Unified Content Operations and the Pluggable Content Supply Chain

Cruce Saunders 
Welcome to towards a smarter world. This is your host, Cruce Saunders, and I am so lucky today to be joined by Mike Iantosca. Michael Iantosca has been in the industry driving every form of structured content from the earliest days. That has included a stint at IBM where he helped to pioneer SGML and the applications of data XML across a huge portfolio, and since then has been working in a number of different companies building and managing content supply chains of various kinds. That's our focus today. Content supply chains. Mike, thanks for being with us.

Mike Iantosca
Thank you for having me, Cruce.

Cruce Saunders
Let's jump right in with the content supply chain. We could talk almost endlessly about all of the history that has led up to this moment in time, but let's just start with what we know, which is that everything is moving quickly in the content world. We have large language models generating tons of unstructured content in various forms. We have lots of innovation across content technology, the advent of many different CCMs and CMS and translation and localization workflows. Lots of tools for semantic management, knowledge graph management, many different tools and processes that have matured over the last couple of decades for delivering structured content out to customers just in time to meet customers in context with support articles. You've been working in this space for a long time and now you're working on content supply chains. That is such a natural outgrowth of everything we've been doing for years. And I'm really curious how you see content supply chains, what they are, how they work, and why they're important for content leaders to be aware of and start driving towards.

Mike Iantosca
Thanks for having me today. It's a pleasure and an honor to work and speak with you. Yeah, content supply chain, let's parse that a little bit. We have a lot of tools and systems and processes. None of them work in a vacuum. They're all interconnected. And the people and everything that the multitude of people involved, what these systems and processes do are all interconnected. There's a lot of communication and integration that needs to happen to make an operation work smoothly from end to end, especially when the organization begins to get sizable. And by sizable I mean anything from a dozen or more people. It doesn't have to be dozens or hundreds of folks. Let's break down what we mean by content supply chain. The way people are thinking about content supply chain today is in a limited way. They're thinking for the most part about the content creation and delivery supply chain. And in fact, that's where most shops have spent or are spending most of their time, they're doing things like moving to structured or combinations of structured and unstructured content. So they've got migrations to handle. They're implementing content management systems, whether they're standard or componentized versions.

They're implementing content delivery platforms, CDPs, and many more tools, taxonomy systems, terminology systems, content quality tools. And they're trying to not only deploy these individual pieces, but tie them together into an orchestrated process for content creation and delivery. And that's the way people mostly think of when they think of a content supply chain. But there's a different kind of content supply chain, and it's one that we do manually, almost universally across the world, and that is the processes and tasks that we have to do outside of the content creation and delivery platform, to weave the entire process and procedures and standards together across multiple people during the planning and management and orchestration. And I call that operations supply chain. And if we break these two apart, we realize that we're doing great with the first one, where most companies have off the shelf technology now that they used to have to build in years past, but now they can go out and buy state of the art tools, state of the art content management systems, state of the art content delivery systems, and state of the art semantic systems. And it's still a little bit of connect the dots to pull them all together.

It's still not totally plug and play. Many of them have connectors to put them together like legos now, and it still requires some degree of customization. But we're living in the golden age, I call it, of content production, or content creation and delivery, because not many years ago, we didn't have all of those tremendous vendors and tools. We had to build everything ourselves. And now we go off and we license them, and then we take an IT team or a small engineering team or third parties to come in and help us assemble these content production supply chains. But we really have a problem in that the other supply chain, the operations supply chain, is almost completely manual. And just as the content has been siloed across different organizations historically, and folks are working really hard to bust those silos of the content itself, the operations are siloed and disconnected.

Cruce Saunders
Yes, absolutely. I'm noticing that in most organizations there's a focus on authoring and getting content into a CMS, and then all of that is still very manual in most organizations, the migration from unstructured to structured and then out to channels. But that is such a limited perception of content operations. Can you elaborate more on some of the larger topology involved in content operations.

