Hey, and welcome back to The Interline Podcast. 

Digital asset management has always been one of those things you either spend a huge amount of time thinking about or you just don’t think about at all. And the difference between those two scenarios was usually down to job role. If you worked in enterprise IT, there’s a good chance you were thinking about digital asset management (DAM) in an architectural sense. If you worked in marketing, e-commerce, retail partnerships, and marketplace interactions, any place where you spent your days sharing visual assets for downstream use cases, DAM was the platform you probably lived in the most. If you worked in the middle, in product creation, DAM was something that happened down or sometimes upstream of you, not something you interacted with or thought about all that often. 

But thanks to some sweeping changes in what constitutes an asset and what it means to manage them, digital asset management as a software category and as a practice is now on a lot more people’s minds, especially in product creation. 

The idea that a digital asset is increasingly a living object or a combination of living objects that have value in themselves and in their relationships to one another is one that’s gaining a lot of traction in apparel and footwear came up in our DPC Report 2026, from just ahead of the holidays last year, in fact. 

So today, I’ve brought on someone who definitely has spent a huge amount of time thinking about DAM and a huge amount of time working on where it crosses over with digital product creation. And that’s Kara Van Malssen, who is a Partner and Managing Director at AVP, which offers platform agnostic consulting services for companies looking to get the most out of digital asset management. 

We also interviewed Kara in that DPC report 2026. So this is our opportunity to go even deeper on a topic that can sound a bit dry and inconsequential, but when you think about it, is actually about the fundamentals of how our industry really operates. So let’s head over to my chat with her.

Listen to the latest episodes / subscribe above, or read on for the full transcript.

NB. The transcript below has been lightly edited.

Okay, Kara Van Malssen, welcome to The Interline Podcast.

Thank you for having me. It’s a pleasure to be here.

No, the pleasure is all on this side of the table. So when we’re going to be spending any amount of time on acronyms or kind of industry lexicon, I always like to start by defining those things early. And I think we’ve got a really big definition to try and get out ahead of at the outset with this one. And that is: what exactly constitutes an asset? 

So in the years I’ve been doing this, which is 16 years now of analyzing technology for fashion, I’ve certainly seen a pretty big change in the way people frame the word ‘asset’. And you can get philosophical here, but, generally an asset until pretty recently has meant something that was fixed – a finished visual, for example, a finished piece of text, finished video, something that needs to be formatted or converted for different use cases and different applications and profiles. But otherwise it doesn’t change. Its form is fixed. 

Now it feels like an asset is coming to mean something different, something that’s alive – like complex objects or assemblages of interrelated objects that are not just being archived and then served up in an optimised way, but being shared as part of ongoing workflows and part of product creation processes, as well as far downstream processes.

Are you seeing that shift? Because I’m guessing you’re more hands on and more on the ground than I am, but it’s definitely something we’ve observed.

Absolutely. We’ve certainly seen a shift, especially in the fashion space, in the apparel and footwear industry. And for most of the history here, an asset was something you made after the real work was done. It’s a campaign image or a final product shot. So it’s something static that you optimised and you maybe moved around, disseminated to different places, like you said. And what’s changed, especially in the product creation space is that some of the most valuable assets now exist before a product even exists physically in the world. And like you said, they keep changing. 

So with that has been a shift from assets as final output to actually assets as input and foundational data infrastructure within the product creation lifecycle. So if we go back to the foundational question here of what is an ‘asset’, that term means that it’s something that has value to the business. And so in this case, we say digital assets, but we also have to be careful with that because if you Google the phrase ‘digital asset’, you’re going to get different definitions. Some of these are financial definitions, you know, thinking about a little Bitcoin and blockchain space. They use the phrase digital assets over there. 

We typically are using the term digital assets to mean creative or visual digital assets, especially in the fashion space. And so the creation and proliferation of these visual assets is kind of inherent within the way that apparel and footwear are being created today. And so we’re seeing that shift, like you said, where we have a lot of different types of assets that are going into the process of creating a product, and their key decision points that come up early in the creation of that product which will ultimately be a physical product. But like I said, these visual assets are kind of coming to life prior to that product entering the world. 

