Podcast: Play in new window | Download
Subscribe: Apple Podcasts | Google Podcasts | Stitcher | TuneIn | RSS
Mike AthertonMike Atherton applies insights from information architecture to help content strategists develop domain models. These models help align content stakeholders and create a powerful way to organize, discover, and display content.
Mike and I talked about:
- his discovery in the late 1990s of the field of information architecture (solving what was then known as “the pain with no name”), via Peter Morville and Louis Rosenfeld’s book, Information Architecture for the World Wide Web
- how his work on the media archives at the BBC led to the development of the practice of domain modeling
- the origins of his domain model in Eric Evans’ book Domain-Driven Design book
- how domain modeling is essentially a research project, a way to pick the brains of stakeholders and help them (subject matter experts, designer, engineers, et al.) come to agreement
- how a domain model gives you a place for any/all content you have in that domain
- how domain modeling can work as a forcing function to figure out which content belongs in a digital product
- his definition of content marketing: “You’re not making content about your product. You’re making content about the things that matter to the people who buy your product.”
- how the ability to target users and customers is more advanced than the personalization and other techniques attempting to address it (here’s the DrupalCon talk on personalization he mentions)
- how well Dyson executes its content marketing program, mapping out a complete 1,200-day customer journey
- the need for a grammar for new practices like personalization
- the use of interactive narrative as an intermediate practice on the way to full-on personalization
- the difference in perception, definition, and application of “content strategy” between product content strategists, tech comms strategists, etc. – and how they can still be tied together
- how he and his co-author Carrie Hane strove in their development of Designing Connected Content to empower non-technical people to apply technical concepts in their content strategy work
- how interface design decisions should be informed by the stucture of content relationships, as described in a domain model
- how domain modeling permits more organic cross-linking and other navigation opportunities
- his thoughts on content marketing, for example, on poorly executed programs: “Doing a thing badly doesn’t make that thing bad.”
- how content marketers would benefit from shifting from a campaign mindset to “thinking about what’s useful for the long term” and how this can help “brands can become that ambassador for their subject domain”
Mike’s Bio
For over 20 years, Mike Atherton has been connecting people to content. A specialist in structuring information, he has chunked, pushed, presented and linked compelling content for the BBC, Huddle, and, in a different age, Playboy TV. Now a content strategist for Facebook in London, he collaborates with product specialists to build experiences from the terminology up. He recently co-authored the book Designing Connected Content.
Video
Here’s the video version of our conversation:
Thank You, Moz
We recorded this episode in a conference room in the Moz offices, a few blocks from Mike’s hotel. Thanks, Ashlie, Ida, and team!
Transcript
Larry:
Hi everyone. Welcome to episode number 44 of the Content Strategy Insights podcast. I’m really happy today to have with us Mike Atherton. Mike is best known – the reason I asked him on the show is because of his book, Designing Connected Content. I’ll tell you a little bit. Mike has- like a lot of us old timers in the field, he has a broad background. I’ll let him articulate his current role, but I was interested in your information architecture background and how you elegantly stitch that into the content strategy profession. How did you come to do that?
Mike:
Hey Larry. Thanks for letting me on the show. It’s good to be another old timer. Yeah, information architecture started for me … There’s a famous book that came out by Peter Morville and Lou Rosenfeld lovingly known as the polar bear book, and that happened at a time when I was cutting my teeth in the mid- to late-’90s. It was fascinating that it was an answer to the things I’d been doing intuitively as a so-called online new media graphics designer, as we tried to stumble to define them in those days, and then realizing that there was this reason why it was hard to figure out what should go in the menu and how things should be organized. That felt like a slightly different problem solving skill to laying out things on a page, for example.
Mike:
When this, I think it was once called the pain with no name about 20 odd years ago, was then dubbed information architecture, I thought, “Yeah, that’s actually the bit that I’m really interested in, more so than color palettes and typography and things.”
Larry:
Right. I remember well when that- it was ’98 when that book came out.
Mike:
Right.
