From Web Pages to Prompt Fields: The Future of Digital Content Experiences

We consider the existential future of content on this week’s episode, with Phyllis Davidson — VP, Principal Analyst at Forrester — who doesn’t hold back in her Provocative Statement of the Day. We also cover everything from content experimentation and strategy to the future of the content landscape and role of generative AI, in this insightful interview with a content celebrity!

Primary and derivative content creation takes center stage as Phyllis unravels the success story behind the best content campaign she’s been involved in. Phyllis sheds light on why this model continues to be a game-changer for B2B enterprises, drawing from her rich experience as an analyst. She also advocates for more interactive content experiences to drive greater engagement. 

As we venture into the realm of experimentation, Phyllis explains the potential power it could have if companies stopped using experimentation just for content promotion instead of the primary content asset itself. Later in the interview, Phyllis challenges conventional norms, offering a visionary perspective on a future where content as we know it might cease to exist — a real Provocative Statement of the Day (PSOTD)!

Join us as we navigate these thought-provoking topics with Phyllis Davidson, a distinguished expert from Forrester, in an exploration that transcends the boundaries of traditional content discussions.

Watch the episode

Listen to the episode

Read the full episode transcript

Hello, Phyllis. I’m so excited to have you here on the show. Welcome!

Thank you, I’m excited to be here.

This is a big get for me. You’re as famous as it comes in the content world. I’m very excited to see where this goes, but let’s go ahead and jump right in with the quick fire. So, first question, what’s the best and most successful content campaign you’ve been involved in?

I would say that was back when I was with Oracle and we did a campaign — we considered it the Oracle story campaign. We wanted this big story about Oracle and this goes back a little way, but this is still an interesting idea. The idea was that by simplifying IT overall, you could drive up your revenue — which is a bold statement. 

It is.

Our big piece of content for that was done by The Economist for us in a research study to prove out as best we could that yes indeed if you could simplify IT you would see greater revenue. Kind of a big idea. Well we wrote this piece in the years before we talked about there being primary and derivative assets. But that’s exactly what it was. The report itself was a primary asset. 

But I squeezed more stuff out of that asset for a good 18 months. In fact, in addition to things like, the infographic, the short piece, the industry piece, all of that. I think one of the really interesting things to come out of it was a webinar series that we did that we referred to as “CIO to CIO.” And we had the Oracle CIO interview big COs all over. Where we could we did them live as fireside chats, recorded as fireside chats, and basically talked about this topic of the importance of simplifying IT. 

So we got so much out of that campaign at a thought leadership and brand level. But what we found was, it was also very successful at including calls to action that could actually get people into a buyer journey with us. So I would call that the most successful and it was nice to do something with a brand like The Economist.

Absolutely. I’m gonna come back to the fires. The fireside chat thing just hits me. I feel like, were they drinking scotch? Like it feels like a thing, sitting by the fire, sipping, holding a glass. Who came up with the idea that we’re gonna call two guys sitting on a couch for a fireside chat?

I wonder did the fireside chat thing come from early television and presidential, calm the people recorded presentations?

I feel like Will Ferrell would be involved in recording a fireside chat. In that Ron Burgundy style. I’m sorry, I digress. 

What’s the worst content campaign that you’ve been involved in?

I’m trying to remember exactly what the content was, but I was at a startup in the health business and we were targeting HR people. To be perfectly honest, I had come from a background where I was usually targeting IT or financial services. So the HR audience was new to me. And I made the mistake that I now, as an analyst, counsel my clients not to make, which is: I made some basic assumptions about what that audience, that HR audience, would want based on my own experience with different audiences. 

In my head I said, well, you know what? Everybody likes interactive content. Everybody likes a certain kind of event. And for that audience, the reality was, it was very different. What I needed to do was more buyer research and better understanding of that persona before coming up with content. 

In the not too distant future, there will be no such thing as content in the way that we think about it today. Click To Tweet

So the content in general, which we came up with, was interactive. I’m still a very big fan of interactive content. Of course, it’s changed dramatically in terms of what that is in the last few years. But the interactive content, like eBooks, that we came up with just didn’t fly with that audience. I don’t think I had the content right, and it’s on me because I did, again, what I actually counsel people not to do and made some assumptions about persona needs. And it’s all about persona needs. That’s what makes content either work or not work.

It is, absolutely! So out of those two, which one did you learn the most from? What’s the take away from that?

