any assets that aren't pertaining to their specific department or unit.
So it's good to have a handle on that upfront.  Obviously, managing that and what's your relationship going to be internal vs. external.  So is your DAM system going to be for the use of only the marketing department?  Is it going to be for the use of the entire university?  Is it only going to be staff?  Is it going to be staff and students?  Or is it going to be staff, students and alumni maybe?  Or are you going to actually have public access to your images?
And again, it's probably going to be a combination of some of those things and so again, understanding exactly who you want to see what and will enable you to actually structure your commissions such that those goals can actually be achieved.
For us at Cornell, we basically have different levels of assets.  We have what we call our  optimized assets and our unoptimized assets.  And basically the marketers have access to everything and the public only access to the optimized assets.  
So basically, the public uses our best of assets, the ones that have been in publication, the ones that have been properly optimized in Photoshop and so on, so they see the really polished stuff.  
And then the marketers, they want access to everything so that they can, you know, if they find a photo and it's not quite what they're looking for for their project, the could go to the rest of that shoot and they can see all the outtakes.  And then the can pick from those if they decide to.
And then e-commerce is another consideration.  Right now I don't believe that Widen has an e-commerce solution, but it may be something that you want to think about in terms of your DAM.  For us, we do, we are a cost recovery unit at our university, which means we charge for our services, so there are certain cases where we do sell our photos for money to either the public or to people within Cornell, things like graduation photos, you know, photos of major events that occur on campus.
And sometimes people will be retiring and we'll sell you know, prints of the university.  And we do calendars.  We do a number of products that we sell also, so that's another thing to consider, whether or not you're going to be trying to monetize those assets directly because that actually opens up a whole new can of worms in terms of you know, privacy and security on your website.  
Okay, so our
Nina: Yeah, go ahead and talk about metadata, Corey, thank you.
Corey: Sure, sure, yeah, when you're tagging assets, again, you really want to avoid complexity, as I said.  The simpler you are the easier it's going to be for you to manage large amounts of data.  
It is however very important to be accurate.  And it's a trick sometimes, especially in higher ed you have to be accurate because of things like, if you look at no. 3 on the slide there, you may have a donor, for example, meaning Joe Clark, in this instance.  And somebody types that into your search.  Well they're going to get portraits of Joe Clark.  They're going to get Joe Clark Hall.  They're going to get every event that's ever taken place in Joe Clark Hall.  They're going to get the Joe Clark Memorial Lecture Series, etc.
So how do you divide things to make sure that people are actually getting what they need.  And this is where the categorization and the filtering of metadata really becomes critical. 
You're going to need a system that is versatile enough to allow you to search through all the metadata fields, specifically.  And Widen's solution to this is really quite fantastic.  Not only do they have filters--you can use filters, you can use categories, but basically the way their system works is you can setup every metadata field that you have as a metadata filter.
So if you do a broad search and you get a bunch of stuff that you don't want, you can then search within that search or you can use, you can filter it by any type of metadata.  So if you have you know, a buildings metadata field, if you have only the building shots, if you have faculty or alumni portrait category, you can filter by that and only the portraits of the person.
So that's really, really important to have versatile metadata systems that accommodate those types of differences because again, with the complexity of things in higher ed.  It's pretty rare where somebody just types in a general search term and they're going to come up with exactly what they were looking for, especially if they've got a large collection like ours, which is in the neighborhood of right about 300,000-400,000 assets at the moment and it grows all the time.
So let's go ahead to the next slide.  This actually kind of ties into some of the stuff, let's see here.  We're going to have a little bit more later about how we've structured our specific DAM in order to accommodate, in order to standardize the metadata fields and make sure that information is being entered accurately.
Basically you want to try and minimize the amount of manual entry that you're doing when it comes to metadata.  We'll talk a little bit more about that later.  
These are some other issues.  Obviously, people we're talking about have big questions about video.  I'll be perfectly honest with you, at Cornell we haven't really solved this problem yet and I think one of the reasons that everyone is asking so much it is because it's not an easy thing to solve. 
Obviously, the cost of storing raw video is enormous and at Cornell we don't even have the means yet to store our raw video offline or out of the cloud, let alone in the cloud itself.  
So what we've done is basically, adopted a policy where what goes into our DAM is our finished products only and then if people want to access the raw footage, they basically, get in contact with us.  We are in the process of researching a large scaled storage system that will serve if not just for our division, possibly even for the whole university, something like an Isilon system or something like that.
And then you know, we're just going to piggyback on whatever storage solution they come up with for video, but right now video is a big issue, obviously.  A great thing about Widen is their ability to transcode on the fly.  Again, because you want to, you want to upload either your high res video or your best, the best version of the finished product that you have.  
And then you can send that out at different sizes for whatever it is that you need.  Or you can directly use the embed codes to pull those videos from Widen and display them on your webpages through the API.  That's great because then you don't have to host all of those gigantic videos on your own servers, which can really help a lot.