Larry:
I remember thinking, “Oh, good. We’ve got-” because I was struggling similarly at the time. But didn’t information architecture- that has a different connotation and meaning. It came out of computer science and there’s a more specific meaning to it there, but I think that book was the first to adapt it to the web and …
Mike:
These things have come from library science and all kinds of different backgrounds. Again, it’s all about figuring out how things are different to each other and how they’re related, and that has defined my career for the last 20 odd years in different forms.
Larry:
Right, and stitching that together. One of the insights that I really appreciated about the Designing Connected Content book is how you took those information architecture insights and that whole field, really, and just elegantly stitched it into the content strategy practice. It starts with … The big insight that I really appreciated was the domain modeling and what a powerful tool that is in so many ways. How did that come about? Did you … I infer from what you said how it might have happened, but tell me how that came to be.
Mike:
About, gosh, 10 odd years ago now, I was working at the BBC. The BBC have some of the best TV archives in the world, amazing, rich multimedia TV and radio content going back 50 odd years, and yet their web strategy at the time was to have all these small little silo teams attached to different TV shows, regional and national TV shows and radio stations, individually making what we would think of as effectively microsites and flat HTML. They weren’t really connected up together. There were all these little different web presences like tents on a hillside. I think it was 200 top level domains at some point.
Mike:
And then, what happened politically was that the BBC was going to get the digital resources cut in half, so they suddenly had a lot less money to play with, and somebody had the bright idea, “Hey, we’ve got like 50 years of amazing content in the archives. The web is now advancing to a point where video is almost becoming the default content type. Why don’t we start to stitch this stuff together?”
Mike:
A bunch of people that I was introduced to, and we’re very fortunate to work with, were taking this technique domain modeling whereby you take a particular subject area, be it history or gardening or food or anything, so the theme parks …
Larry:
The theme parks, I think of yours.
Mike:
… and you say, “Okay. What are the things in that world? What are they called and how do they connect to each other?” So a theme park brand, they’d have the typically resort and that resort might have a number of parks like you get in Disney world and you have places like Magic Kingdom and you have at Epcot and those kind of things. Within each park, there’s a whole bunch of lands, and the lands have attractions, and attractions have characters, and characters have creators, and the attractions also have designers. And rather than the strictly hierarchical system of old, what you get is what we’d call today a graph, a very wibbly-wobbly jumble of stuff that are connected in very specific ways. So you can say that this attraction was designed by this imagineer, Winnie the Pooh was created by A. A. Milne, and that bit in the middle, that predicate, is the thing that teaches you how those two things are connected.
Mike:
Through that, we could map out any subject. We could map out natural history, which, of course, the Linnaean taxonomy goes back 200 years or more. We could map out gardening, we could map out TV programs. And better still, all these different silos within the BBC, we could start to stitch them all together because all these different domain models could be developed independently but connected by the things that they had in common.
Mike:
There was a great example where we showed the TV show Mad Men, who’s got that great ’60s soundtrack and stuff, and the metadata data that we were getting back from the TV playhouse system for the show gave us the track listings and we could make all of those track listings linkable to artists in the music domain. So you could click on Bob Dylan in the track listing on a TV episode of Mad Men and be taken to the Bob Dylan page that shows all his albums and all this and all the times it’s been played out in radio and all that amazing information. As someone who also nerds out about theme park architecture, it was exactly like that. It was all these hubs and spokes that were all richly connected together in ways that just made sense.
Larry:
Exactly, yeah. It reminds me of back when I did entity relationship diagrams for designing relational databases.
Mike:
Oh, absolutely.
Larry:
Very much, but not exactly that, right? There’s that camel case connection you use to connect entities. That’s kind of like the way ERDs work, but not exactly, is that true?
Mike:
I think that’s the important bit of it. Domain-driven design in Eric Evans’ book, which is a huge doorstop of a book for software engineers that practice this thing, that was when those named relationship values came into ERDs. We’re getting very off-topic from content str-
Larry:
No, no, no.
Mike:
Content Strategy podcast. It was having those values that makes the relationship explicit. We’ve all seen in websites and stuff, you’re reading an article and then it has a bunch of so-called related links, and there’s no real clue as to how they are related or what the relationship is. You can intuit it often as a human, but you can’t tell that to a machine that this person is married to that person or this person, lived in this town or whatever, that would give meaning to those relationships.
