IEN Third Annual DAM Practitioners’ Summit – Breakthrough Solutions for the Management of Enterprise Digital Assets

This feature article was contributed by Jeffrey Marino.

Insight Exchange Network hosted its annual gathering of DAM professionals late last January at The Third Annual DAM Practitioners’ Summit in NYC. There was also a pre-event mixer at the Rum House bar which was hosted by IEN and NYC Digital Asset Managers. The conference chairperson was Dan Piro, Director of the Digital Asset Archive at the National Hockey League.

Presentations focused on DAM challenges, solutions and stories. Topics covered strategies and tactics for user adoption, governance, metrics for measuring DAM value, integration of DAM with other systems, workflows and more. Case studies, Q&A and a breakout session rounded out the agenda.

In Increasing User Adoption to Maximize your DAM’s ROI, Frank De Carlo posed qualifying questions to panel of DAM managers from major non-profits, sports media and the food and beverage industry.

Which comes first – integration or UX?

Henryk de Gyor, Principal Consultant, Another DAM Consultancy, presented the panel‘s unanimous opinion:

“UX is the number one thing. No-one has the patience for a terrible UX.”

Jessica Berlin, Digital Asset Management Director of the American Cancer Society, said:

“Integrations might look advantageous; but there is so much consolidation within the industry and amongst the different platforms, you have to consider how that might affect their stability.”

What about metrics?

The technical baseline for the system is quantifying speed of uploads, number of downloads, and sheer number of assets, said the panel. It’s quite another thing to be able to quantify usability, usefulness or the user experience. Leah Carlson, Global Digital Content Manager at McCormick & Company advised,

“Indicators need to be in context of the strategy for the DAM. For example, the number of users can decline but the dam could still be successful.”

Each panelist concurred on the priority of connecting directly with users as a major input – metrics for quality need to be considered alongside any quantitative measurements.

The qualitative baseline comes from user-support requests, says Anne Graham, Assistant Manager, Media Management at Turner Sports. She values anecdotal field research to identify issues – especially with beta users – and working directly with users to resolve them. The personal touch pays off by fostering collaboration:

“We partner with producers to get production assistants to fill metadata gaps, especially of older material. And that makes the archive more current and more useful.”

Quantitatively speaking, Carlson measures how often searches result in a download, and uses that as a conversion rate. Changes in conversion rates may point to gaps, ie availability of assets versus their actual usage. This in turn can provide input to the archive strategy. Berlin generates regular dashboard reports not only for the assets being downloaded but also for where those users are.

“If there are differences in activity for different regions, that gives our team an opportunity or reason to reach out to them.”

For Berlin, one as-yet-unrealized part of the dashboard is a null searches report. How many searches fail? What are the search terms involved? Correlating such details may go a long way towards improving the user experience.

A key takeaway from this first panel of the Summit: stay focused on good user experience. As de Gyor put it,

“It’s users that adopt the system, not the other way around.”

In the next presentation, Can AI Effectively Tag for Empathy? A Discussion on Artificial Intelligence and the Perception of Empathy,  Rebecca Schneider, Executive Director of Avenue CX, continued her exploration of a complex topic. As background: empathy has complex cognitive, emotional, and compassionate components, and processing a range of ‘messy’ inputs is generally thought of as what makes us ‘human.’ When one tags an image – Schneider showed one in monochrome, with children, and a cat, looking out a railcar window – and discussed how it unavoidably gets tagged with empathy. We interpret the image according to our experience and culture, with bias. From an AI’s perspective, such an image is not ‘clean’ data; it is riddled with ambiguity. For an AI to interpret (i.e. learn to tag), it requires a reference. How accurate is the process of tagging reference images, Schneider asks? As a human process, it is by definition subject to bias. As an example, Schneider cites a project called ImageNet where millions of selfies (user-generated facial images) were tagged by humans (‘mechanical turk’ individuals), generating tens of thousands of categories (a taxonomy nightmare) and keywords (lots of them inappropriate).