I think in general, we learn the most from things that don’t work. But in this particular case, I think I learned an important basic fact from what didn’t work, which is that you have to spend time getting to know the actual persona and you should just keep out of your head, your own experience. You can maybe bring that into it once you’ve understood the persona and inform that thinking with your own experience as a marketer. But an understanding of that audience has to come first. And I’m now such a huge proponent of that. I think having some of the bad experiences like the one I mentioned supported that. 

On the other hand, with the positive experience, I became a major champion of the primary derivative approach. In fact, we even did, while I was at Oracle, some really interesting workshops with our teams to teach them how to do that. How to make the most out of a piece of content that you’re investing a lot in, money, people, whatever. So I think I learned from both of them.

I think the take away on primary/derivative content, that model still works so well. It’s so important to us as a business right now, creating big, important guidebooks that really have very little to do with our product that will help to educate anybody — not just people that buy our product. 

They can then be customized to be very us-oriented in those derivatives, but people don’t come to work to read about us as a vendor. They come to work and ask questions on the internet about problems that they’re having and ways to solve those. If we can solve those problems, it’s more likely that people will pay attention to us. So that top level is the education piece that leads down to all of those derivatives that become much more actionable.

You need to look at your own data. Even if you don't have a lot of data, it's really important to understand what real buyers are consuming. Click To Tweet

Right. The best way to deploy a primary and derivative strategy is to think about it as not just derivatives that are a cut from the primary, but that are customized to continue with people through their journey where possible. So you can really hit both. You can customize something so that you can get it to a different audience. And you can also build on something, I would still consider that a derivative that goes to the same audience the primary piece went to, but actually pulls them further into a journey with you. 

I actually think too that given generative AI, I’m sure we’re gonna talk more about generative AI here in a moment. I think that gives us a way to create derivatives very quickly. I don’t believe that we’re gonna be autonomously using AI to put out content that we don’t read. But if you wanna take a primary asset and ask ChatGPT to turn it into a blog to add industry information, it’s certainly worthy of the first draft in many cases. So I think that we can deploy that approach even more utilizing generative AI. I don’t think it’s a matter of just saying, you know what, I don’t need to do any work anymore. I’m just going to let generative AI write it. I’m not buying into that.

No, it’s a fun idea thinking that I can just give a quick prompt and get all the content that I’ve ever needed, but that’s not been my experience thus far. 

No!

No, seems more challenging than that. And that’s something that you’ve been talking about, been researching, the concept of experimentation. Tell me a little bit about that.

Yes! Okay and I have to tell you, I read a book that got me so excited about experimentation. I’m kind of a nerd that way. The book is called Experimentation Works and it’s by Stefan Thomke, who’s a professor at Harvard.

I’ve heard of that, yes.

Have you heard of that institution? I think you probably have. I’m guessing, you know, I didn’t go there personally.

I also didn’t go there.

He’s got a scientific background and he realized that this experimentation and the scientific method are actually being successfully applied to business. He really went with that and covered some fantastic case studies. One of them is a well-known Harvard Business Review case study and it’s actually on booking.com. 

So this idea of experimentation, when we think about it in business and in particular in B2B use cases, certainly for let’s call it the last decade, there has been a lot of experimentation. It’s been primarily with the way you’re promoting an offer, right? So this is done, certainly on the B2C side. Perhaps it started there and it’s rather simple, it’s simpler I should say on the B2C side because you’re generally not trying to get people into a complex journey with you. You’re trying to attract them quickly. 

So early on, let’s use booking.com as an example, it was worthwhile for them to try different colors and banner ads, different placement of text in banner ads, different text all together. Very simple A/B testing, multivariate kinds of stuff also has been done with the promotional, we can call it content, but it’s not the asset when we think of you and I think of content as an asset. This is really just the promotional elements that are probably promoting either an asset or an offer. And in the case of Booking.com, that’s getting you right to their page to hopefully book something with them, right? They did literally thousands of experiments at any given time and then are moving very quickly to operationalize something that worked. What Thomke realized was that the scientific method was really underway already and he just exploded that across many more use cases. So it’s a really interesting read. 

But here’s what I realized. We’re really still even in B2B at the point when folks talk about customizing content and when vendors, CMS vendors, digital experience vendors, et cetera, talk about the ability of their platforms to customize content, the reality is what they’re talking about is customizing the promotion of content. I haven’t run across anyone yet who’s operationalized, operationalizing, and automating the customization of the asset. Because the reality is that it’s hard. It’s much harder to do, even with generative AI and other tools.

