From 2007 to 2021, I served as Supervisor of Digital Asset Management at Crispin Porter+Bogusky in Boulder, Colorado. In my time there I worked between the IT and Production & Creative Departments to assist the company through the evolution of asset storage from 100% on-prem local file & ftp servers to a cloud based, enterprise digital asset management platform featuring integrations with our internal project management and finance systems. These integrations became the basis of automated workflows for assigning project metadata and filtering broadcast video and photography assets into dynamically updated secure client portals. Using this system, my team also generated streaming conversions for OTT and web ad placements, providing both a revenue stream and efficiency gain for the agency and cost savings for clients.

In January of 2022 I was contracted to build out a digital asset managment platform for Plume Design, Inc., a Palo Alto based technology company creating revolutionary Wi-Fi management hardware and software for consumers and ISPs. As a rapidly growing company that produced large quantities of articles, product & technical datasheets, whitepapers, and DTC & B2B marketing materials, they needed a system to capture critical technical and business organization information related to their assets, as well as consistent version control for distribution to their sales, marketing teams, and ISP partners.

While Digital Asset Management (DAM) is a rapidly expanding business need, many people aren't quite sure what it means or how it differs from using local file storage or consumer level cloud services like Google Drive. Below is a synopsis of some fundamental concepts of Digital Asset Management and case studies showing the different types of DAM implementations I built and managed at CP+B and Plume Design.

As companies and organizations generate ever-increasing quantities of digital files and transition away from traditional on-prem servers and physical storage media toward cloud-based file storage solutions and a remote or hybid workforce, the need for cost effective, efficient organization and simplified, secure sharing of assets and content has become increasingly important. At the same time, managing versioning, usage rights and all the relevant information associated with those those assets results in the mass proliferation of disconnected, multi-tab spreadsheets that are labor intensive and fatiguing to manage and keep updated to the point where they become a problem in and of themselves. The age of on-prem servers and folder-based storage is rapidly being supplanted by secure and cost-effective cloud-based abstracted metadata based file storage for graphic, video, and "virtual" assets such as online forms, lottie animations, and other website elements that are shared across production, sales, and marketing teams. As both a concept and a solution, Digital Asset Management bridges the gap between abstracted, cloud based file storage and the need for organizations to be able to quickly and easily organize, find, track and distribute assets across multiple departments, teams, and external channels with minimal siloing and duplication along the way. Though the goals and objectives of Digital Asset Management can vary greatly depending on the specific needs of each company or organization, metadata is the foundation for every effective digital asset management solution.

Ok, so what is this METADATA stuff?

Definitionally, metadata is data that describes and provides information about data. While it sounds confusiong and redundant, in reality it's a very simple concept that is at the foundation of the way people organize information in their own minds and is a much more natural and efficient way to organize data than the server> folder> subfolder> sub-subfolder> file name methodology that we have been accustomed to associate with digital file storage. In fact, metadata is the organizational system used in every online store and music or video streaming service.

The two greatest weaknesses and inherent inefficiencies of the traditional server>folder method of file organization are isolation and duplication. There is never a single way to adequately classify or describe an asset and an asset cannot live in more one place at a time without being duplicated across multiple folders, so there are only two possible solutions in a traditional folder structure-- create a shallow folder structure that only describes an asset at a high level, or create a deep folder structure where an asset is prolifically duplicated in multiple folders.

Consider the red dress image on the right as an example. If this dress were organized in a traditional folder structure it would have to be lumped into a single high level folder that ignores its details, buried deep in a nested folder structure that would isolate it from other items that share certain traits or duplicated across 10 separate folders with descriptive names, or worst of all, kept on an individual's computer destkop. By using metadata, those pitfalls can be avoided and the single asset can appear in 10 different metadata-based virtual folders simultaneously (typically referred to as "collections" in DAMspeak). In addition to providing the ability to thoroughly and completely describe an asset, it also makes managing the asset much easier. If the dress in this example is updated for a new year, it only needs to be replaced one time in a single location. Since the the classifications are based on metadata instead of folders, it will dynamically update in all the collections it's associated with, eliminating the risk of having an incorrect asset used in published materials.

But metadata isn't just restricted to descriptive "tags" for an asset. It can also be used to direct assets to specific company divisions, teams and departments. Taken a step further, metadata can be even used to automate workflows and build integrations with systems outside of the DAM platform itself. Consequently, developing a metadata taxonomy that covers all aspects of an organization's structure, workflows, assets, and objectives is a critical step in building an effective Digital Asset Management solution.

CASE STUDIES

Crispin Porter + Bogusky

In 2016, after nearly 20 years in business, award-winning advertising agency Crispin Porter+Bogusky was named Global AOR for a luxury automotive brand. The scope of work included a requirement for an online distribution system for tracking usage rights and sharing video and photography assets with regional clients. Using traditional FTP server and 3rd party file sharing apps for a high value global client was deemed an insecure, labor intensive, and "off-brand" band-aid solution. CP+B was also coming to the realization that continuing to store their entire historical library of broadcast masters on phyisical tapes and hard drives in offsite vault storage facilities was an outdated and inefficient method, so the decision was made to implement an enterprise level, cloud-based Digital Asset Management system to digitize storage in a cloud based system and improve internal workflow efficiency, overall. As Digital Asset Management Supervisor I saw this as an excellent opportunity to completely re-invent the way the company managed its content internally and distributed it to clients and vendors. The ultimate goal was to create a model that could be scaled up across the entire MDC Partners agency network.