Mike Iantosca
Yeah, that's a great one. Let's start with content strategy, right? It's the most overloaded term in our discipline. When we think about content strategy, or if you're in a room with ten people and you say, let's talk about content strategy, I guarantee you that with ten people, you're going to get ten different answers as to what they think content strategy is. Content strategy is really a collector of many strategies, so let's just mention them, right? We have content modeling, we have content migration and conversion. We have content design, we have content creation strategy. We have content reuse strategy, publishing strategy, governance, quality and governance strategy, localization strategy, compliance strategy, and knowledge management strategies, to name a few. If that doesn't make your head spin, I don't know what would. Because what ties all of that together? Certainly not the content creation and publishing systems that we're assembling and call the content supply chain. That's the content creation and publishing supply chain. So how do you have a strategy if you don't weave those together and orchestrate them? Number one. And here's the second provocative question. In the middle of some talks, I'm giving one of my favorite questions to ask the audience is, it's 10:00 p.m. Do you know where your content is?

Ask that question in a practical sense to almost anybody and almost any company and say, do you know? And could you give me an inventory of all of the content produced by all the different organizations in your company for product X that are internal or external facing, and you will draw a blank stare. That's a pretty sad statement, because if you don't know holistically what and where all your content is, let alone coordinate it and manage it, how do you have a content strategy to begin with? And that's the elephant in the room that nobody has seemed to want to address, because it's been difficult enough to try to bust the content silos themselves, let alone bust the content operation silos that exist across many functions of our company.

Cruce Saunders
Yeah. And when you don't have the ability to understand how the entire portfolio is composed and how it's working together across actions, there's no way to optimize that set of workflows. So having that analytics view that gives you a sense of all of your assets, where they are in their lifecycles, and for example, which assets are depreciating in value because they're no longer updated, useful, or accessed in the way that they were when they were initially created. So just being able to have visibility allows for responsiveness and adaptation and refreshing of the value of a content asset portfolio. I'm also aware that in many organizations, there's a whole set of content lifecycles for individual assets that gets away from one given system of record. So once it hits the Jira ticket or the Workfront application, it just kind of lives there. But it's outside of the CMS. And then once it gets into the CMS, it's outside of visibility of the project management application. And then once it hits the process for localization, there's not necessarily an automated trigger for getting that content back in and published. Along with the base language versions, there's this connective tissue that's needed in operations, isn't there?

And how can organizations help to get their hand on sort of these bigger picture portfolio management dynamics of content velocity and link checking and trademarking and accessibility, and other sort of high level portfolio management dynamics that involve all of the vectors that drive value in a content portfolio?

Mike Iantosca
That's a great question. I love it. And this has been the crux of the problem forever. So let's go back to two of the notions that you covered there. One is how do we manage the knowledge and the information about everything that we have? Well, if you're a good publications organization, at least a good product pubs organization, or a marketing organization, people talk about having a publishing agenda where they maintain on a nice Excel spreadsheet, their publishing agenda. And that is not an uncommon system of record. And it's locked up and it's siloed. And the other thing you mentioned is systems like Jira and similar. These are transient systems of record, right? A Jira ticket gets created, somebody creates something, they fulfill the request, they close the ticket. And that is a transient record, that is not a permanent system of record. So the key here is not very complex. The key is that you need to have an enterprise, a centralized, enterprise wide system of record for every publication that the company is going to publish for a given product or service or offering. And that persistent system of record lives for every publication in a central system, for the life of that product and document, until it's time to archive it and set it aside.

And this can be years, right? We create maybe short term content in the form of support kcs articles that might live for six months or a year or two years. And then we create product documentation that could exist for 3,5,10 or more years. And where is the central system of record that has all of them, so that we could weave together all of these strategies and operations and begin to implement automation and communication across all the parties that are involved. It's one thing even to have a system of record for just one functional area. So I do know of one or two shops that have a central system of record, but it's only for the product documentation. There's no common system of record for all of the other kinds of content that are created, right? We create content for developers. We create and usually created by separate organizations, right? We create content for marketing and sales. We create content that's learning content. We create support content. So these are all, if you will, silos of content planning and management that don't come together in any central way. And unless you bring them together into some central system of record, you can't manage the whole lifecycle of all of the content for an individual product, let alone a solution.

Remember, sometimes, many times, very often, companies have products and their product stacks. It's not just one product. In the case of Avalara, our flagship products are Avatax, for doing sales tax computation and returns. What people file, right? Sales tax returns. Those are two products. And what if we want to look at all of the content for not just Avatax, but we want to look at all the content for Avatax and returns together as a solution. And in some cases, three and four products get together are combined by a customer to create a solution. So we need a way to look at not only the aggregate amount of content across the enterprise, because the customers don't care who inside the company created the content. They don't care that a learning organization created learning courses, and that a separate organization created the product doc, and a separate organization created the partner or the developer doc. They just want a comprehensive access to the content and don't want to know how or where the content was created. They want what I call the total content experience. And the total content experience is the goal. But you can't have a total content experience and manage a total content experience unless you have a central system of record, which I called the total system of record.