But the other thing I want to mention here is that, you know we said an asset is something that has value to the business. And so that differentiates assets from just plain old files. And I think that distinction is important. So some level of stewardship is really what elevates files to assets, things that have enduring value to the business. And so those are things like metadata, governance, reusability.

So without some level of stewardship, assets actually can quickly become a liability – without clear versioning, status, ownership, rights. And so it can start to slow things down or even risk misuse. So, I think what we’re seeing in that shift is there’s a proliferation of digital assets, which are just inherent in the whole digital product creation space. It’s kind of synonymous with digital assets. 

I think so. Humour me for a second, then. So let’s picture a garment for argument’s sake. It can be a handbag. It can be a piece of footwear. Just a product. Give me one or two examples of what would constitute an asset in the non-traditional definition then. I could easily reference finished catalogue shots, finished static renders, that kind of thing. Just give me an example of what you’re talking about. 

Where are these things that merit stewardship and that need governance and that need control? What kinds of things are we talking about? Are we talking about materials? Are we talking about components? Are we talking about data? Just give me some rough outlines for pieces.

Yeah, absolutely. So in the DPC space, we kind of think of a style as a container for many component 2D and 3D visual assets. So that includes, like you said, materials, trims, and a lot of those things are reusable across different products, right? So it’s not just that that material and that trim are specific to that style. It’s being used across multiple, but it is being used in this style. So we need to capture that, we need to record that and associate those assets with the style. Then, of course, you have things like logos, graphics, CADs, and your colourways and then patterns, construction details, 3D protos, all those types of things. Some are, like I said, specific to the individual product and style and some are things that are reusable across multiple. 

So I think one of the key aspects of this is if a visual can be reused, referenced, compared or evolved across seasons or teams, it becomes that type of asset that I’m talking about. It becomes part of a library and it needs to have that level of stewardship and management.

I think that’s fair. I think that traditionally you would have used that framing for… I’ll pick a photo shoot again, like a finished photo shoot where you can take one or more of the same images or similar images from a campaign and reuse them across multiple other campaigns. That would have some element of reusability, but only within a fairly narrow lane, only within fairly well circumscribed applications. What you’re talking about here are things that have that level of enduring value to the particular product or to the particular collection, but that are also reusable, extensible, and composable for a bunch of other different assets. A lot of these are things that are not created and then kind of ramrodded down a particular road but things that actually have broad utility.

And so the utility and the kind of broadness of those lanes vary a little bit. If you’re thinking about materials and kind of material libraries, that has obviously broad application within your brand. If you probably have specific materials that you go to over and over for multiple products across multiple seasons. But then when you get to, you know, a specific style and the CADs for that style. Now maybe those CADs are for this season. If we carry over that style to the next season, those CADs are going to become useful, but we’re not using those colourways anymore. So maybe now we have used those as a reference, we can then replace the colourway as long as that style is not changing in any way, those CADs are reusable. So it’s a little bit narrower than that broad application of those materials, but yeah, it definitely has application across seasons and across products.

Okay, great. by the same logic, the logic we’ve just described, given that the nature of an asset has changed, or at least the framing of it has changed, the definition of digital asset management must have evolved as well, not just because of that change of framing, but because of what that change of framing means for the importance, the impact, and the growing scale of those kind of reusable value-packed assets that we’ve talked about in different stages of the product lifecycle. 

So we’ve defined what an asset is for our purposes. Let’s try and define what digital asset management is for our purposes today. And maybe just give me your perspective on how that definition’s evolved in the last few years.

Yeah, so first I’ll just start with saying DAM is a capability and a practice, not just a platform. So a lot of times when we talk about digital asset management and we say DAM, we mean a tool. And of course it is a tool, but I’d like to argue that there’s more to it than just that. It’s a people, process, technology, data, triangulation that we need to keep in mind. ⁓ But that DAM capability is something that enables organisations to trust digital product representations at scale. 

So the definition of DAM and how it’s evolved, like you said, it’s similar to how the definition of digital assets has evolved or expanded or what we can consider digital assets is much, much bigger than what we used to think of. And so the definition of DAM, the role of DAM and where it sits within the organisation has evolved. But it’s not just because there’s new file types. It’s not just because now we have 3D and we didn’t used to. It’s because DAM now kind of sits at the core of product creation and not just downstream in the marketing side of the house, which is where it’s historically and traditionally lived. So DAM as a capability and as a set of functionality, at the core, that’s a pretty timeless concept. And that really hasn’t changed.

You know, we have a central solution, location, platform, something where all the digital assets are stored, managed, and can be disseminated. But what’s changed is, like you said, DAM used to be kind of a destination, an archive, a place where teams went after work was done. And now it’s more of a dependency where other systems and workflows assume it’s reliable and can leverage the data that it contains.

So I think one of the big shifts is, digital asset management systems were kind of islands, siloed from the rest of the organisation or from the other parts of the ecosystem. So if we think about a decade back or a little bit more, that was pretty true. But now it’s part of the workflow. It’s integrated into the business process. And so in that sense, it’s really foundational for DPC. And I’ll argue it’s often one of the missing pieces. And actually I was listening to some of the other episodes of the podcast and I realised I’m not the first person to suggest this. You had a guest from H&M last year who said one of the biggest gap areas for many brands was not centralising asset libraries into a single DAM because it’s very difficult to scale without doing this. And I wholeheartedly agree with that. So digital asset management is this underlying foundational capability that allows you to centralise visual assets as data, along with textual information and integrate that into the workflow to enable scale.

Yeah, so we mentioned digital product creation and 3D a couple of times already. I know that’s not the sole catalyst for this shift in framing of what constitutes an asset or what constitutes DAM, but it does seem like the growing uptake of having 3D representations of clothing, footwear, and accessories that capture a product in active development or refinement or iteration by capturing its constituent parts. seems like that growth in uptake has been a driver behind some of this framing. 

Tell me what that blend of finished asset management and then work in progress product creation actually looks like in practice when we think about the kinds of products that we are designing, simulating, visualizing, fitting, sampling and so on in 3D and what it looks like to manage the governance for that kind of not just a complex multi-part asset, but a multi-stakeholder, multi-stage workflow that’s built around those assets? Because you’ve already said there’s more to this than just putting different file types in the system. That’s not the shift that 3D and digital product creation represents. It represents, I think, a different way of thinking about sharing and collaborating and iterating.

So what do you think about that? And give me an example of what that kind of workflow looks like if we put a 3D shoe, a 3D garment at the centre of it? 

Okay, sure. I think the difference is that we’re talking about using visual assets as tools to help manage decisions in motion. So if you’re developing that shoe, you know that there’s probably gonna be several milestones that you’re hitting across that go-to-market life cycle. And during those different milestones, you’re kind of iterating on the development of it. So it starts out kind of at a low fidelity and then obviously needs to get to a place where the accuracy and the fit details and pattern need to be, you know, perfect for manufacturing purposes. So over that entire process and through those different milestones, these visual assets are playing a key role in helping with various decisions by different stakeholders.

So it’s like the DAM kind of becoming the connective tissue between those workflows, those users and those decisions as a central place where all of these assets and all of these key details about them can live. And so the ability of a DAM system to capture not just the assets themselves, because you can just do that in SharePoint or wherever, but metadata about them. And it might be metadata about the status of this thing within, you know, something that’s specific to a certain milestone or something that is, you know, relevant to it, to a specific part of the approval or things like that. So, if we take this product, this imaginary shoe and we think, okay, it’s a new style or a new season. Maybe it’s a carryover from a past season. Maybe it’s a brand new style. So from a design input perspective, there’s a designer, they’re likely going to the DAM, they’re looking for certain templates, working tools like blocks, like for footwear lasts or avatars for apparel and white models potentially for the footwear design. So things that have been approved for reuse and for fit. And so those kind of working tools are there for them to use. And then you have things that carry across multiple styles and products like logos, et cetera. So the designer is able to get those things from the DAM. They know that they’ve been approved. They know that they’re valid for their use.

So then when they’re done with their outputs, again, multiple iterations of design outputs, because they’re doing this through multiple milestones, they deliver those back to the DAM. And so that can be done through integration with the design tools. So maybe your CLO, your Browzwear, or Illustrator, whatever tool you’re using, kind of putting the assets back into the central platform rather than saving it on your local desktop, your Drive, your team’s Dropbox or whatever. So it’s going to this more central tool because you as a designer are not the only person or role that needs to use those files because you now have product developers who are gonna be needing to put together final tech packs. They’re gonna be getting those assets through the design process, patterns, construction details, those logos, and they’re happy to be getting those through the PLM. So through integration with PLM, they’re able to grab that stuff, get it into their final tech pack. 

And then kind of thinking back full circle to visual line planning, if you’re at the beginning of the season, next season, we can pull in the pallets, the CADs, the renders from the past season to kind of help with that next season’s line planning. So it kind of becomes circular and cyclical. So that’s just like a few examples that come to mind of the types of roles, the types of touch points, the kinds of things they’re going to be looking for using, putting back in etc. during that process.

Okay. And kind of related to that DPC discussion is the idea that’s floating around that 3D digital product creation is on the wane. And that generative AI offers a shorter route maybe to at least some of the key use cases that companies have put at the core of their DPC deployments. Primarily the idea that if you want to bring an idea to life, visually, and communicate that with people, 3D maybe represents a long circuitous and quite labour intensive way of doing something that generative AI in theory at least offers a straighter shot to. I’m sympathetic to that to some extent, but I don’t think it’s as simple or as universal as people make it out to be. And I think the framing that you have here, which is assets as assemblages of living objects that workflows and collaborations are built around, actually spotlights the fundamental difference between 3D and AI. 

But before we move on from talking about 3D assets, what’s your perspective on that? Do you see DPC strategies being paired back, being scrutinised? Do you see that kind of push for people to replace some of this with AI? You’ve got a good vantage point on the industry and I’m keen to see what that looks like.

Yeah, my perspective from having watched the industry from the last few years is that there’s not an either/or here. There’s a complementary thing to be considered. 

I guess, first of all, I just say I’m aligned with the viewpoint that DPC isn’t just 3D. It’s just any type of digital asset that contributes to the product creation lifecycle. That could be 2D, that could be 3D. So I don’t think it’s just 3D, but just in terms of the 3D aspect of this, my read on the situation is there has been some level of right-sizing to using 3D where there is the most value rather than attempting to use it fully across the product lifecycle. Rather than it being the solution to everything, let’s do everything end to end in 3D, kind of bringing it to the workflow where it makes the most sense, where you have the most capability in terms of talent, skill, infrastructure, et cetera. And then where you need the type of information that 3D can give you that 2D really can’t. 

And again, rather than having to go straight to physical product, is there a space where 3D can kind of fit in and help answer certain key decisions in the product development process? And I think, yes. So from my perspective, I see brands are still using 3D to a great extent. It just might be shrinking or right-sizing a little bit. 

So I don’t think we’re going to see it go away anytime soon. It may be a matter of updating the definition. I’ve seen organisations previously talk about their work as DPC, and now they’re just calling it product creation. And maybe that’s a recognition that, you know what, there’s 2D, there’s 3D assets in that process. It’s not just 3D. So they’re kind of like neutralising that definition a little bit. 

Yeah, I would agree with that. And I think it’s almost inevitable on any kind of tech maturity adoption and diffusion curve that eventually something is not sufficiently new that you don’t need to talk about it as its own thing anymore. It just becomes so well integrated into the way that you work that it’s kind of unremarkable. I get that. So I think I agree with you there.

And I think we’re not there yet with AI, but we need to be.

No, we’re definitely not there with AI. That certainly needs to still be spoken of as its own thing. Tied to everything we’ve just talked about, I think is also the idea of a pretty fundamental shift in the user base for DAM systems. Again, I’m using the software sense rather than the practice and the discipline here. You know, marketing and communications teams, they’ve always been primary users as dictated by the concept of fixed assets. But now that we’re talking about extending access to the same tools and workflows and philosophies to everybody in product creation, designers, developers, engineers, sourcing teams, and so on. Does that change the way these systems are designed and implemented? Does it change the roles they’re intended to play? And what have you seen when it comes to mapping these kinds of user-based expansions and use case build outs to existing DAM platforms? Is it the case that we have a lot of legacy solutions that were built with that old audience and that old definition in mind, and that it’s now an uphill struggle to update them? Or are we actually in a reasonably mature market where there are a whole bunch of DAM platforms and workflows built around them that recognise the more living definition of what constitutes an asset?

I think we’re somewhere in the middle from a marketplace standpoint. A few changes from the platform side, from the marketplace, the industry in general. So yes, DAM as a solution was primarily developed for marketing kinds of use cases. And feature development has kept pace and aligned with this. This is still true today. And so if you talk to a lot of the DAM platform vendors out there, “MarTech” is the only language they speak. They understand marketing. They have, you know, spent their entire careers, sales reps and others at those companies learning how marketing technologies work and how they fit into it. So that’s still true, but there have been a few platform vendors out there that recognise they have built a product and a platform that is capable of being molded to a wide variety of situations and use cases, it’s not just for marketing. 

And so we’ve seen DAM enter into a lot of really interesting spaces. So not just marketing, but we’ve seen it enter, for example, medical use cases like for surgical video being captured and brought into DAMs and then used for kind of AI analysis and for teaching and for research. This kind of thing was never envisioned when people were building these DAMs. So I think digital product creation and apparel and footwear creation is kind of a similar shift. And so there have been a few vendors who recognise that there is a need for DAM in this space. They’ve responded by shifting some of their positioning and marketing and targeting this market to demonstrate how these platforms can be adapted and configured to the product ecosystem.

And some of the best DAM platforms out there, like from the enterprise DAM space, have a great deal of flexibility. So they can kind of be configured to support those use cases with their current capabilities to a good extent. So a lot of the fundamentals are the same, you know, the same features – you need versioning, storage, metadata, derivative creation, sharing and a flexible data model that allows you to kind of build those product containers and those relationships between all these different components, those reusable materials, et cetera, et cetera. 

I think the biggest shortcoming is in the 3D support aspect. So the DAM platforms, you know, have not been asked to support a great deal of 3D. And as I’m sure you well know, there’s a lot. It’s kind of a wild west out there in terms of formats. And so that’s a lot for them to try to respond to. They’re all pushing for, why don’t you guys normalise and standardise on one file format so we don’t have to support every single proprietary design tools flavour. So that’s one of the areas where I think we’re considering, we’re gonna expect to see some evolution. 

Some of the main DAM platforms have started to build out 3D viewing, interacting, rendering, et cetera. So that’s evolving. But there’s also 3D DAMs that are popping up in the market with a lot of cool functionality and native integrations with design tools. But there’s the risk of siloing if you kind of go to a tool like that. If we back up and think about product creation as this combination of a variety of different types of assets, whether 3D or 2D, if we kind of go forward with a 3D DAM, we’re really covering part of that world.

That’s a really good point because I think when you think about 3D-specific capabilities for DAM, so just the couple that come to mind because I’ve seen them referenced over the years are, you know, automatic load creation through texture compression and mesh decimation and stuff to get 3D assets to where they are usable across a bunch of different platforms. That’s only useful when you truck with 3D. There’s very little there that is transferable to 2D. There’s very little there that kind of cross-pollinates across to your other different business functions or anything. And I can understand the draw of those kinds of capabilities if you are a company that does a lot of 3D. But equally, I feel you there. I think that’s something that people ought to be aware of and to take into their decision-making is that improvements to 3D only capabilities. It’s not a tide that raises all boats situation.

Yeah, I think what we’ve seen some brands do is put those 3D DAM tools kind of upstream a little bit further, closer to the designers’ space and aggregate those specific assets and then output, you know, the right flavour to actually go into the enterprise or centralised product creation DAM. So they can kind of get the best of both worlds. It’s more tools in the ecosystem, but I think that’s an option that can work. The question is: will we see more adoption of these 3D-specific DAMs or will we see better development of the 3D capabilities within the traditional DAMs? I don’t know which way that’ll end up going.

My next question might also be a “don’t know” situation, but when we think about the user base for any platform or solution, the natural following question to, is the user base growing, is how far is that user base crossing over with the audience for other platforms that might be feature competitive, capability competitive, or at least pressed into service to do the same thing or to do something like it? So when we were talking about just static assets, I don’t think anybody would argue with the idea that they should reside in a PIM or a DAM solution rather than in PLM. But now that we’re shifting the window to encompass a lot more in the way of living objects, I can see how a company that spent a lot of money on a PLM platform or someone who’s embraced some of the kind of PLM-like or PLM-lite functionality that’s been built into 3D designer simulation tools might say to themselves, well, why can’t those live assets stay where they are? You know, I am doing my product creation in PLM and CLO or Browzwear or what have you, why do I need something else? Where should the assets live? What would your answer be to that? Is it integration? What is the question? You know, brass tacks, why not there? And why in DAM?

Yes, so if we go back a little bit to that concept of, DAM as fundamental foundational infrastructure for all this, I mean, one of the key reasons to think about that is that DAM isn’t just serving human actors, it’s also serving machines. And those machines need access to all the assets in one place in a way that they can easily access and reference and do things with. So some of those machines are APIs, some of those machines are AIs. And so when you have the assets sort of staying in one place, because there isn’t one place where they could all stay other than the DAM, right? Some of them could be in the design tools. Some of the design tools have very DAM-like kind of repositories built in with them, right? So they could maybe some of them stay there. You’ve got other things that live in PLM like construction details and things like that. But those systems don’t house all of the assets, whereas the DAM, that’s its role. And so allowing those assets to flow into a centralised platform that the other tools can draw on for different purposes also allows for this sort of machine access and use as well, which I think will be key to starting to think about how generative AI may be able to enter the product creation space as well. If the assets are siloed, if the different types of assets are siloed in these different tools and platforms, it’s going to have a hard time leveraging those. 

So, in short, integration is probably the answer because these tools are not DAMs. The capabilities just are not the same. Permissions, version control, file transformation, dissemination, they’re limited in a system like PLM. It’s not built to do stuff like that. It’s built to do one thing with those assets, put them in a tech pack. You know, for example, it’s not built for these other needs. 

And the other thing I guess I’d say is, you know, not all participants in the product creation lifecycle are at home in PLM. That serves the product development teams, material developers, but it’s not where the designers hang out. So we’re going to end up with these silos if we take that approach. And I think that’s the risk: how do we scale? How do we allow for machine learning? How do we train our generative AI models? How do we kind of allow the flexibility of reuse in these other systems if we don’t have them all in one?

I think that that’s the answer, but I do see a lot of confusion around this. And this question comes up a lot, like why can’t everything just live in the PLM? And a lot of people don’t understand what DAM is, what the role it can play to scale workflows. And so they also just maybe don’t realise there’s a lot of assets out there that don’t end up in the PLM at all. And those are in 501 SharePoint, you know, repositories, like laying around the organisation. So there’s that big mess that can pop up when there’s no kind of central home.

Yeah, I think luckily we’re talking at the time of year when people are in the kind of mood to try and get stuff all in one system as opposed to keeping it in multiple places. 

[Laughs]

You find me very much in that frame early in the year. It’s the time when people really try and audit their solution landscape. 

You mentioned AI and I’ve got two questions about it. And the first is related to the inputs to digital asset management, because it seems like we’re staring down a potential explosion in the volume of assets that can be created by removing the overhead that used to be required to create them. You know, as we talked about earlier, whether we’re talking about 3D, 2D, it’s easier than ever to create a huge volume of stuff, visual stuff specifically. And that huge volume of visual stuff is also not separated into parts and layers and components. It’s not something you can pick apart in PhotoShop, it’s not something you can pick apart in CLO. The output of generative AI is fast and flexible, but it’s also fixed in its own unique sort of way.

Do you see a role for DAM in maybe stepping into this whole space around AI provenance and the explosion of assets there? I do have another question about AI, which is more related to the AI and the capabilities of DAM, but this one’s been weighing on me, which is when we’re talking about objects and assets and we’re sitting at a juncture where it’s easier than ever to just create avalanches of those things, feels like something’s happening there.

Right, I think it comes back to that point earlier, that definition of asset versus file and the stewardship that defines whether something’s an asset or a file. Because if we accept that everything that a generative AI is going to output is an asset and then therefore should be staved and stewarded, we’re going to have kind of compounding debt or asset debt. We’re gonna have metadata debt. We’re gonna have version control. We’re gonna have all kinds of debt. We’re not gonna know what those things are, what their value is, why they’re here, you know, what are we doing with all of them? So in the DAM world, we have this concept of selection, you know, what is selected? What is kind of in scope for long-term reuse? And those are the things that we need to save. So I think part of it will come down to that, making some policy decisions around which generative AI outputs need to be saved and which are just one-offs, you know, to be recreated. So that idea of libraries, what are the libraries that the generative AI needs to be able to create a good output, those things we need to save. But the outputs themselves, not everything needs to kind of go in the DAM forever and ever.

I think DAM can help with the provenance and traceability. Even if we’re thinking about capturing things like model, prompts, rights, source, you know, what was the thing that created this image, that could be captured as metadata within a DAM for the outputs that are determined to be worth saving. So I think, yeah, just two viewpoints on that is: 1, what needs to be saved? And then 2, how can we capture the information about the thing that made that thing, that asset in a way that is going to help with the provenance? 

I like that. That’s a really good answer to that question. 

And the second AI question is, what role do you see for AI within the capabilities of DAM itself? I think you hinted at this a little bit before, but just go a bit deeper. If I was out there shopping for a DAM, seeing a lot about them having additive AI capabilities. How would I distinguish the useful ones from the non-useful ones?

It’s no surprise that the vendors are investing a lot here as all kinds of platforms and technologies are. I think we’re still going to see the answer to your question shake out over a little bit longer. You know, what we’ve seen in the past two years is a shift from, this feeling like, you know what, most of the human aspects of DAM are going to be gone very soon. Humans won’t need to be creating metadata, humans won’t need to be managing rights, humans won’t need to be doing XYZ. You know, agentic AI concepts came along and kind of pushed that even further. And then people started implementing the tools and they realised, you know what, this is not going to replace my job tomorrow. In fact, in order to even use these tools, we have to have so many strong fundamentals in place in terms of metadata, taxonomy, rights, permissions, et cetera, that we’ve got a lot of work to do here. 

So there are things that the DAM can do that the AI capabilities are getting better and better at. You know, there’s a certain amount of metadata generation, normalisation. This has always been the bane in the existence of the DAM manager – keeping up with all the metadata that needs to be created to really make the assets valuable. So there’s some improvements there, but it’s not a silver bullet. It requires a lot of training, which requires a lot of data. 

So I think there’s some powerful search capabilities within a lot of these platforms that are AI-driven these days. So AIs are good at pattern recognition and we can see prompt, chat-like interfaces popping up in certain places to kind of help with search asset, you know, simple stuff like asset deduplication, things like that. 