Let's see what else can we talk about here, handling large numbers of photos and assets.  This is probably, if you're in higher ed you're going to have to deal with this.  One of the things that we do is basically we do our tagging in bulk.  
I basically approach each shoot that I do separately and I attempt to categorize and to keyword the entire shoot together.  Sometimes--it works really well for like portrait shoots and shoots where you have homogeneous subject matter in the shoot.  It works well for things like events where you have different people in every shot and they have different locations and so on.
So you get a wide range of stuff.  And basically what's going to happen is there are going to be some shoots where you're going to be able to get through in five minutes.  And there's going to be other shoots that are going to take you a half an hour, an hour to shoot or to keyword.  And that's just, that just goes with the territory.
And then you're going to have to decide what kind of documents you want to ingest into your DAM system.  So basically, any decent DAM system is going to have the ability to adjust pretty much any electronic file type.
The vast majority probably of what you're going to be ingesting is photographs and video, possibly documents, possibly things like you know, PDF web presentations, PowerPoint presentations, that sort of thing.
Widen's system has the ability to adjust all of those things.  And then you're also going to have to decide okay, and are we going to use the system to basically manage our workflow, and by that I mean are we going to be ingesting assets that are not yet optimized, that are maybe parts of projects that are not yet finalized...are we going to be using it in our workflow to get things like approvals.  
Or is this basically going to be a showcase of our work?  Are we going to upload only finished projects, only polished assets, only final videos?   All of these are basic considerations that you're going to have to solve for your own university because you know, according to what your goal is for the system.  I've talked with some people who are very comfortable with simply having a place where the public and the university can go to basically get a good idea of what types of images are available.
And then they're perfectly content to have those people call them and request what they need.  Other people want everything to be completely automated, from ingestion, to access, to workflow.  So you know, you've got to decide for yourselves where you want to fit in that timeline.  Let's go onto the next one.
Nina: Yeah, thank you, Corey.  Now we're going to get into kind of the real nitty gritty about Cornell and their DAM, and Corey is going to share real specific things with you about how they planned and structured their DAM system.
Corey: Sure, so this first one is probably the very, well, probably the most important one of all.  And that is you need to talk with your users about what they want, what they need, what they expect to get out of your DAM system.  One of my biggest rude awakenings as a digital asset manager was just realizing how differently other people conceive of our assets and what they need to use them for.
A lot of asset managers tend to be very detail oriented, very analytical, very structure based.  And when you are interfacing with marketing or design areas of your organization, they tend to think in much different terms of much more artistic and creative terms.  And they may not be searching for something extremely specific when they come into your DAM. 
They may want to look for some sort of nebulous thing like excellence in education, or diversity, or some sort of marketing priority or term that they're hoping to get results from and you haven't categorized any of your assets that way.
So it's really important to understand how your users are searching and then incorporate that structure into your DAM from the start.  Much, much easier to do it at the beginning than it is to sort of get everything in there and then adapt it afterwards.  
I can tell you that just because I've had to do it both ways.  And do it from the beginning is much, much easier.  
Basically, when I was hired for Cornell I was handling an existing system that didn't work very well for what people needed it for.  We had about six years of digital media, basically just photography.  So you think out of 1200 shoots a year, probably somewhere in the neighborhood of 800-1,000 shoots with I don't know, 30-60 images per shoot.  So a lot of that would get in there.
One thing that is really great with Widen is that they, if you're looking to have a system with them is, and we're going to show you this in a minute, but they have something called a, their, I think it's called a setup document where they actually take you through step by step what you need to do to get a really smooth installation going.
So basically it, there you go, a site setup checklist.  So for those of you who are interested in the real nuts and bolts of you know, from when your decision to get a DAM system all the way up to implementation, these are basically the steps that Widen will guide you through.
And following these steps gets your assets in line, gets them ready to be ingested in a way that's going to be good for you.
Let's go back to the previous slide.  One more.  There we go.  Okay, so yeah and then your timelines.  So when we went with Widen I was actually very surprised at how quickly things went.  It was probably somewhat due to the fact that we actually already had a relatively organized system for our assets.  We already had a good numbering system.  We already had things on CDs, by shoot, by date.  So we had a constructed system that we could basically just move, replicate in the DAM and then move the assets in very quickly.
The total time it took was about three months.  So if you look on that site setup sheet, I think 99 days is the total number that they have there if you stay on your timeline.  And ours was actually a little bit faster than that because we were able to skip over certain parts of this setup just because we'd already done them or you know, there were certain features that we didn't choose to add to our DAM, so all those steps were not needed.
Let's go on to the next slide.  Okay, so how did we structure the system?  How did we come up with our categorization structure?  What our metadata fields were going to be?  How did we decide on what the value lists for those were going to be?
