This article was written by Jeffrey Marino.
DAM Chicago 2019, The Art and Practice of Managing Digital Media, kicked off September 24th. Before the sessions began, I chatted with David Lipsey, the conference Chair, and asked for some insight on what was in store for us at the ‘the Midwest’s largest conference dedicated to Digital Asset Management.’
“Chicago is a very interesting expression of DAM in the marketplace. There are hundreds of companies in the region that rank in the top thousand in the US. Their imaging, branding and e-commerce needs are a significant opportunity for DAM technology and personnel, and it’s happening right now.”
Lipsey then kicked off the conference with the baseline definition that digital assets are information assets, and they need to be considered in the context of a long and complex supply-chain. Recognizably embedded in the real world, “these assets are a shadow currency in contemporary life, operating outside corporate norms of audit and accountability.”
While yesterday’s digital asset has value as ‘a single source of truth,’ today its value is as a transactional device, “a core currency and handshake of engagement.” Thought provoking and intellectually generous, Lipsey is also the director of the newly stood up DAM Certification curriculum at Rutgers University.
Praveen Moturu presented the keynote address, Digital Assets, Digital Eco System & Digital Transformation: How Mars Inc. is leveraging Digital Engine & Platforms to connect the Digital Eco Systems. His mile-high view as VP and Chief Enterprise Architect illustrated the complexity of the organization’s as-is state as a ‘hairball’ while depicting its transformative should-be state as a true metamorphosis.
The DAM initiative (like the caterpillar) is subject to birth, growth and risks to viability before it can mature and take off. “Many caterpillars die without becoming a butterfly,” he advises, so take good care. Be judicious with what does and does not go into the DAM. And stay aware of the challenging environment. In particular Moturu cites these constant disruptors:
- the always increasing volume of assets and distribution channels, spaces and contexts;
- the variety of types and usage such as VR, AR and 3D;
- and the everlasting need for compliance with external and internal regulations.
In the next address, Reframing the Conversation – Innovations in DAM, Collections Information, and Data at the Detroit Institute of Arts, Jessica Herczeg-Konecny described the the museum’s transition from departmental to enterprise DAM, and the process of migrating 60,000 assets to a new platform. With a small staff of ‘one and a half’ at her disposal, she acknowledged that “sun-setting the old system was hard,” and advised “it’s important to have a good exit strategy.” With the bulk of the Institute’s 90,000 assets migrated, her role is now to ‘captain, cheerlead and champion’ the new system and its users – collectors, custodians and contributors at the organization.
Christine Gibbs, the Collections Database Manager at the museum, said that with today’s maturity of DAM technology, ‘APIs are the new hub,’ and is looking ahead to connect the new DAM to the museum CMS and to broadening its audience. In an interesting and practical example of DAM and the real world, Gibbs used asset data on physical dimensions of objects to calculate shelf space requirements of the physical plant – very useful info to have on hand when capital expenditure and budgeting opportunities come around! Another size-calculus illustrates the scope of digitalization of paper records into the DAM: over 17 square miles of documents. (In a wry Midwest gibe, Gibbs qualified this area as “enough to cover half of Manhattan.”)
Next, Gil Comeaux and Travis Garrett presented Lessons Learned from a 2 year DAM Expedition by Tyson Foods. Managing content at a major producer in the food supply chain is a big challenge given the constant critical updates – not only to package design (artwork) but also to accurate, and legal, labeling ingredients and nutrition. The DAM also serves packing and shipping – the labeling workflow of cartons with their contents and destination, automatically customized for tracking and accountability. Their small team (2 people) succeeds in a large enterprise by maintaining clear data standards (what should go into the DAM vs what should not), and extensive communication with a global user base (feedback, training, webinars, surveys, and more). Operational keywords for the team were to ‘unravel’ and ‘refit’ a wide range of business processes and to set DAM – which they code named ‘JIVE’ – at the very center of the Tyson Foods digital asset supply chain. Based upon creative production and rework costs, they calculate the value of their 150,000+ assets at an estimated $16MM. And in case that valuation didn’t get the attention of their stakeholders to justify the initiative, they ALSO produced an amazing video, JIVE THE QUEST. Check it out, it’s a must-see!
Next we attended DAM Innovation: Technical Experts Address Your Really Tough Questions. The core challenges for DAM are their users’ expectations, said Lisa McIntyre from Nuxeo, citing how all consumers think of Google and Instagram as the be-all and end-all of UX. Other panelists chimed in about network speed (as in, ‘why is our corporate wifi so slow?’). The core opportunity is that with good connectivity, APIs today make integrations of on-prem to cloud-based apps and storage easier than ever. Of course, said Rich Carroll from Sitecore, we should realize that large media files like 4K and 8K still belong on-prem.
Regarding copyright, McIntyre summarized that if asset reuse is the raison d’etre for the DAM, the 1st priority for that implementation needs to be setting up rights and permissions properly. “Ask for permission rather than count on forgiveness – you’ll never go wrong with that.” To that very point, in a later session we were reminded how Taster’s Choice used an image without permission on jars of freeze-dried coffee resulting in a good-to-the-last-drop $15MM legal settlement.