Right now, it’s just at a place where it’s kind of incremental improvements. And I’ll also say that a lot of AI or really machine learning is not new in DAM. There’s been a lot of things there for years and years around image recognition, OCR, speech-to-text for time-based media and things like that. So it’s some of the generative capability. And then of course, agentic is coming into these platforms fast, but what workflows are you going to apply it to? You know, a lot of them are going to be things around quality control and then things like that. 

So it’s not magic and it’s not always super sexy stuff, but hopefully we’re going to see AI helping with some of the tediousness. But we’re still at a place where if you need accuracy and trustworthiness and consistency, you know, we don’t always see this consistent outputs of generative AI tools, of course.

Yep, I think that’s a very fair summary. Last question is, beyond what we’ve just talked about, all the new AI capabilities of varying degrees of utility that are being added, what do you think is coming next for digital asset management as a technology category and as a discipline? Anything we haven’t already covered that you think is going to exert a serious pull on things over the next one to two years?

I think we are going to start seeing the DAM as a source of data for AI to learn from. I think that’s a big shift that’s coming and has not been there historically. So when we’re creating assets, when we’re aggregating them, how we’re configuring the system, how we’re configuring metadata, you know, it’s not just for human users anymore. We need to be thinking about that as for machine users and having these tools be and the data sets they contain be available for AI tools to learn from is going to be a big shift coming. But there’s a recognition of, like I said, the fundamentals have to be in place for that to work. So a lot of organisations are still kind of at that foundational stage. Like, let’s actually get this stuff really organised. And it’s things that people like DAM admins and librarians within companies have pushed for for years.

But the shift to AI and the push from leadership to adopt AI may be finally the catalyst that allows, that provides the resources for companies to put what they need into the platforms, in terms of fundamental data foundations that they can build off of. Maybe it’ll be something positive that kind of pushes these things into the future. 

So that’s my prediction: DAM for machines, not just people.

I think that’s a pretty good one. And I think that’s also something that we can potentially pick up in our AI Report this spring. I’m currently very much on the hunt for topics that are going under-recognised and kind of under-acknowledged on top of the ones we already have planned. And I like that one. I think there’s some legs to that story and I think we can cover it. 

Kara, no more questions from me. I’ve really enjoyed talking to you. This is a really interesting and evolving space. I know we’ve talked about digital product creation, we’ve talked about AI, we’ve talked about digital asset management. There’s a lot all kind of swirling together here that seems pretty fundamental to much of what I think the fashion industry wants to accomplish. I’ve enjoyed this conversation very much on that basis.

Yeah, thank you very much. I know digital asset management is not the most sexy or exciting thing for a lot of organisations, but it’s important and I hope everybody finds this useful.

Well, The Interline’s audience is the right one for it, I think. We’re used to these sort of back office, less outwardly sexy discussions. 

But, Kara, thank you so much for your time. Hopefully we can pick up this conversation at some point in the future.

Yeah, great. Thank you so much for having me.

And that’s my chat with Kara done. I hope we managed to define all the key acronyms, terms, and ideas there because I’d hate for anyone to be put off thinking about and evaluating digital asset management at a time when the definition and the importance of digital assets is at an all-time high. 

As a companion piece, I’d encourage you to grab The DPC Report 2026, free from The Interline – not just for the interview with Kara herself in the second half, but for some of the editorials and brand case studies which really highlight just how much re-interrogation of the idea of what an asset is, is really going on right now. 

Much more to come next week and the week after and so on and be sure to listen to any back episodes that you might have missed. We’re covering a lot of process and technical ground with the show and my plan is to keep expanding which parts of the fashion technology spectrum we get into. So there should be something for everyone in our back catalogue and there should be something for everyone coming soon as well. 

For now, thanks for being a listener and I’ll speak to you again soon.

Ben is a senior fashion technology storyteller, analyst, and moderator with more than a decade's experience in retail.