The first thing that we did was we created our existing organization structure, category structure in the DAM.  So basically we had done it by data.  Each shoot that we do has a shoot number, so we just basically did it by year.  We have what's called a chronological list of our shoots, and then it's divided by year and then by asset, by shoot number.
So anyone at any time that's familiar with our system can go into that list and access any shoot if they know the shoot number, very simple.  And we're going to have a look at that in some of the next slides here.
One thing that I found--that's okay, we'll just go through the three of these and then we'll hit the slides and show everyone.
Another thing that I came across was that when I ingested all of our assets off of the CDs, I found--obviously, all of those are the original, unoptimized assets, but we also had huge drives full of optimized assets.
These are the ones that have actually been chosen for publication, repurposed in Photoshop.  The ones that we actually had a high res that were really nice marquis assets, those were not in our structure.  They were basically just haphazardly laying about on external hard drives.  
So basically what we had to do was structure those assets as well into some sort of you know, structure, so that we could get those into the DAM.  That was very, very difficult, but it was very, very important because those were actually the assets we wanted to see.  And again, one of the great things about the Widen DAM and any other DAM that you might want to get is probably going to have this, is they have what's called a versioning system.
So if you ingest say a low resolution copy or an original copy of something and then you have a better high res optimized version, you can actually import that new version and apply all the metadata and so forth to it automatically without having to redo anything.  So you don't have to reenter the metadata or any of that.  You just create it as a version and it'll automatically categorize it, add the metadata that same as the original asset.
So that saved us incredible amounts of time because what we could actually do was just take those drives and drag and drop them.  And the system itself would tell us if that asset was already in there and it allows you to compare the existing version and the new version, and then you simply pick what you want to do.  Do you want to delete the old version.  Do you want to add the version as an asset.  You know, it gives you all these fantastic options.
So that is an incredible timesaver for us in terms of the Widen system.  And that sort of informs that second point there, which is the lack of closeout policies in our marketing department and then I guess this is going to depend on your organization, but what I find in our DAM is that people are very eager to logon and find assets for their projects.  
They're not nearly as good about sending back the finished product to be ingested into the system at the end.  So what I ended up having to do was creating workflows where I got access to the final project folders for the marketing department and I go in periodically and pull any new optimized photos, any finalized video of projects that were in there that should go in the DAM that nobody because really I'm not in the marketing group.  I work in the photography department.
So there was no way for me to get access to those final products unless they were actually given to me or I went and found them myself, so that's something to think about.  It's very easy to just take everything off your CDs and your drives and put it in the DAM, but what you, if you really want a DAM that tracks your history, categorizes your products, showcases your marketing efforts, you're going to want to make sure that you get your finished products in there and that you are uploading your optimized versions and so on.  And that can be a little bit different than what a lot of asset managers are used to.
And then the last thing on the slide is the single sign-off option.  This is again, something that a lot of DAM systems offer and it's basically a way for people to log onto the system the same way that they logon for anything else that they have access to at the university.  
So most of your universities are going to have a net ID system where you have an authentication program, like [inaudible 32:52], where people just enter the net ID and password, and then they get access to the university websites and services.
Well, Widen and I'm assuming some other DAM companies will largely set things up so people can use their existing net IDs to log into your system.  They don't have to have a new set of user names and passwords to remember, which is fantastic.  It's also a huge time saver in terms of sorting out who's going to have access to what.  
So basically when somebody registers with their net ID, the system is going to pull their directory information from your university directory, so you're going to be able to tell who they are, where they work in the university, and then that will allow you to very easily manage their permissions and so on.
So now we're going to show you slides of our actual DAM and how those things have been implemented in it.  Here's our single signon page, so you can see basically the first link at the top there is where Cornell users can login with their internet ID.  
And then the second section at the bottom is where we can also get access for people who don't have a Cornell net ID.  So if you're a member of the public or maybe you're an alumni that doesn't have an ID anymore or whatever, you can register at a non Cornell user, create your own username and password, and all that's done automatically.  The DAM doesn't have to be involved in this process of registering the users at all.  It's fantastic.
Go to the next one.  Okay, so here is a snapshot of the category structure for the chronological list of shoots that I mentioned.  So we have our top category, which is the chronological list of shoots, then we have it done by year.  We have under 2010s, 2010-2013 and then again, because we do so many shoots, I have it further divided into groups of 100 shoots.
So you can just see basically how each one of the bottom level there is a shoot.  We put in the shoot number at the beginning, the 13 for the first of the year.  Then we have a set of, these are the taxonomy tags for the units at Cornell.  So AAD is Alumni Affairs and Development, PAFF is Public Affairs and so on.  And then a brief description of what types of photos are in that shoot.  