Adobe is doing a great job, so is Optimizely, at automating changes in elements of images using generative AI so that you can very quickly pump out a different banner ad, a different landing page maybe. That’s relatively easy. It’s still of course, keeping a human in the loop for the most part. They call that content customization. In the industry, it’s considered completely acceptable to consider that little element, a piece of content. If you can customize it and personalize it, that’s something that everyone wants to do. But when it comes to saying, okay, well, wait a minute, that banner is promoting an eBook. I wanna customize the eBook. That’s a different story, right? People aren’t doing that yet. I’m a huge proponent of the fact that it must happen. And this will be in the report that’s published.

We think of that as activity metrics, not performance metrics. But let's face it, in most B2B organizations, those activity metrics are considered performance. And that's a problem. It's one of the reasons why we end up just having a really hard time… Click To Tweet

I think of that as taking customization of content the last mile into the actual asset. We could also call it customization at the edge, right at the point of delivering the actual content to the target audience. That’s not really happening now. I want to let you ask me more questions, but I think that we’re going to see things move forward, not just with generative AI, but we have to begin to have some modular thinking about the way we approach content assets.

Well, and the infrastructure that would support that because what you’re talking about makes total sense and I actually see it amongst a number of financial services companies. Think advisory organizations like TD Ameritrade for instance, and I’m making this up for them. I don’t know that they’re doing this just to be clear if you work at TD Ameritrade. But the idea of creating content for all of the advisories that I support, all the individual advisors and small advisor businesses, and providing them with a very customized annual report, for instance, that’s for them and their customers. That’s a concept that they understand. 

But the infrastructure to support that, like I can easily create that content. I can probably even do it without humans being actively involved because there’s only so much that I’m gonna customize in that. But the storage, the service of that, the delivery of that content, 

The assembly.

That’s all much more complex and doesn’t exist in current CMSs. My favorite CMS is Paperflite. And so how do I automatically create replicated collections in Paperflite for individual channels of delivery that go down to the single person small company like that, we’re not ready for that yet. So the fact that I can create all this content, it’s the same story. We can create all the content in the world, but all the other pieces of the equation aren’t ready yet.

Right and I do believe that we’ll see organizations moving in that direction. I’ve certainly seen a lot of interest on the advanced client side. Honestly, many B2B organizations that I deal with are still barely customizing those promotional assets that I mentioned a few moments ago. B2B tends to lag behind. 

Understanding what works for real buyers when they're in a journey with you by actually looking at everything a group of buyers have done who actually you've closed business with. Click To Tweet

This is just one of the areas I find I’m always trying to help people with at the same time saying, look, don’t feel bad that you’re not doing this yet. Yes, you hear us talking about, I mean, everyone is talking about the fact that personalization is an absolute thing that must happen, right? We’re personalizing B2B experiences full stop. What does that mean though? It means so many different things. I think probably the hardest thing to do is to actually deliver that completely unique asset. That long copy, whatever it is, it could be a long video, it could be any number of types of content, but can you customize at that level? I think there are going to be multiple ways that we’ll see organizations starting to do that and vendors starting to offer options for automating that. 

I’ll tell you what, a place to look where this is happening is actually on the sales enablement side and in sales content solutions. The notion of modular content has been there for a while. And the easiest use case that makes sense, I think this immediately resonates with all of us as something that makes sense and seems relatively simple, is marketing or a sales enablement team is creating presentations for sales reps. You set them up in modules so that a sales rep can then say, all right, I gotta go talk to a government office about product X, Y, and Z. Let me pick from these approved modules in my seismic, right? And put together the presentation I need for that meeting. Oh, this is great. They’re already approved. They’re packaged in such a way that I don’t have to try to create a segue between one section and the next section. That’s already thought through for me and I can present something that I know is valid, is current, and so on. So that’s a nice example. 

On the marketing side, we would need that picking and choosing to be done in an automated way, but technology exists to do that kind of assembly. It’s a matter of taking that technology and connecting it to this kind of use case.

So if I’m a Forrester customer, I can read this research that you’ve done on experimentation. If I’m not, I should probably contact Forrester and see about becoming a customer.

I don’t know exactly when this is coming out, but I’ll be doing a blog that can be read by anybody on that piece as well.

Shameless plug for Forrester. Everybody should go talk to Forrester.

Everyone should talk to me. I love to talk to people.

Fantastic. You’re the tip of the spear of what’s happening in the marketplace. And I know you’re looking a lot at content-driven demand generation and how content’s driving pipeline. Can you talk a little bit about that?