Kathryn Gronsbell of Carnegie Hall raised the baton for the next session, Show & Tell: A Look Under the Hood with Several Live Walkthroughs of Real-World DAMs and Some Novel Uses. Her recent project was the rollout and release of a newly public facing portal of the DAM, with 85,000 assets of historical interest. The organization’s Digital Collections site drew an audience of four thousand users in four months without any marketing, all the while having low impact on the department’s ongoing core operations of archive and preservation.  Baseline data indicate a low bounce rate (20%), with users spending extensive time on the site, viewing on average five pages per session. Gronsbell’s experience is a good case study of how feedback-driven projects can generate lessons learned, which can in turn apply to assessing, determining requirements, and planning improvements to internal tools as well as new public facing initiatives.

Demetrios Vasiadis, Manager of Digital Asset Technologies at Conde Nast, has cycled through several (he says ‘two and a half’) DAMs in his 12 years at the company. Despite the complexities – his DAM supports 26 brands, nearly 300 publications and over 8MM assets – the problems his team face (and their solutions) are common to any large DAM operation:

• Problem: Not visually appealing
• Solution: Create a landing page for each of the properties (publications)
• Problem: Extend the utilization of the DAM
• Solution: Make non-published content available
• Problem: No linking to rights
• Solution: The rights management process begins at upload (and with the uploader)
• Problem: Need for rapid upload and image processing at scale
• Solution: Use Amazon Web Services Elastic Transcoder to shift technical debt
• Problem: Lack of usage reporting (aka team needs to spend hours, even days on a report)
• Solution: Create a module called ‘The Registry’ to pull a comprehensive report of any asset’s use, within seconds, using blockchain unique identifiers
• Problem: Lots of null searches
• Solution: Methodology for weighting machine-learned tags versus people-entered tags
• Problem: Give it a name?
• Solution: ‘Montrose’

In the final part of the Show & Tell, Melinda McLauglin, CMO of Extreme Reach presented the anatomy of a fifteen second spot. MAM (Media Asset Management) assets are broadly categorized into running footage (let the cameras roll), segments (groups of shots making up a scene), shots, rich media ads & any number of local versions. “Assets naturally accrue metadata at each step,” says McLauglin, with many hands involved in managing the work in progress. Rights management is hyper-critical in the advertising world, driven by expensive media buys, residual payments to talent, and usage expiries that carry steep penalties if breached. While the typical DAM manager might not interact often in this sphere, McLaughlin’s notion of ‘precision compliance’ speaks directly to the ever-present need for best practices in managing copyright risk, usage and permissions.

In the sponsored session, Kenneth Hurta, Solutions Engineer at Brandfolder, talked about assets relate in their DAM via ‘containers’ and AI capabilities. For example, a container could hold images of retail products and their SKUs; the SKU image gets automatically scanned via OCR (optical character recognition) and becomes text searchable and firmly linked with the corresponding product.

In Elevating the Role of DAM Professionals — Career Pathing, Hedging Against Job Loss, Demonstrating Value-add, & More, veteran DAM practitioners prioritized three things: stay close to the end users, automate wherever possible, and have a thick skin. Asked for advice on resources for newcomers, they cited meetups like and NYC Digital Asset Managers and New Jersey DAM, various blogs, the new certificate courses at Rutgers and SF State, and conferences like Henry Stewart.

A straw poll of conference attendees revealed that this year’s audience was made up more of veterans than aspirants:

How long have you been in the DAM universe?

• One or two years: a couple
• Three or four years: a few
• Five or more year: some
• Ten or more years: many

Surveys got taken to another level in Sussing Out the DAM’s Worth at Dell Technologies, presented by Margie Foster, Digital Asset Management Librarian and Sarah Sundbeck, Digital Asset Manager, both from Dell Technologies. As background: Dell’s first DAM was managed out-of-house and had no reporting capability; the next DAM added reporting and brought its administration in-house. Next step for the team: obtain actionable feedback by surveying users.