Larry:
Got It. You’re right, that does go a little far afield. But that was the core insight that I loved about your work is that you pulled that whole field into content strategy and make it actionable. Because you don’t have to understand all that deep level stuff to do a domain model and to … And I think one of the powerful things about that domain model, to me, anyway, it seems like one of the best hacks for stakeholder alignment, one of the key tasks for any content strategists. Was that inten- How did that emerge, that insight that you’re getting literally everybody on the same page?
Mike:
Yeah. Well, from 20 years of consulting and working in different corporations, it was very clear that whatever the discipline, the number one complaint people have in the field, is basically how do I let the company do the job that they’ve hired me to do and how do we get them to agree on things.
Mike:
But what you find often, and I found this a lot when I was doing more UX design and what have you and wire framing and all of that, is that you’re typically asking- you’re putting people in their uncomfortable place. You’re getting some business executives into a room and you’re painting a bunch of wire frames on a wall or something. It’s like taking a drawing to your mom and saying, “Do you like it?” You’re asking people to judge things on criteria that they’re not necessarily comfortable with.
Mike:
So what’s good about the domain modeling technique is, essentially, it’s research. So rather than going into a subject matter experts with a page layout or even an article or something and saying, “Can you review this for me? Can you judge my picture?,” what you’re doing is you’re picking their brains at the things that they know about so you’re satisfied. “Tell me all about 401ks and how do they work and how are they different to other investment plans.” And you get them on the thing that they’re excited about and knowledgeable about, and they feel that they can then contribute to the process on their own terms.
Larry:
One quick thing about that. I want to go in the weeds a little bit on that. Are you typically working- There’s a lot of ways this can unfold, I’m sure, but physically getting people in the same room to having a subject matter expert, the technical people, the creative people, the business people, all …
Mike:
How we’ve done it is that we’ve- there’s various ways. We’ve done lots of interviews where you would do one-on-one and you have particular things that you want to know. Meanwhile, and particularly if it’s a subject domain that you as the content strategist don’t know a lot about, maybe personal finance or something or a particular health issue or something, you’re doing a lot of research yourself so that you have the right questions to ask.
Mike:
And they’re helping unpack your understanding and flush out any complexity. “Are these two things the same? Are they different? If they’re different, how are they different? Is this always true or are there some exceptions to it?” Like you would interrogate anything. And you’re building up this picture. Once you have a sense of, “Okay. I think this is how it works,” that’s the time to then get not only your SMEs, but also engineers and designers and the people who are going to build the thing, into a room and then it starts to be post-it notes and sharpies on the board or bits of string to connect the two things together, and you start mapping. And that is a collaborative and often a creatively combative exercise where you’re picking apart how things fit together and what’s always true and where the relationships are.
Mike:
But you’re only doing that. You’re not trying to do that while also trying to do layout or trying to do navigation or anything like that, or even trying to deal with the content that you know you’ve got. You’re just modeling your world. And then the step after that is, “Okay. How do we pin bits of content from what we’ve got to this?”
Larry:
So any one piece of content is just a detail about that world.
Mike:
Ideally, yeah. The theory in the book is that domain modeling is a great forcing function for getting rid of content that shouldn’t be there or producing content that should. If you take the premise that … There’s that quote from Howard Gossage, the ad man, that says, “People don’t read ads. They read things that interest them and sometimes it’s an ad.” Well, people don’t read content or watch content because it’s content. They do it because they’re interested in the things in the content. If you buy into that, that they’re watching this video not because they like videos but because they like LeBron James or basketball, then you start to think, “Okay. My customers are interested in a particular world. What does that world look like? What are the things in that world?” And if I make content about each of those things, that will be useful to the person that I’m trying to serve.
Larry:
Got It.
Mike:
And that, to me … There’s a lot of snippiness of content strategy versus content marketing, but that’s what I think content marketing is. It’s against the regular marketing.
Larry:
Right. Quick aside. How would you define- because I think, to my mind, I have a pretty well-established definition of content marketing in my head.
Mike:
Right.
Larry:
It’s non-pushy, non-hypey, objective content that’s designed to illustrate you as an expert and maps to a customer journey.
Mike:
Yeah.
Larry:
Does that mesh with your …
Mike:
Super fancy. It’s John Deere with the furrow in 1895. It’s P&G making As the World Turns. You’re not making content about your product, you’re making content about the things that matter to the people who buy your product.
Larry:
You just reminded me. A friend shared on Twitter … She was at a presentation, she was in Europe a month or two ago, and I think you had a slide that said people don’t care about your content, they care about what your content is about …
Mike:
Right. And that’s it. Because anyone who’s engaging with digital content, particularly from a brand or something, that isn’t their end intent. It’s a means to a bigger end. It’s because they want to buy a car or it’s because they want to plan a pension. They’ve come with some particular information seeking task. Your job is to engage them and also to keep them as a customer. Peter Drucker said the purpose of any business is to create and keep a customer. Well, to me, keeping a customer is about keeping them interested in the thing that makes them a customer.
Larry:
Right.
Mike:
Again, John Deere, they’re not about keeping people interested in agricultural machinery. They want to keep people interested in agriculture.
Larry:
Plowing a field and harvesting crops.
Mike:
Yeah, because those are the things that keep them interested in the stuff that the company sells.
Larry:
Yeah.
Mike:
We see it all the time now. Modern brands are doing this. Airbnb are doing it with city guides. MailChimp do it when they put out these guides, not on how to use MailChimp, on how to do better email marketing, because they know that their very tightly-defined customer segment has particular challenges in particular jobs to do. And that’s where I think content marketing is that brand relationship drive. Because it says, “We get you. We understand the things that you’re going through, and we can help you with that.”
Larry:
Yeah. Stitching that back to content because … well, a couple things about it. First, I’ve talked to so many people over the last few years. Literally, everybody I meet in any kind of a business context, I’ll ask them about what does the term content strategy mean to them.
Mike:
Right.
Larry:
80-90% of the time, they think it’s a marketing thing.
Mike:
Yeah.
Larry:
It’s embedded in it, but it’s also this scalable thing. You can be strategic about content marketing content but, ideally, it’s happening in a bigger context, right? Is that …
Mike:
For sure. Yeah, you’re right. People have quite nebulous definitions of both content strategy and content marketing. I think a lot of people in the content strategy community think of it like things like Outbrain and clickbait stories and the stupid lightweight microsites for FMCG brands that we all used to build in the ’90s and stay out far too late from agencies, for things that would never be seen by anyone other than the client who commissioned them.
Mike:
But what I think content strategy’s real beef with is actually marketing. I think content marketing came about to almost challenge that wastage in marketing to say, “Okay. We could do this smarter. We could actually deliver something of value and use that as leverage in the relationship building.”
Larry:
Yeah. This is reminding me of some of the threads- before we went on the air, we were talking about some of the other inputs into content strategy.
Mike:
Yeah.
Larry:
A couple things just occurred to me there, kind of Seth Godin’s permission-based marketing. And then, there’s this flip to permission-based, customer-focused marketing.
Mike:
Yeah.
Larry:
But also in the computer world, the whole HCDI and other fields, where human-centered design as a discipline. It seems like a lot of that stuff is converging.
Mike:
Yeah, those are certainly … I’m in town for DrupalCon and there was a great talk by Jeff Eaton yesterday about personalization and the signals to personalization. Yeah, we’re starting to be able to build up a very targeted picture of what people want and we can start to serve them advertising or what have you. At the moment, that exists in rela- the targeting stuff is pretty advanced, but what we do with the targeting is still quite unsophisticated. We’ll just throw an ad at them or something, rather than have some valuable, useful and usable content in the wings to to do that.
Mike:
And not every brand. I have a great case study in one of my presentations about Dyson, the vacuum cleaning people.
Larry:
Oh, the vacuum. Yeah.
Mike:
They map out their entire customer journey for, I think it’s 1,280 days, and they know that after the first week of purchase they’re going to hit you with an email to register your guarantee for the product so that you’re covered. They know that after three months, you’re going to need to change your filter, so they’re going to send you an email to remind you to do it. And then later on, there’s things like just household tips, how to get a wine stain out of a table and stuff like that. Because their relationship with you is about being house proud and household clean, and you bought an expensive vacuum cleaner so you obviously care about keeping your house tidy. And it starts to drift more and more away from being about the product to being about the world of the product.
Larry:
Right. What you just said, and just thinking about … and I think personalization is one of the things that a lot of people are concerned or thinking about in the field, but it’s going from- it’s almost looping back around. Especially coming out of the UX world, which is a world you’re very familiar with, the development of personas has been so important there. But now it’s almost like we want to just start … the personalization is undoing the personas and getting down to the actual individuals.
Mike:
Yeah, yeah. And I don’t know whether we’ve got really a grammar for that. We’ve never had the tools to the extent that we do now to target people. So things that have become before personas and then before that broader demographic research like Nielsen type stuff have always been fairly blunt instruments, and businesses have been built around those blunt instruments because that was the best that they could do. Yeah, when it comes to very hyperlocal, hyper-personalized content, we can absolutely identify the person or the people, but what do you do with them? What do you give them? We haven’t quite got the high pick personalized content.
Larry:
Have you come across anybody who’s do- because I had a guest on recently, Neal Sofian from a local [crosstalk 00:21:07] called Tuzag, which does medical … They’re a medical concierge service, where you come in as a consumer and they get the usual demographic stuff. They get psychographic stuff from your behavior and then they have location data and weather data and all that kind of stuff. They stitch that together and then they actually serve up personalized content based on that. They’ve been working on that for 15 years, and they’re just now finally getting there. Are you aware of others who are beginning to figure out ways to put together that personalized content so we can do more sophisticated …
Mike:
No, not as such. no. There’s a few people who are dealing with interactive narrative. There’s Avox, that are doing that kind of thing where you can start to involve people in a progressive user journey and personalized content along that way, but it’s not quite the same thing as targeting, “Here’s the perfect, completely, absolutely for you. This is your Tarot reading. This is something that’s absolutely down to the individual.” It could increasingly be, talking about content strategy, that that becomes a necessary skill set. Tech comps for years and years have talked about content chunking and reuse and all that kind of thing and how everything is a dynamic remix of content chunks.
Larry:
Right. Well there’s-
Mike:
Still, that feels a very crude tool, to start to cut and paste something together based on these Barnum statements, almost, that are supposed to apply to a particular person.
Larry:
Yeah. Well, that gets into the whole world of DITA and the highly structured technical content.
Mike:
Yeah.
Larry:
It’s stitching that together with these marketing targeting.
Mike:
Oh, good Lord. Stitching that together even with the rest of content strategy is challenging at the moment. Content strategy as a field has these- I’m sure there are many camps or different ideas of it, but there are people increasingly within the product space that define it as effectively in-product content development and content guidance and microcopy and that kind of thing. And then there are people who have come to it from the tech comps world, and they’re all about structured metadata and DITA and schema.org and that kind of stuff. Again, I think there’s jostling within the communities because these people aren’t quite having the same conversation and aren’t quite sure whether they’re supposed to be.
Larry:
Exactly.
Mike:
But I like to just cherry pick the best of all of it and see how it rolls.
Larry:
And you’ve been so good at that, the way you stitch it. Because a lot of what we’re talking about here, those seem like plugins to your domain modeling.
Mike:
Yeah.
Larry:
The idea of figuring out how to incorporate Schema into the overall- just have that be a part of any content strategy.
Mike:
Absolutely, yeah. The hardest part is making it interesting to the people who are increasingly positioned to run with it. We’re at a stage where nobody has an information architect within their product team these days. All those kind of skills have fallen between engineering, maybe a bit of PM, potentially content strategy. But if content strategy as I would like to see it within the product space also incorporates taxonomy and terminology, the whole ontological kitten caboodle, what is this thing and how is it defined, then it does start to get under the hood and it does start to get into these things which bleed out into software development for the terrified kind of thing.
Larry:
That’s a great- that’s probably a book already. But it seems like the work that you’ve done with and describe in Designing Connection Content, it seems like a lot of the stuff you’re talking about really is just additional practices that could be incorporated into that.