I don't know if you want to give it a name, but we need a sort of a content central, that is an operations content supply chain to do that.

Cruce Saunders
Yeah, that ability to essentially put railroad tracks on operations or have a nimble, pluggable supply chain, really, I mean, more than railroad tracks. It's more of a framework that allows for interoperability of process and workflow, analytics and metadata movement, along with the various stages of a content lifecycle. And you see these kinds of workflow management applications that are inside of individual CCMS, right? So individual tools will have their own workflows. It's just that meta layer where we need to connect content across different functions and organizational silos when we're reusing component based content across an enterprise. And it does feel like as we move into a more coherent set of knowledge for an enterprise, that's not going to be just a bunch of unstructured content that LLMs train on and we build a big vector of unstructured blobs in the enterprise, right? It's going to be some level of structured and unstructured content working together, and it needs to be held together across some common infrastructure for sharing all of these facets of the content records that we need to leverage for different parts of the supply chain. For example, which languages does this content object need to be translated into?

And what are the timelines around it? When does it need to be refreshed? How do we evaluate value for this particular asset? In what ways can it be reused? What are the governance frameworks around this? Is this content subject to regulatory interrogation or review? And if so, to which standards does it need to adhere? All of those are forms of portfolio management that really are relevant beyond an individual silo. As organizations are working across lots of different customers that need to interact with many different lifecycles of their customer experience, presales and post sales with an organization across lots of channels, siloing content or just blobifying it and throwing it into an LLM and hoping for the best. Neither of those are the best solution. So there is a missing content operations framework to help bring all of this together. And Mike, I don't know how you're thinking about it, but in my mind, you know, as we've talked about in the past, it doesn't feel like this is necessarily one vendor. It's not like one consulting company solution or one CCMS or CMS vendor solution. It feels like that this is actually more of a framework based approach to content operations and supply chain that is really needed to be shared within the industry.

There's something underlying this whole dynamic of moving to a more intelligent way of working with structured and unstructured content objects and intelligent systems, especially with AI now our biggest consumer. How do we move towards this kind of shared infrastructure? I know that was a big, huge set of statements, not just a question, but in sharing ideas with you, my brain is on fire with realizations here.

Mike Iantosca
That's absolutely correct. If this is going to work in the industry we need, you might want to characterize as content mission control, right? And I like to joke and say, houston, we have a problem. We have no control at the operations integration layer and breaking those silos. So ideally, what we would have would be an electronic system here, right? A digital system which would consist of a back end data management and a front end portal, just to give you a vision. And it would be pluggable in the sense of it would be API constructed, API driven, and connector driven, so that we could take any of the major CCMs or CMS out there and plug that CMS into that framework, and we could take any of the CDPs, the content delivery platforms, and we can plug in the semantic management systems into this framework. And what you end up with is the ability to automate and orchestrate workflow across the entire lifecycle. And this is, as you said, a meta workflow. This is not the individual workflow of just one content management system. This is the overarching governance and orchestration and the planning. So a system like this would be fully pluggable and extensible.

It would be API driven and it would have connectors, ideally out of the box, that would connect to the popular systems so that any individual company somebody has Heretto, somebody else has IXIASOFT, somebody else has Paligo. It shouldn't matter what CMS they have or what CDP they have, they should be able to plug this in to this central system. And then with the central system, they can start creating the records, the plans, the content plans, and those content plans can be done individually, functionally, within the same system. So we could have a plan for the product content and a plan for the marketing content. And each of these plans contains the records of every publication that each functional group is going to create. But because all of these plans exist in the same framework, we literally can start getting an automating generation of how all the content fits together and have dashboards and metro maps and analytics that span the entire that total content experience. And so we're not forcing individual functions into a singular system, but we're sort of federating them together and giving them the flexibility to operate like they want to in terms of their functional needs.