So that's basically the simple structure that we migrated over from our file system as we started. 
And then on the next slide you'll see we actually employ a number of different ways, category structures, ways to organize in the system because I told you earlier about how the marketing people were thinking of things differently.  
So the number system is great for the photo people, for us, because we know about the shoot numbers.  
Well, those numbers really don't mean anything to anybody else except for our department.  What most people are looking for are images that have something to do with their department or their unit, so we also have what's called the academic units category.  This is just done alphabetically.
So you can see we're under A here.  And it's basically one gigantic listing of every department's program, college, unit, whatever that we have on campus in a giant alphabetical list.
And the reason that we did it this way is because Cornell obviously has an extremely large and complex administrative hierarchy.  And a lot of times we'd find that people were coming in and they would be looking for images from some department that was buried five levels deep in a college and they had no idea where that department actually sat in the administrative hierarchy, so there was no way for them to really--so if we'd only set things up, if we had nested everything in its hierarchy, people would never have been able to find what they were looking for.
So basically we eliminated the administrative hierarchy and put everything in alphabetical.  And because the DAM system allows you to place assets in more than one category, you can actually have it in the college category and you can also have it in the department category or the program category.
So this actually worked out really well and it's much, much easier for people to find what they're looking for because they just go to the name of the department that they're looking for.  They don't have to know what college it's in, what department it's in, etc.  So this is another possibility you know, if you don't want to set your DAM up at all by a numbering system, you just want to do it by college or unit of your school, you can do that as well.
And then this is the third type of organization that we do and this is by marketing priority.  So again, our primary client is the marketing department at Cornell.  We went to them and said well, how do you want us to organize the assets?  And they said here's our five year marketing strategic plan.  These are the major topics that we're marketing over the next five years.  You can see those ones at the top there for 2010 and 2013.
And then they also gave us a list of other priorities that are things that just in general we tend to emphasize in our marketing materials.  So now when I'm ingesting assets I also categorize them according to marketing strategy and that's very, very helpful for the marketing people.  
Nina: Okay, Corey, I'm going to move us along just a little bit here and I'm hoping you can tell us a little bit about how you approach your metadata and search for Cornell.
Corey: Sure.  So no. 1 is you want to get what I call official data.  And you may have to do some digging for this because as with any work culture, people use acronyms, people use abbreviations, people use nicknames, people use all of this language that is not really true official language.  It's not the language that's going to appear on your websites or in your publications.  
So what you want is to get an official list of data, so for example, when I was putting together that list of department names I actually went to the university administration office and said can I get a list of the official names of all the departments and that's what I put in.
You can add your folksonomies later because obviously it can be useful sometime to incorporate those acronyms and nicknames so it's easier for people to find things, but you don't want to organize according to those because they're not going to be used universally.  So it's something that you can basically add later.
Another great thing as I mentioned before, Widen allow you to setup all your metadata fields as filter searches.  And what that means is whenever you do a search in Widen, let's say you get a thousand assets returned.  On the left side you're going to get a filter for every, every metadata field that you have that you're then going to be able to further narrow your search by.
And so the more fields that you set up as filters, the more accurate you're going to be able to get when you're searching for things.  And Widen will actually tell you how many assets you have in each one of those filter categories, which is fantastic, so you can see where the majority of your assets lie. 
How many metadata fields should you have?   This is a big question.  It's going to depend basically on the nature of your collections.  You may want to setup more than one metadata type, so for example, if you are primarily doing video, you may have a different set of metadata fields than if you're doing photography.
Or you may have different collections.  So if you're having, if you have research collections you may want to try different information than if you have primarily marketing collections.  
So we actually have a number of different metadata types in our DAM.  So for example, the vet college likes to track things like you know, animal type, and species, and disease, things like that that don't have anything to do with our photographs and photos.  So you can setup separate metadata types for those types of things.  
But basically there's going to be a number of fields that are basically universal, like your creation date, your description, your creator, whether that's the photographer, the videographer or the person that writes the document.
And then you're going to want to branch that out to information that's important for you, important for you and your systems.  Most of the time the sweet spot that I've heard is anywhere between about 20 and 30 fields.  You probably don't really want to go over that because you're going to be spending an inordinate amount of time actually filling all that data out for your assets.
Cornell has 24 fields and we're going to have a look at them in a little bit.  And that actually, we started with a little bit more than that.  We actually started with 31 or 32, but managed to whittle that down just because you know, it was taking a long time to add all that.