The next step was deciding to "buy or build." Building a proprietary system in-house that's owned by a company or organization is always tempting, but with limited IT resources available for engineering and ongoing support, we quickly came to the conclusion that building a system from scratch was an unfeasible option. With the myriad of proven SaaS (Software as a Service) options available it made little sense to spend extensive time and money building something that already exists. Instead, we opted to utlize a SaaS solution and customize it by implementing proprietary integrations with our on-prem infrastructure and internal finance & project management systems.

Each year there are new DAM services coming online and improvements made to existing systems. All have their strengths and weaknesses, so when it comes to choosing the best one the first step is defining the goals and requirements-- what is it that we wanted to acheive with a DAM solution?

• External Facing Portals: Need a simple, efficient way to expose viewable transcoded video & still assets of all types and formats (.mov, .wmv., .mpg .psd, tiff, jepg, .png. indd., .docx, .ppt, .pdf, etc) working files and documents to other CP+B offices, clients & external vendors directly from the file server and control user & group upload, download and view permissions in custom branded portals specific to each client.
• Asset Request Management System: Need a module to manage and fulfill file requests from 3rd parties.
• High Speed Global Distribution: Need an integrated system for delivering high res video and print assets globally using high speed non tcp-ip digital transfer (i.e. Aspera)
• Annotations and Approvals Module: A module to facilitate commenting, markup, discussion and approvals for creative development and client approval
• Automation: Ability to auto-ingest assets from SAN and associate metadata from folder structure & internal project management system API.
• Custom metadata schemas: Create custom metadata schemas to embed on assets for purposes of searching, tracking, identification and rights management.
• Rights Managed Asset Status: A method to highlight and send notifications when assets with time limited usage are nearing expiration and visual representation of status in UI.
• Support: Robust network for 24/7 technical, hardware & user support requiring minimal CPB resources
• Single sign-on integration
• Job-Jackets API MetaData Assignment
• Heterogeneous : Ability to utilize the tool fully on both Mac and PC clients
• Access controls - Role based and access layer security with ability to restrict access to any components of the portal or digital assets, including but not limited to: editing/revising, downloading, sharing, and deletion of files/folders within the system.
• Cloud based (SaaS) and/or hybrid solution where we leverage some type of bridgehead server to cache common assets for faster local access.
• SSAE 16 Compliance
• Platform should be hosted in a Tier IV Datacenter

Click for larger Automated DAM project creation and metadata assignment workflow; triggered by job status change in the project management system

With our list of requirements established, we began evaluating DAM systems and setting up formal online demos. Having a list of requirements made this a relatively simple process and we quickly narrowed the field of potential options. Most of our requirements were easily met by the systems we demo'd, but there was one that became a sticking point-- the ability to sync assets on our on-prem SAN to the cloud and build the integration with our internal systems for automated project creation workflows. This was an important requirement for 2 reasons:

1. Usage rights management for Global client assets can be very complex, so all assets needed be associated with and consolidated under their originating agency job numbers to facilitate organization and backtracking to associated contracts and POs. This was the foundation of our metadata taxonomy.

2. Automated project creation workflow vastly increased system efficiency by creating a predetermined holding space for assets on both an on-prem SAN and the cloud platform with a standardized, automated naming convention and folder structure. This automated workflow was triggered when a project estimate was signed and the Finance team changed the status in the Poject Management system from "Open" to "In Production." This would eliminate scattered silos of assets which are ubiquitous in traditional folder based file server organization. This way, the job information in the DAM plarform where assets were stored would always match the information in the Project Management system.

In the years since 2016, the Digital Asset Management industry (and the web-based application landscape, as a whole) has spawned an entire service sub-sector devoted to system integrations, but at the time this type of service was hard to come by. Either the systems themselves had not yet developed open APIs or they lacked the internal professional services teams to create custom integrations for their customers. Only one of the systems we evaluated was able to meet all all our requirements and had an internal professional services team to build the integrations we needed-- AdStream North America. Apart from being the only DAM vendor that could meet the integration requirements, their platform also incorporated a full service global broadcast TV trafficking module with complete reporting and analytics.