Here’s how their approach evolved:

1^st – a single question customer satisfaction survey using a five-point scale (from ‘very satisfied’ to ‘not at all’) to provide a basic sense of the status of the DAM, but little insight.

2^nd – survey using a ten question System Usability Scale, the ‘SUS’ in wide use for evaluating users’ perceptions of effectiveness, efficiency and satisfaction of systems. The result was a surprising low score. “Don’t be afraid of that kind of result,” said Sundbeck, as it may illustrate opportunities for improvement, not to mention “an opportunity to understand what you should and shouldn’t measure.” Foster also realized that the low score was not necessarily valid, discovering after the fact that her SUS process was incomplete. The correct use of the Brookes SUS method is to generate multiple outputs, ie compare the scores of two or more systems, or two or more versions of the same system, instead of the single measure she made of their DAM.

3rd – Armed with these learnings, the team created a custom survey of 5 questions specifically designed for their DAM environment:

• What is your email address?
• Which type of DAM account do you use most?
• How satisfied are you?
• How do you usually locate assets in the DAM?
• In which way could the DAM work differently and better for you?

Note how this custom survey mixes quantitative and qualitative questions to generate both scores and feedback. The team plans to issues it twice a year  so the results can be compared over time.

The takeaway: data on users’ perceptions that becomes information can help direct the evolution of the DAM.

Day two, the first panel was Buyer’s Remorse and Lessons Learned: What I Wish I’d Known and What I’d be Sure to do Next Time. The first ask:

What do you want from your vendor?

Jennifer Anna, Photo & Digital Asset Manager at World Wildlife Fund emphasized,

“Be very selective of the functions you want to start with. There are myriad out of the box functionalities, but what matches up is variable. And priorities will shift according to users.”

Suzanne Saylor, Digital Asset Management at Boeing, mentioned that as you obtain requirements from your stakeholders, vendors need to roll with you. Consultant Howard Kloc advises empathy for the vendor: “They are trying to deliver the impossible.” Still, one panelist asked, “Have you ever seen an out of the box system work perfectly?”

Lessons Learned?

Jennifer Anna reports how important it is to a) socialize DAM practices prior to implementation; b) diligently follow-up; and c) not just rely on training and newsletters. Assemble key stakeholders early because “there will be surprises,” says Kloc. He narrated a story from a kickoff meeting where IT – while shaking hands with the DAM manager, the ad agency, and the business stakeholder, injected a 300-page questionnaire to the project requirements list.

Service Level Agreements between vendor and customer are important to contractually define and manage aggressively. Organizations may want to be able to exact penalties for the breaches of uptime guarantees, for example. Closer to the ground, be aware of dependencies that may affect operations even when the system is up. Think global in terms of connectivity. World Wildlife Fund, says Jennifer Anna, has field operations in remote areas. “Our testing should have included using the system with that remote bandwidth as a consideration.”

Following the thread on surveys was Case Study: Stories from the Field—Laying the Groundwork for Successful DAM Integrations, Natalie Daller from American Needle described her process of marking out the user journey and zeroing in on usability testing. To investigate ‘the state of the DAM,’ she used methods included card-sorting, a survey with a five-point Likert scale, and an open ended questionnaire. Importantly, she leveraged the skills of an in-house survey expert to design the survey and ensure its validity. Daller decided to baseline the DAM experience by surveying new employees – new users of the DAM. ‘Journey testing’ helped Daller substantiate issues and bolster the business case to make improvements regarding:

• poor usability for the users (numbering in the tens of thousands)
• search being hindered by an unfamiliar UI
• wayfinding being a challenge, with a ‘disjointed’ UX
• potentially redundancy of content in platforms like the LMS (learning management system)
• limitations of reporting, when other platforms are not integrated with the DAM

Daller’s takeaway: when there are unused assets, both productivity and time are wasted, and money is left on the table – this makes sense to senior management. “We were able to declare the DAM a central service that was important” and proceeded to plug the DAM into other platforms using APIs.