First item on this slide, also extremely important -- avoid the use of manual entry fields and use standardized pick lists for metadata entry.  This is how you make sure that everybody is entering the data in the same way.  You don't allow them to type it in.  You don't allow them to misspell words.  You don't allow them to use acronyms.  You don't allow them to use nicknames.  
You put everything in a drop down list that they have to pick from.  Again, you're going to use your standardized vocabulary and you're going to use your official documentation lists to make sure that you have all the names and so on correct.
One of the great things again that Widen has is that for these fields you can actually start typing the name of what you're looking for in and it will whittle down your field lists to only items that contain what you're typing in.  So even if it's not first word in the entry, you're still going to be able to find it very easily and that's extremely helpful, because again, somebody might call it something differently than you do.
You know, somebody might call it the office of diversity, whereas another one might the diversity office. 
If you type, just type in diversity you're going to get both of those entries and you can pick which one is there.
Avoid the use of unexplained the acronyms.  That's just another facet of this.  We did develop a use of a Cornell taxonomy with "official" tags or acronyms for each department.  It's okay to use, it's okay to put an acronym in there, but what you might want to do is put the acronym in there without actually writing it out and explaining what that means because--and that comes into the third point, which is...
You have to enter your metadata as if it's being searched by someone who has no familiarity with your institution.  Obviously, that's especially important if the public is going to have access, but even within a large university there are so many people that work at the university that have no idea about the language or structure of the university outside of the department that they work in.
So they may not know how to search for certain things outside their sphere of influence.  So you really need to avoid the use of things that only a small number of people are going to know what they mean.
So this gives you a peek into where you create the metadata value lists in Widen.  It's very easy, you just type them in on the left and add.  Widen has this great little sort alphabetically button, so this list is actually about 600 units long, but you can see where basically I'm typing in the full name of the department or programmer unit and then I have its abbreviation--you could call it an acronym or taxonomy tag or whatever you want--in brackets.
So let's go ahead to the next one.  
This shows you how to set your metadata fields up as filters, so you can see over on the left for each field you can choose field type.  This particular one is a palette field.  Some people call that the select field.  Basically what that means is you can make more than one selection.  
So sometimes we, for example, might shoot an event--it's a collaborative events between more than one unit of the university, so you actually add more than one unit to that shoot.
And then on the righthand side, these are all the options for your fields.  So you know, do you want people to be able to see this field?  Do you want it to be editable?  Is it an acquired field?  
Quick search, Cornell has--I'm sorry--Widen has a number of ways to search, quick search being one of them that's very useful if you want to do complicated searches.  You can add that there. 
Do you want to be able to search your results by this field?  Also extremely useful, especially for things like dates.  And do you want this to be available as a search filter?  And basically I have all of my metadata fields set as filters, except for you know, ones that don't make sense.
And here is our list of 24 metadata fields.  The ones that have UP in front of them are specific to university photography and a couple appear at the top, description creation date, that are basically universal.  So they don't have the U-photo prefix there.
But this for, for photography this is probably basically what you're going to want.  Your content type, your description, your creation date and so on, I'm not going to go through them all, we're a little pressed for time.  
But some things to draw your attention to, there are fields here for whether or not you need release forms.  Can you just go back for a second?  And also for things like you know, associated with projects or publications, that stuff is down at the bottom...your copyright information and so on.
And another time saver, Widen and other systems will allow you to do what's called metadata mapping.  If you are tagging your assets outside of the system itself, so for example, we tag a lot of our images in Adobe Bridge.  You can see at the top there on the lefthand side, this is the Adobe Bridge metadata panel where you can enter things like your description, your keywords, your photographer and so on.  
And then you can map these bridge fields to the metadata fields that you create in your DAM system.  And if you're the only one importing, this can be extremely useful because you can do all your metadata entry, you can embed all that metadata outside of the system in bridge on your desktop, and then you can just drag and drop your assets into your DAM and all of that information will be automatically populated to your metadata fields, which is great.
Of course, you have to make sure you have the same pick lists in your bridge as you do in live.  
And then here we can just go--did you have something there, Nina, you wanted to say? 
Nina: No, I was just going to say this is really the last section for you know, specific to Cornell and after that how you migrated your assets over initially, so kind of disparate drives into the DAM system.
Corey: Right, so that's, I mean everyone I talk to has this issue where they you know, we've got 16 different drives, and then we've got all these drives on machines, external drives and so on, and how do we get everything together?  How do we make sure we're not being redundant.  
So basically the way we did it was we had our six years of CDs and then we had, all the photographers had their own hard drives.  And the first thing that we did is we migrated the digital images off the CDs and onto what we're using, which is basically rate[?] storage, network attached storage divorces.
We had everything in the hierarchy on our file system and then you just basically go through them.  Because they were already all numbered it was relatively easy to drag and drop things all into the same folder and find out you know, when we had duplicates.
Widen also offers a deduping service when they create your DAM system, so that's, they can actually dedupe them for you, which is great.  And obviously what you want to do is you want to keep the largest, best version of each asset that you have.  So Widen works on the premise where you upload one high resolution master asset and then you can then resize on the fly.  
You can deliver different sizes of that asset and the system itself will do those conversions for you and send out the product, so you don't have to have six different sizes of the same asset in there.  You only have to have one large master asset. 
So basically that's how we did it.  We started, we adjusted the basic version of every asset and then we, then we took the optimized assets and drag and dropped those and we waited for the system to show us the conflicts.  And then you go through the conflicts one by one and decide which one is better.
It takes a little bit of time depending on how much you have, but the end result is that you have the best version of every photo in the DAM system, which is obviously what you want.  
And then in terms of metadata, basically we embedded what I would call the basic metadata for all of the assets.  We did this in bridge before we uploaded them into Widen.  And by basic metadata I just mean the file number, the photographer, the shoot location, things that are basically going to be the same for every image in that shoot.  
What we did not embed were things that are going to be different for each image.  So things like the description, primarily the description and the keywords were the things that were not--because they're going to be different for each image.
And then obviously if you were going to do that beforehand it would take years, at least it would for us in order to get that information in there.
So we got the basic information in so that we can actually search and locate.  Now, most of my job is going back and filling in all that missing metadata.  
Upload your assets in a measured and controlled way.  So basically we just started with the most recent assets and we worked our way backwards.  Obviously, because the most recent ones are the ones that most people are interested in.  You know, and we do go back 10 years about, so you know, the longer we went, the more we had and the deeper our collection got.
And then on the side you can just see what I already said, which is that you want to apply the metadata to all these assets and then you're going to have to go back and fill in the blanks later.
Nina: Okay, thank you so much, Corey.  We're getting into the final section of the webinar now and we're nearing the end of our time, so I'm going to let Corey kind of breeze through these and then we're going to take a couple minutes for questions at the end.
Corey: Sure, so yeah, basically, tips and advice.  Do it right the first time.  It's very difficult to, when you have all these assets in there and you want to apply a new metadata field or you want to organize things in a different way, basically every time you make a change you have to go back and apply that change to every asset you have in the system.
And if you have tens or thousands of assets in there, it's almost an impossible task to actually do that.  So you really want to make sure that everything in there from the beginning, all your metadata fields and you know, that way you won't have to waste your time later on.
Obviously, it's going to be easier to keep things organized.  
Assign an admin early in the process.  Yeah, that's obviously key.  This is not an easy thing to get over in the higher ed world with budgets and so on, but if you really want your DAM system to work, you need to hire a full time DAM person.  
I don't know how many times I've heard someone say oh, well, we'll just get a DAM system and then we'll let the summer interns tag off the photos and whatever.  That doesn't work.  You need someone who's familiar with the technology.  You need someone who is familiar with your institution and the administrative hierarchy of that system so they can properly administer the permissions and so on.
You need to know somebody who can be a technical liaison for support.  You need somebody who obviously has good time management skills and attention to detail because there are so many little nuances that are going to come up.  And your average person is not going to have any idea how those things differ unless you're a full time employee and you know your administration inside and out.  So that's very, very important.
Important of user governance.  Again, single sign-on is great for this.  And assigning permissions is just basically going to be something that you have to work out before you start because you're going to want to create your rolls and your permissions at the very outset.
Most users are not going to have access to everything.  In fact, as the admin you're probably going to be the only one that can see everything in your DAM, so you need to understand again, who the people are, what they should and shouldn't have access to.  And a good admin is one that can actually deal with the complexity, but hide that complexity from the user.
So if you have Joe Shmoe from Dept. X, he logs in.  What you want to present to him are things that are going to be of importance to him and without bogging him down with all the other stuff, and without making his life complicated.
Last but not least, again, we return to the issue of storing raw data.  Right now, Cornell, what we do, we store all of our raw assets offline on rate[?] storage and we upload low resolution jpeg previews into our system so that people can access to everything, but we don't have to pay exorbitant storage fees.
But then we do upload anytime an image actually gets used or optimized, that master version goes into the system as well.
Storing raw data in the cloud right now is basically a pipe dream for a lot of people unless you're a gigantic broadcasting corporation or something.  Almost nobody does that.  As I said earlier, we haven't really solved this issue yet either. 
Right now we're just storing our raw data again on network attached storage with backups.  It's basically what you can do.
People always say oh, we shoot as much as we want, the cost of storage keeps going down.  Yes, that's true, but file sizes also keep going up.  So if you're a photographer, you know that in the last five years or so the size of raw images has tripled.  So even though your cost is going down, the amount of storage you need is going up.
I recommend photography edits as much as possible, although it may be next to impossible to get your photographer to actually do such a thing, which has been my experience.
But obviously there are a lot of photos taken that are never going to be used.  You don't want to put up the portraits of people with their eyes closed, for example.  So we have a lot of images that end up in our system that will probably never be used and it's simply because we don't have the time to weed them out.  So more careful shooting and better edits is also going to end in a more succinct system for you.
That next point I already explained.  And then yeah, for video assets, as I said, we're only uploading our finished products at the moment.
Nina: Thank you so much, Corey, and you know, we have a couple minutes to answer some additional questions, but I wanted to let everybody know again, this recording is going to be made available to you as well as the slide deck.  We're going to share Corey's contact information with you in just a minute here and he's also provided some links to some nice additional information for you about best practices and taxonomy at Cornell.
I'm also going to give you additional information so you can contact DAM advisors at Widen if you would like to further a conversation with us.  But I wanted to open it up to any questions that people might have.  
Corey: Sorry.
Nina: No, go ahead, Corey.
Corey: Why don't we put up that last slide there while we're doing the questions.  Yeah, so there's my email address.  Please feel free to email me with any more specific questions you might have.  There's a link there to a little taxonomy and best practices page I have on our Cornell website.  And then the bottom is the link to our DAM system.
And what I wanted to say was any of you can go to our DAM system and register as a browser to check the system out.  And if you are more technical minded and you really want to get into the nitty gritty, I can always up your permission so you can see some of the behind the scenes stuff if you're curious about the Widen system.
Jake: Thanks, Corey, this is Jake just with a couple quick questions from the attendees here.  The first one is what metadata do you add before import through bridge?  What fields?
Corey: When we first started I actually added everything except for our marketing priorities.  Basically, anything except for the fields that we multi select fields because I could only add one value to those fields.  So basically I was doing the description, the tags, the photographer, the copyright information, the rights usage, as much as I could.
That was at the point when I was the only one importing assets into the system.  And it was also before Widen had something called the upload wizard.  The upload wizard came out with the last release and it, what it is is it's workflow within the system itself.  It actually allows you to add all that metadata as you import assets.  
And we also have other people from around our universities start uploading assets for themselves as well, and they were not familiar with the mapping structure between the bridge fields and the Widen fields.  
So basically what I had to do was stop mapping a lot of those fields.  Right now I'm only mapping the manual entry fields, which are the description and the tags field.  And the rest of them I'm adding during the upload process using the upload wizard.  And the reason for that is because for those, those are the pick list fields, those are the standardized fields, so anyone can upload assets using upload wizard.  And they will be able to get the correct metadata in there as opposed to having to know in advance what those entries are in the lists in order to do it in bridge.
So it all depends on, how much you map is going to depend on you know, how many cooks you've got making your soup.  If you're the only person, you can make it really fantastic and you can have all your fields, and it's going to be fantastic.
If there's more than one person doing it and you need to be concerned about the standardization of the entries, then you're going to be able to map less.
Kara: Hi, this is Kara from Widen.  Just to add to what Corey said and actually a little background, I actually worked with Corey on the Widen side to get him up and running with his system.  But you actually can map now multiple fields on bridge.
Corey: Oh, yeah?
Kara: Yeah, you can do multiple fields now.
Corey: Ah, there you go.  
Kara: Yeah, there you go.
Corey: Fantastic.
Jake: Thanks, Corey, and thanks Kara.  One more question, a hot topic right now is the lifecycle of asset management, particularly when it comes to archiving images.  What are your plans and do you have plans for archiving some of your older images and what do you use as a database for archival?
Corey: Right now we are just at the threshold of dealing with that issue.  Historically when we weren't shooting digital, when we were still shooting film, we kept at our office, we kept 10 years of assets and anything older than that we would ship to the university archives, which is in the library and they would take care of that.
Right now we're at the cusp of where we've not shot 10 years of digital and our first digital images we did in you know, ML3, so we now have to decide whether we're going to be the custodians of all of our material going forward or if we're going to be transferring digital data to university archives.  
That depends on whether or not the archive has the ability to ingest that data and make it easily accessible.  Right now they don't, so it looks like we're going to be maintaining our own archives for a while.
Within your DAM system there are a number of ways to approach this.  Widen has the ability to do what's called expiring your assets, which basically means that if you expire an asset, but only admin can view that asset.  And it doesn't show up in search results for people, so basically it isn't visible to anyone except for people who you give permission to see them.
So that's one way to get rid of old images cluttering up your server results.  Another thing that we do is that we use the starring system in Widen, which is a way of having particular assets flown up to the top of your search results regardless of how you select them.