When it comes to content driving pipeline, in fact, I just also released a piece based on the idea of looking back to see ahead. Basically understanding what works for real buyers when they’re in a journey with you by actually looking at everything a group of buyers have done who actually you’ve closed business with. We talk about the fact that you don’t have to look at every single case. If you take a handful of cases to see what is the trend in terms of what’s really being engaged with as someone goes through a journey with me. Because part of the problem is when you ask this question about what’s working, most organizations have no idea because what they’re doing is they’re looking at activity. So they’re looking at, okay, how many page views did I get? How many downloads? But connecting that to actual journeys, usually that information is known more on an anonymous basis. 

I think of that as taking customization of content the last mile into the actual asset. We could also call it customization at the edge, right at the point of delivering the actual content to the target audience. That's not really happening now. Click To Tweet

We think of that as activity metrics, not performance metrics. But let’s face it, in most B2B organizations, those activity metrics are considered performance. And that’s a problem. It’s one of the reasons why we end up just having a really hard time evaluating the value of content and offers. In terms of what we do see working, it’s a very connected story. What becomes more important than anything else, more important than the asset type, certainly depending on who your buyer is, there are certain asset types that work better. We always recommend to clients that they’re looking very closely at the preferences of their audience members, both for content type and channel. So this is largely about getting those things right with content that makes sense based on what they last consumed. 

So, each individual piece of content is less important than the way the content is connected in a journey. Now, granted, we know people aren’t on a linear journey. That’s okay. We still need to tell that connected story, even if people are moving back and forth across it. So one of the things that we always say to our clients is don’t give a tiny bit of attention to the call to action. Oh yeah guys make sure you throw in some call to action at the end of that brief. Uh-uh. The call to action and that connected story is what’s super critical. 

Back in the old days Chris, and I think I think you’re old enough to remember these days, but we used to actually do focus groups to try to figure out what’s working, and what do people want. Wouldn’t it be cool to do that again today, to try to get an understanding of, in particular, these complex buyer journeys? Can we just talk to a group of customers about what really worked? It’s hard to do. 

We do produce buyer preferences reports, and they’re extremely helpful. In fact, we cover something like 22 different personas. So our clients love these reports because it certainly is great information to help them figure out what are my buyer preferences. Let me inform that by Forrester Research. But here’s the thing that we always tell people, yes, we have this and absolutely you can use it, but your own data is more important. What’s actually happening is more important. 

You can think of it this way, if you had to sit down and fill out a survey from Forrester that was asking you a lot of really specific questions about what it is you need to know when you’re in the market for something, what your preferences are, when you look at this, what is your buyer journey, getting into a lot of detail. The way that you answer those questions in a survey versus the way you actually consume are probably not the same, right? So we recognize that and we always tell our clients, this is helpful, but this shouldn’t be the very basis for what you create. You need to look at your own data. Even if you don’t have a lot of data, it’s really important to understand what real buyers are consuming. So I’ll stop there on that question.

Amazing, amazing. We’re at my sometimes favorite part of this show. It’s the PSOTD or the Provocative Statement of the Day. It’s a thing that you think that maybe not everybody thinks that maybe starts fights every so often. But I’m gonna get out of your way. What is your PSOTD?

I love that you turned it into an acronym. I’m gonna have to write that down so I can post something about it. Okay. In the not too distant future, there will be no such thing as content in the way that we think about it today. Was that a mic drop?

When folks talk about customizing content and when vendors, CMS vendors, digital experience vendors, et cetera, talk about the ability of their platforms to customize content, the reality is what they're talking about is customizing the promotion of… Click To Tweet

That was dropping my pen. OK.

I know I still want to have a job in the future. You probably do too.

No, I do. But I’m very interested. Go on. What do you mean?

Okay, so let me back that up with a few statements. So look, the fact of the matter is generative AI, even though what I tell clients today is, don’t be scared, use this as a writing companion and forget about the idea that you’re gonna have generative AI autonomously and automatically share content it creates without review. That’s just not happening anytime soon for the majority of B2B businesses. And I feel like that’s okay. We can still get a lot from generative AI. 

However, taking the long view with generative AI across so many things, not just creating content, but everything that generative AI will create. And the way, this is the big part. The way that it’s going to completely change the digital experience for all of us. What’s going to happen at some point in the future is we won’t be looking at web pages the way we do today. 

Agree.

We’ll simply be going into a prompt field and asking for information. So my belief is that early on when that starts to happen, the information that we get will probably be essentially web pages. It’s just that we’re not going to go looking for them. They’ll come to us. The information we get may well be content assets. That’s in the beginning. Down the road, we’ll simply get a response that’s created in the moment by generative AI. It won’t be anything like content as we think of it today. And it won’t be assets. The whole idea right now of what we think of as assets and the importance of asset types, like a white paper. I think that will go away. I think that we’ll have chunks of content that answers specific questions, again, that’s created in the moment. 

Now, is this 10 years down the road, 20 years? I don’t know. I think that at some point, we’re not going to be typing in that prompt box. We’re simply going to be talking to it. I mean, let’s be real. We’re all almost there already. We can talk to our phones all day long and just ask it for information. So I just don’t think that our children, our children’s children are going to consume content in anything like the way we do today. So there you have it.

So I’ll buy your discussion. I think that makes sense, but I still think that there’s original thought as the genesis for all of this. So yes, they’re not gonna go look for the handbook on inclusive language as written by Acrolinx. Nobody’s gonna search for that. I’m gonna ask the question of how do I manage the creation of … shit, I see the problem. Why do I care about inclusive language? Because I’m not creating content, I’m telling the prompt to be inclusive. Wow, this gets cyclical really quickly. 

But there’s gonna be something new, like when you buy your hovercraft in five years, to Jetson your way to work, there’s gonna be a product manual. Even if you don’t use it or see it or touch it, somebody needs to write down where do I put the key in my new Jetsons car? So there’s still a need for that first original thought or generative AI can’t tell you when you ask the question because it doesn’t know. 

It's all about persona needs. That's what makes content either work or not work. Click To Tweet

That’s the safety net that I think content people need to see is that you’re not creating the white paper but that genesis of thought. Something new is born, is built, is created, and put into the world, and the only way that anybody’s gonna know what it is if the people that created it tell the smart robots that give us all of our information what it is and how it works. 

Maybe it means that technical and complex content creators continue to be very important, and marketing content becomes derivative of that. I only need to create one piece of content, technical as it might be, product documentation, and then I can say, hey, can you write this for marketing material? Can you create this as our answer for when somebody asks the question? Wow. Well, that rocks my world for the day.

Well, so hold on. So here’s the thing. It’s my provocative thought, it doesn’t mean I 100% believe it. Because you actually bring up something that has been on my mind as a concern. As we talk about all the different ways generative AI can help us, including coming up with themes and things like that, here’s the thing. Let’s say I’m a part of a business that’s decided to really go crazy with generative AI. And I let generative AI come up with a theme, create the content for the theme. Yes, I review it, but you know, not that involved. At some point for my company to have thought leadership, I have to have some new human ideation that will then be fed back into the machine. 

Yes, absolutely.

But if we, here’s the fear, if we get into a mode where we’re really depending on those machines, we risk losing our creative abilities to some extent. We can’t let that happen. So I think that the unknown is how it is that we’re going to maintain control. I mean, maybe this is a derivative of the big question about AI overall: How do we maintain control of the conversation and ensure that what it’s creating is built off of our ideation and that we continue to ideate. We need to continue to feed our own ideation. 

I think in terms of are marketers just gonna take stuff written, pieces of content that are written by the product people. There still has to be messaging. And even if we think of no asset world, just blocks of content that answer questions, in the background there still has to be all the messaging work. So what changes is what comes out of the machine. What goes into it’s still largely the same other than writing those final assets. You still have to have the messaging, the thinking.

Well, it has to be. So ChatGPT knows all the words, it’s read all the words, but it’s only read what exists today. And so let’s sit back and let it start writing. You can write everything now. I’m just gonna ask you to create a thing and you create themes and content and a whole nine yards. It’s gonna run out of things to say because it’s now using the things that it’s created as its source content. It’s just all rolling into itself until it’s just typing the word “the” over and over and over again because there’s nothing else to say. 

I made the mistake that I now, as an analyst, counsel my clients not to make, which is: I made some basic assumptions about what that audience would want based on my own experience with different audiences. Click To Tweet

So there needs to be the human in the loop regardless of bringing that new idea to the machine so that the machine can use that to create something usable. So if you really want to go dark, it’s sort of a matrix-like pod world of people just being held in suspended animation thinking and the computers are pulling the thoughts out of their head and using that as the basis for new ideas. So we just become feeders for the computer. I don’t know if that’s a thing. 

So that’s scary. 

It might not even be a provocative statement of the day. We might have gone off the rails.

I think somebody’s already made a movie about that.

Possibly, possibly. Phyllis, thank you so much for being on the show. This has been super valuable. If people want to find you, the best way is LinkedIn? 

Yes.

Fantastic. Excellent. Phyllis, have a great rest of your day. Everybody else, thanks for joining us.

Thank you, Chris.

Important links

• Phyllis Davidson – LinkedIn
• Forrester
• WordBirds – LinkedIn
• WordBirds – Home page