But we can pull all of that and leverage all of that content and that entire system of record in a holistic way. So once we have that centralized content plan or plans system, we can start doing amazing things, because we can start aggregating all of the collateral and analyzing it and providing analytics and doing things like, oh, you've just created a new publication record or an update record to update, guide X and it needs to be localized. Okay, well, the planner in that record just clicks localize this by this date, and I need these four languages. And it spins off another planning record in the localization portion of the same framework. And the localization planner, with their automated systems, fulfill that without having to have separate systems and having to have people call each other or send emails. The whole thing has orchestrated messaging and email notifications and role management and accessibility capabilities. So it automates a lot of things. So that when a publication is finished, let's say we assign a publication plan record to a writer. The writer automatically gets a Jira notification and a Jira ticket to complete. And when they complete that transient record, we automatically publish the content to the staging system.

Nobody's involved, but people get notified, and then we have final approvals that are done through notifications and emails. But we now have a system of record that loops through the entire life of every publication that we can leverage. So if we could develop this, say an open source version, potentially, where any vendor can plug and play, even develop connectors to connect to this hub, we could move the entire industry forward in a unified way without having it locked into any one vendor's proprietary suite, if you will.

Cruce Saunders
Yeah, this is really exciting. The universal pluggable content supply chain. It really is something that I believe is going to need to exist in some form in order to develop this whole next generation of AI driven, personalized, really highly specifically relevant and enterprise ready interactions across multiple channels. There's going to need to be a way for the humans and the machines to work well together and to be able to get the various silos of technology interoperating against a common set of understandings around what is a content lifecycle, what is an individual workflow, what is the metadata associated at an individual object level and within group of objects, and how are those shared across tools and technology? This, I think, is something that every vendor stands to gain from. There's no content management technology provider who ultimately can retain customer value in a locked silo. That era has long since passed. Now the real value for any given vendor is how well does their content and their process and workflow connect to the ecosystem of their customers. 

Mike Iantosca
Yeah. Are you a player in the ecosystem? The comprehensive ecosystem? And that should be a value. And each participant, every vendor that participates in that ecosystem can offer different levels of value in that ecosystem.

Cruce Saunders
Yeah. And their value grows exponentially along with interconnection.

I think that that's a law of content, that the value grows with interconnection. So it's almost a square of the root value. The base value of any given content object grows with its interconnected exposure to systems and channels, which because then therefore it extrapolates exponentially the ability for that content object to perform work across an entire ecosystem of touch points, right? And so it has to be able to function across an entire ecosystem of technology. So I believe that ultimately vendors will support an open source, shared effort like this universal content supply chain. To me, that is one of the most exciting revelations in the last decade. I've seen a lot of really good progress made towards content standards and linked data standards and adoption across technology that's allowed content to be more portable and more discoverable. Things like DITA and schema.org and structured content schema, visibility and portability and search have all been innovations that have made content more available in more places and easier for enterprises to work with and leverage. But this next generation has to ante all of that forward, because now there's lots of standards, there's lots of systems, and there's a lot of AI consumers that need the content.

So we need a pluggable content supply chain as a shared level in order to take this next epocal shift from where we're at with our content management practices into this next era of true intelligence, where we're able to, at an object level and at a portfolio level, really orchestrate these composable experiences and help all of the humans, the authors, the creators, the designers, the storytellers, the business users, the strategists, to interact at a much more functional level, not just within a silo, but apply all their brilliant intellectual property across an entire knowledge base within an enterprise, the larger enterprise knowledge space. It's very exciting, and I'm really looking forward to seeing what happens next as this idea begins to form into something that I hope will become a shared industry effort to build a pluggable content supply chain framework.

Mike Iantosca
That's exactly correct, and you really summarized it well, because what this does is build workflow, automation and communications across all the different roles across the enterprise. Let's wrap it up with a big red bow here and just say that that's not the end game completely, because we didn't delve into AI today or gen AI today. But then just imagine with this enterprise operations supply chain, we begin to bring AI and gen AI into the operations process, because we do have a centralized handle on it and the many things that we could do with AI on the operations side of the equation. So why don't we save that for another conversation.

Cruce Saunders
Amazing, self healing, self correcting, self evolving knowledge networks that work seamlessly between humans and machines builds a brighter, more connected, more coherent world. Building knowledge together within an organization, across organizations and communities, and across the world. Thank you, Mike Iantosca. This has been an absolute pleasure.

Mike Iantosca
Thank you, Cruce. It's always fun to push the industry forward and move the needle in big ways.

Cruce Saunders
One step at a time, towards a smarter world.