And basically we use that for our best images, so basically you know, if you had images that you still wanted to be searchable, but you don't want them showing up in those sort of the marquis area, you can unstar them and they'll go back down to wherever it is that you sort.
So there are a number of ways to do that.  Also, there is one feature and maybe Kara or someone can speak about this, there's an archive feature that goes along with the versioning process.  And I haven't really used it, so I don't know what it does, but maybe you can talk about that.  
Nina: Yeah, absolutely and I'm sure we could include more information with well, but essentially it makes it unavailable for ordering, but still maintains that asset on the site.  So you would have the ability to access it from within the DAM, but it's no longer available to be order.
Corey: Okay, so there you go.  And then what about this, you mentioned on the first 
slide today something about glacier storage. 
Jake: Yeah, that is something that we're looking at as part of the long-term evolution of the product, just knowing that people have that concern for the lifecycle management.  Not available and no timeline to it, but we are investigating. 
Corey: And how exactly does that work?  
Jake: Well, it's long-term cold storage.  More cost effective storage for the long-term preservation of your digital assets. 
Corey: Okay, so there you go.  And yeah, I mean that's something that you know, everyone is going to have to tackle this issue of whether those older assets have to be available like now or if you can afford to put them in "cold storage" where it may take longer to retrieve them, but you're going to save on you know, on your storage costs and so on.
Nina: Right, right.  Is that the last question?
Jake: That's essentially it.  We've got one more question; the question about ROI is a good one.  And the question is were you charged with demonstrating potential ROI to the university when working through the DAM selection process and what did that look like?
Corey: Yeah, we definitely were, although I have to say that ROI when it comes to DAM is a little bit unorthodox.  It's sometimes difficult to actually assign a dollar value to the savings that DAM provides you, but we were able to do things like show how we could do away with inconveniences like having people physically come to our offices and search through CDs.  
We could do something like okay, it took you know, somebody the inconvenience of physically having to come to our office and spending an hour searching un-keyworded images on CDs, whereas now they can have a targeted search by the same asset in five minutes.  So then you can take that you know, an hour down to five minutes as a savings, as some sort of return on investment. 
You can also show things like you know, we can only, a person can only go through this amount of assets in a particular time period if they're on CD, but if you have them on a searchable database we can extend the number of assets that a person can search through almost exponentially because it's no longer a time issue.
So you can use things like that.  Now, I suppose you could quantify that in a dollar amount if you, you know, then assign dollars to that time or whatever.  We didn't quite go that deep.  Basically, for our department it was already self-evident that this was just going to save us a lot of time and money and they were pretty much gung ho at the beginning.
I am in a situation now where I get asked for stats on a periodical basis and Widen has provided a whole new set of stats tools from the last release which are great, but it's still a difficult thing to apply. 
So for example, I can go to my superiors and say we have this number of image downloads in the last quarter, but I can't tell them how many of those images were actually used in publication or on a website.  So it's difficult.  You have to, you have to come up with formulas and you have to use averages and so on. 
I think that, was there a Widen--I don't know if it was a white paper or a PDF or something that we sent out that originally had some really good suggestions for how to communicate ROI?
Jake: Yes, absolutely, we do have an ROI section on our website.  And you're right, Corey, there is no clear-cut way for calculating that, but there are a lot of different things to consider.  We do have a slide deck we can make available to share some ROI models that other customers have put together industries, something we're happy to share.
Corey: Right, I mean it's very, very difficult to assign a particular dollar value to an asset because you know, you can never tell what asset is going to resonate with somebody and we have certain assets that are you know, 12 years old and people still come to us asking for those specific images over, and over and over again.
And then we have other assets that the photographers think are absolutely fantastic and nobody ever uses them.  And so you're you know, your ROI then is obviously that older asset is more valuable to you, even though you know, at face value you wouldn't think so.  So it's a tough thing.
Nina: Well, this has been really great, Corey.  We can't thank you enough for being our special guest on today's webinar and we wanted to thank everyone who was able to join and participate today.  Like we said, we're going to make this recording available to you as well as the slide deck.  And we hope that all the information proves helpful to everyone.  Thank you. 
Corey: Thank you very much, everybody.  

comments powered by Disqus

The Widen Blog

Where marketing and creative teams find actionable advice, practical resources, and success stories to flourish in a world connected by content.


Like what you're reading?

We can send you our best articles.

Top DAM solutions

Forrester Wave: Digital Asset Management for Customer Experience, Q4 2019 Preview

Widen recognized as a “Strong Performer” in the Q4 2019 Forrester Wave report.

Get the report


The DAM Evaluation Toolkit

The DAM Evaluation Toolkit: Your path to the right DAM solution

Your guide to finding the right digital asset management (DAM) solution for your organization.

Get the toolkit


Widen Summit

Inspiring change: The 2020 Widen Summit Preview

Connect with your world-wide Widen network at this year’s free, virtual Widen Summit.

Register now

Are you ready for a DAM solution of your own? Request a demo