Mike:
That’s right. When we were writing the book, Carrie and I, that was one of the questions we were asking ourselves. Who is this for and at what level should it be pitched? We wanted to aim it at people who were jobbing content strategists, so who understood words, understood language, understood the problems that they were having, the things that they were coming up against, these boundaries, where it’s like, “I can’t quite structurally get upstream enough. I believe that somehow linguistically this product could be better if it was under the hood, but I don’t know quite how to get there.” And a lot of content strategists, particularly in those teams, are struggling for legitimacy and a seat at the table and all of that, still.
Mike:
So we wanted to pitch this book in a way that was explaining these things, which seem quite technical to somebody who is non-technical, in a way that they could be empowered to act on them. And you can do it all with pen and paper or post-it notes and sharpies and that kind of thing. You’re never actually writing code or anything like that. But it’s thinking in that structured way that engineers do and [crosstalk 00:26:30] do.
Larry:
Yeah, and then you can come in. I want to circle back to one thing. I’ve been looking for an opportunity to get it back in, it just hasn’t come back around.
Mike:
Yeah, yeah.
Larry:
You talked earlier- when we were talking, you mentioned Karen McGrane and I immediately thought- I did this little brain loop. Anyway, I’ve been working on a blog post, and somebody’s free to steal this idea, Death to Lorem Ipsum. It’s a common sentiment in the world, in this content-first world we’re emerging into, but a lot of the conversations about that, like when I Google that phrase and read stuff, there’s a lot of conversation about … for the most part, people are like, “Yeah, fake content is no good. You want to model real content.”
Larry:
But Karen made an interesting point in a discussion about that, that sometimes you want either gibberish or a way to explain that something is low fidelity because, as you’re prototyping stuff, it’s sometimes better to just have something meaningless if you’re talking about one thing.
Larry:
The way this relates to that whole … I think a lot of the content-first conversations I’ve heard are at the page level or the individual chunk of content level. My hunch is that you would think content-first, but content in a nicely organized database that’s nicely mapped and … Is that …
Mike:
Yeah, yeah, for sure. And not to disagree with Karen at all because I think that’s true. I think it depends what it is that you’re getting feedback on or you’re trying to assess, and if that’s purely layout, then you don’t want to get people hung up on the specifics of the content, particularly in a client meeting or something. But, yeah, my sense is if you have the content in place first, that can organically drive the design of it. I should correct myself. The interface design of it. Because that comes from the structural relationships in the content.
Larry:
Right.
Mike:
One of the big wins of having your content described à la a domain modeling approach is that you’re not limited to this top-down, big bucket classification. You can have much more organic crosslinkings and cross references between things, a bit like you do on Wikipedia, for example. And if that becomes your primary mode of moving between things, that’s going to influence how you design an interface around it. You don’t use Wikipedia with big top-down menus or anything like that, or drill downs. You either search or you’re clicking from one thing to the next, and that’s the experience.
Mike:
So I believe … and it’s not about factionalism or one team having to own it over another, but I believe that if what you’re fundamentally presenting is information, is content, be it user generated or published, then that has its own form and it has its own way that it wants to be expressed. And that comes from how it’s related to one of them.
Larry:
Yeah, cool. Well, thanks. I’m glad I circled back to that.
Mike:
Yeah, yeah.
Larry:
We’re coming up on time. It always goes so fast and I could talk forever, but I always give my guests … Is there anything last, anything that I haven’t brought up, anything that’s just on your mind about content strategy or any insights you got at DrupalCon or anything like that, anything last you’d like to share with my folks?
Mike:
Yeah. I think content marketing gets a bad rap a lot of the time. Sometimes it is done badly, but doing your thing badly doesn’t make that thing bad. But what I’d really like to see, where I think there is a limitation, is rather than thinking in traditional campaigns and that sort of short burst of activity, thinking about what’s useful for the long term. What kind of stuff could make you an authority on a subject and how you build a brand over time. And that’s where I think we can really start to see these worlds come together, where business goals are aligned with user goals, are aligned with some differentiation of the information that’s out there, and brands can become that ambassador for their subject domain.
Larry:
I agree. That sounds like what a brand will probably be in the future.
Mike:
Right. I would like to think so, yeah.
Larry:
Well, thanks so much, Mike. It’s been a great conversation. I really enjoyed having you here.
Mike:
It’s a pleasure.
Larry:
Cheers.
Leave a Reply