With the help of AdStream's professional services team, the foundation of the CP+B metadata taxonomy was auto-generated via an API integration with the internal project management system to capture and assign the following information to assets ingested into the system:

• Client Name
• Job Number
• Job Name
• Job Type & Subtype (Video/Broadcast, Interactive/Digital, Print, Photography, Design)
• Campaign Name
• Product (where applicable)
• Team Members (Creative Director, Art Director(s), Producer(s), Business Affairs, Talent Manager, Account Manager)
• Launch Date

By using the Client property in the metadata taxonomy generated by the integration with the project management system, assets were dynamically filtered into the client portals. From there they could further organized by the Job Type classification, campaign name or any other metadata value or client specific tagging required. Using this metadata-based organization, the client portals required no separate maintenance to update as new assets were added to the platform. Different types of assets (video, photography, graphics) related to a specific campaign (i.e. Holiday 2018) could also be easily associated with one another without duplication from asset type collections to campaign collections.

In September of 2020, MDC Partners (CP+B's parent holding company) held an internal "Shark Tank" competition where employees and teams from the agency network presented ideas to solve agency problems. This was a period of time in which there were efforts being made to have agencies within the MDC network collaborate on client projects, so I created a presentation outlining the features and process of building the CP+B Digital Asset Management System and proposed scaling it out across the entire MDC network to facilitate inter-agency collaboration. Titled "Digital Asset Management and the Advent of the Agency Cloud Ecosystem," I was chosen as one of 5 finalists to be presented in an online forum to the MDC leadership team and over 200 employees withing the agency network. A recorded video of the complete presentation is available here:

Plume Design, Inc.
In January of 2022 I accepted a contract position to build out a Digital Asset Management system for Plume Design, Inc.-- a Palo Alto, CA based company that has developed a cutting-edge suite of software and hardware solutions for Smart Home & IOT WiFi management. As a rapidly growing company that creates large quantities of marketing and sales assets for consumers, small businesses and ISPs, they faced a common problem among companies with a largely remote workforce spread across multiple countries and continents-- they had no central repository for their assets. Files were spread across disparate department folders in Google Drive and having no consistent naming convention, duplication was rampant and version control was non-existent. Within the organization, various divisions (Brand, Sales, Marketing, Support, Events, B2B, D2C) also had their own specialized systems for tracking analytics on lead generation and partner marketing resources which leveraged large numbers of cross-utltized assets such as whitepapers, case studies and documentation for their software and hardware products which were frequently updated as new versions of their software were released. The situation was becoming increasingly difficult to manage via spreadsheets so the Creative Services team made the smart decision to build out an enterprise level DAM system to consolidate and efficiently manage the volume of assets they were producing. As such, these were some of primary goals they wanted to acheive with their DAM system implementation:

1. Consolidate all cross-utilized sales and marketing assets into a single platform (logos, software datasheets, borchues whitepapers, case studies, articles, product photography, D2C and B2B brand videos)
2. Version controls
3. Integrations with Sales and Marketing systems for automated file syncing
4. Source file linking (assets linked to Figma working files)
5. "Hero data" tracking
6. Connecting/Associating localized versions of assets.

At the time I was brought on, the key stakeholders at Plume had already gone through the process of evaluating and choosing a DAM system so I was able to to immediately jump into developing their metadata taxonomy. Unlike my previous experience at CP+B where the metadata taxonomy was relatively shallow and centered around clients, job numbers, and job types, the Plume metadata taxonomy needed to be much more detailed and cover all aspects of the business organization from divisions to departments, product suites and their associated apps, asset types & sub-types, and originators, all the way down to specific data and statistics cited within assets that needed to be tracked within the system so they could be updated as needed. Creating a taxonomy as vast and comprehensive as this can quickly become unmanageable from an asset ingest standpoint, so it required developing extensive metadata dependencies within the taxonomy. This is a strategy for condensing metadata properties by associating sets of options with specific values-- an "if A, then B" strategy. This facilitates the accomodation of a large number of options for specific metadata properties when a only limited number of options are relevant to what is selected as the parent property.

The table below demonstrates how metaproperty dependencies can be utilized on a single asset type (Video) with only two asset sub-types (Consumer Video & B2B Video) to accommodate 22 possible options. Because the options only become available when specific values are selected, this facilitates creating a detailed, comprehensive asset taxonomy that still remains condensed and concise within the ingest process. This also allows collections or filters to be created for sub-types that will automatically display only the options that are relevant to them. For example, in the dependency table below, only Broadcast Videos will display the duration and region options. Instagram and Facebook videos will offer an option to assign aspect ratio, but YouTube videos will not.

In addition to extensive use of dependencies for managing large amounts of descriptive metadata for various asset types, the Plume DAM system also leveraged the metadata taxonomy for creating filter based collections for specific departments. With one click from the platform landing page, division and department teams could have all the assets relevant to them in one place where they could be further sorted by types, campaign, originator or any other taxonomic value desired.

The metadata taxonomy was also used as the foundation for an API integration between the DAM system and the Sales Enablement software platform. By selecting a Syncing metaproperty option on asset ingest, an automated workflow would push the file (or new versions of existing files) from the DAM system to the Sales Enablement platform, ensuring the Sales Enablement Team always had a comphensive library of the correct and current versions of all relevant assets. Creating these non-descriptive metadata properties within the taxonomy opened up the door for a variety of automated asset organization options, which greatly enhanced indvidual and organizational adoption of the new system.