The panel then tackled distinctions between Artificial Intelligence and Machine Learning. Many (or most?) DAMs feature some sort of AI under the hood. “AI has real use cases and value in DAM,” says Jim Hanifen from Brandfolder.
He polled the audience for a show of hands:
- How many of you are using AI now? (one hand, maybe two)
- How many of you want AI? (about a dozen)
- How many of you have budgeted for AI (zero).
As we pondered these results, a comment from the audience comment broke the silence.
“We had to turn our AI off after one day. It was kind of psychotic!”
(More on this in the last session of the day.)
Neil Grant from Tenovos explained a bit about Machine Learning, and one of the ways AI seeks to improve itself “Machine learning is about looking at large datasets, and it learns by looking at the data around search results, to identify patterns there.” Datasets can be quite large in DAM (though nowhere near as big as those in ecommerce, downstream of DAM), and applied learnings can help DAM managers know more about their users and how their assets are used, and even ‘heal’ metadata.
Finally, product development by vendors is, by definition, ahead of user adoption. Brandfolder’s Henifen is sure that “our tools are far beyond our users’ knowledge of how to make use of them.” How is a DAM practitioner to cope? In the next panel, DAM’s Role in an Integrated Technology Environment, Michele Dickinson Heuer, lead Digital Asset Manager at Nestlé Purina Petcare North America, says “it’s important to try what’s in the box, and question how one is working,” before clamoring for new (or newfangled) features. At Nestlé , best practices are to focus on the tools at hand, support superusers, achieve wins consistently (small ones are valuable!) and gain DAM champions in the process.
Introducing the closing panel, Why Metadata Matters, Chair David Lipsey posed the challenge question, What are Ostraca? hinting that ‘it’s the original metadata.’
Pottery shards are artifacts in the archeological story, and ostraca, as objects inscribed with writing are indeed akin to assets tagged with metadata. They lend additional meaning, importance and value to the set of objects telling that story. Metadata, Lipsey says, “is the spirit of a physical, intellectual and creative asset.” Does the alchemy of spirit and intelligence beget life and… personality?
- “Our AI is a comedian,” says Kim Johnson of Hilton Hotels.
- “Ours is a sociopath,” says Charlotte of Coty, Inc. “It mistook chrome nail polish for a knife.”
- “We researched DAM for 7 to 8 months and leapfrogged right to AI,” says Anne Adami from VideoFashion. “It’s improving and learning really well.”
The ability for AI to thrive at VideoFashion is directly related to its environment. VideoFashion is an eponymous single brand holding exclusive rights to its video assets: it’s the largest private fashion library and archive in the world, encompassing 43 years of models, runways, haute couture, glamor and celebrity sightings. While their AI was highly skilled ‘out of the box’ in OCR (optical character recognition, or ‘reading’) and speech to text (‘transcription’), Adami was able to provide it with hundreds of scripted shows for its core education, and set her AI up for machine learning success. Not everything, though: facial recognition still needed refinement.“We had to teach it not to identify Yves St Laurent as Henry Kissinger.”
A more challenging environment for AI is Hilton’s 6,000 properties, 6,000 websites and 6,000 sets of assets. Regional and local differences are critical components of the business, and it’s understandable for Johnson’s asset managers (‘the largest DAM team in the world’) to be amused by the AI, but holding it at arm’s length.
Coty is similarly cautious about next steps with AI. With 50 brands, 3 internal divisions and international differences to work with, one working solution is to support 3 metadata configurations – reflecting perhaps the differences of ‘spirit’ in how different region communicate and conduct business.
Here’s a bit of the lively Q&A with these panelists.
Q: Do you support synonyms in your approved taxonomy?
A: Charlotte (Coty) – Yes
A: Kim (Hilton) – No (but we have a freeform keywords field)
A: Anne (VideoFashion) – Yes! If the AI recognizes the synonym, the AI will figure it out.
Q: What about usage rights; what is most important?
A: Charlotte (Coty) – Expiration and territory
A: Kim (Hilton) – Tracking usage across channels
A: Anne (VideoFashion) – No worries! It all belongs to us.
Q: What about archiving?
A: Charlotte (Coty) – Retire, yes. Delete, no
A: Kim (Hilton) – We have a live-archive-deep archive model for our assets. Licensed assets expire automatically
A: Anne (VideoFashion) – Never delete. Just add storage!
With two simultaneous content tracks to choose from, the excerpts above represent only a bit more than half of the offerings.. We had a nice opportunity to get to know other attendees in a more intimate round-table setting hosted by Pacific Media Technologies: What Happens Before DAM: The Risks of not Digitizing Physical Assets, which was so interesting we didn’t get to visit any of the other seven or eight round-tables happening nearby. Henry Stewart DAM Chicago packs a lot into a single day!