Daller’s new DAM has four API integrations. They’ll collect baseline measurements – in particular the search-to-download conversion rate – after six months in operation. For users to get good results from search, she reiterates, it all depends on good metadata. She wryly summed up,

“You’re never really done. Don’t half asset.”

Consultants Elena Brodie-Kusa and Roy Walter joined the next panel, DAM Integrations — What’s to Be Gained and How Will They Impact Your Workflows? along with Oksana Dersovitz, Senior Project Analyst, DAM at Wakefern. Numerous comments earlier in the conference about frustrations with IT caused Dersovitz to stand up in defense.

“I’m in the IT department. And I am the DAM owner and sponsor.”

“We get 10,000 requests a day for images updated by manufacturers,” she says. Wakefern owns major supermarkets like Shop Rite, so the product and image supply chain are mission-critical. At Wakefern, the utility of the DAM is driven by a major ‘predecessor’ system: managers enter product data into the PIM (product information system); the data flows downstream to the DAM via the API so that product searches always correlate DAM product images with accurate PIM data.

“Once we defined the PIM as the predecessor system, we mapped the data and worked out places where we could automate. We did a metadata review with the stakeholders to align different fields as needed. And then we used the API to feed into the DAM.”

With only two engineers trained in the vendor’s proprietary API, Wakefern paid money to the vendor for additional support. Be aware, says Dersowitz, that not all APIs are available to all customers. “The vendor might require that you upgrade to an enterprise plan in order to take advantage of an API.”

The next panel was Best Practices Deep -Dive for Governance Program & Rights Management. The DAM platform can control user behavior automatically through permissions – who can add metadata, who can upload, what can be uploaded. Manual methods for controlling user behavior include the carrot (praising users, issuing rewards for good metadata practice) and the stick (photographers don’t get paid if they don’t enter metadata properly.) Lauren Dohr of Mediaverse Information Management summed up:

“There isn’t one single point of governance. You need to standardize processes and work on integrating your information, your workflows and processes across the organization.”

The last session of the day was The Pros and Cons of Work-In-Progress Versus Final-Assets DAM.

What’s in a DAM?

• Ninety percent of assets in our DAM are final (Natalie Daller, American Needle)
• Mostly final assets (Stacey Jurgenson, PBS)
• All final assets (about 10% of the audience)
• Work-in-progress assets (about 10% of the audience)
• Final assets and work-in-progress assets (about 20% of the audience)

What about workflow tools?

• We saw resistance on integrating Workfront into the DAM; it got too complicated (Fred Robertson, Digital Asset Manager)
• We looked at the app that’s within the DAM and found that it’s not the best process. It’s best to have the user leverage the tool where the asset is going to wind up. (Stacey Jurgenson)
• When integrated, they might have some impact on work-in-progress (Natalie Daller, American Needle)

What about the black hole of storage?

• A hard-to-answer question. Two bad practices: naming folders ‘old,’ and blind deletion (Fred Robertson)
• We have a to-be-reviewed section, which is itself subject to periodic review. We then go to the content owner to finalize the asset, or purge. (Stacey Jurgenson)
• Defining experiences: a) needing to prove the use of assets over time, and b) a thirty-year anniversary scramble, leads to c): Archive forever. Do not delete. (Natalie Daller)

The IEN DAM Practitioners’ Summit offers participants insights from veterans in the field who are ready to share their successes – and to a lesser degree, their issues in planning, implementing and administering their DAM. That only a few vendors have tables simultaneously means there is less pressure on the exhibition floor overall and better access to the marketers eager to present their ware – but with less on-the-spot comparison available. This conference was a great forum of information sharing among those well-seated in the industry, and the new round-table discussion provided an excellent opportunity to learn what other audience members are up to.

Share this Article: