Podcast: Play in new window | Download
Subscribe: Apple Podcasts | Google Podcasts | Stitcher | TuneIn | RSS

Sam Bhagwat is the co-founder of Gatsby, a popular framework for creating content experiences.
He is also the author of Modular: The Web’s New Architecture. If you’re not sure how the terms “headless,” “decoupled,” or “composable” might affect your content work in the future, the book can help you understand these concepts and the technical ecosystems that enable them.
We talked about:
- the Gatsby origin story (we recorded this just before Gatsby was acquired by Netlify – more on this below)
- his book, Modular: The Web’s New Architecture
- the tech and design trends driving the adoption of modular architectures
- the ways that modular architectures help different kinds of businesses – merchants, publishers, and marketers
- the “solar system” of different classes of development tools that support modular architectures
- workflows and practices that support the federation and orchestration of modular systems
- the future of content roles in a modular landscape
- the benefits of thinking of content both as data and as code
- how coding practices like version control and multiplayer collaboration are likely to be adopted in the content world
- how the emergence of a search utility like Algolia shows the benefits of modularization
- his hypothesis that the move to full modularity may retrench a bit in the near term as parts of CMS functionality are re-bundled
- the excitement he sees around “new integration and orchestration ways of pulling our existing pieces of the puzzle together” – including Gatsby’s new Valhalla product
- a note regarding the Netlify acquisition that Sam shared shortly after our interview: “The news this month in the Jamstack world is Gatsby joining Netlify. We hadn’t finished the acquisition talks when Larry and I recorded this episode, so unfortunately we couldn’t discuss them. But our conversation should give a lot of insight, since we’d joined Netlify for a lot of the reasons Larry and I talked about. The last decade of the modular web has been about creating the right primitives. The next few years will be about integrating them together and making them easier to use.”
Sam’s bio
Sam Bhagwat is a principal engineer at Netlify. He’s the co-founder of Gatsby and author of Modular: The Web’s New Architecture. Prior to founding Gatsby he was an early engineer at Plangrid and Zenefits. Sam lives in Sacramento with his five-year old son and three-year-old daughter.
Connect with Sam online
- email: sam at gatsbyjs.com
Video
Here’s the video version of our conversation:
Podcast intro transcript
This is the Content Strategy Insights podcast, episode number 136. For most of the history of the web, content practitioners have managed their work in integrated content management systems like WordPress, Drupal, and Adobe Experience Manager. Many organizations are moving away from these monolithic products to headless CMSs that are a part of an emerging modular ecosystem for building web experiences. Sam Bhagwat’s new book, Modular: The Web’s New Architecture, can help you understand and navigate this new way of working with content.
Interview transcript
Larry:
Hey everyone. Welcome to episode number 136 of the Content Strategy Insights Podcast. I am really happy today to welcome to the show Sam Bhagwat. Sam is the co-founder and the Chief Strategy Officer at Gatsby, which is a… Actually, It’s not a static site generator. What do you call Gatsby, Sam?
Sam:
I would call it a modern JavaScript framework for content sites.
Larry:
For putting sites together. And your role there, you co-founded it and you’re now the Chief Strategy Officer. A couple quick things. What led you to found Gatsby and what is being a chief strategy officer there entail?
Sam:
Yeah, so rewinding back a few years in 2015, my best friend Kyle Matthews created an open source JavaScript framework to build websites based on React. So this was really early in the React days and it started taking off, among developers that were kind of excited to build with this new framework. And Kyle was very early in that community, very much felt that React would be the future of web development. I started getting involved as Gatsby became more and more popular and one day we just turned to each other and said like, “Let’s make a company here. It seems like we should be doing this.” And that’s how Gatsby as a company got started.
Sam:
So these days now, fast forward five, six, seven years and it’s pretty clear that React and modern JavaScript is going to play a huge role in web development. I think that’s pretty broadly accepted. The Jamstack, the composable web, the modular web, goes by a number of names, is getting just a ton of mainstream adoption. And so right now at Gatsby I mostly spend a lot of time with our customers and prospective customers helping people understand how to put all these different pieces together.
Larry:
And one of the ways you’ve helped people put it together is with that book you wrote. You just wrote a book recently called, Modular: The Web’s New Architecture: (And How It’s Changing Online Business). Did that book arise out of all that education you’ve been doing?
Sam:
Pretty much. I mean, we were having a lot of conversations with folks who were getting into this world and folks had come from a lot of different backgrounds. Some folks had come from the content world, some folks who come from ecommerce world, some folks would come from more of a startup engineering world. And each of group of folks understood certain things about the architecture but also didn’t understand certain things about the architecture. After we had a lot of similar conversations, we wanted to tie it all together with a little bit of the history of how all these pieces came together, as well as a practical guide of how to use the different pieces together.
Larry:
And I love that. I mean, there’s a lot to about the book, I’ll just say. I really appreciated it and enjoyed it. But one of the early in the book you talk about the rise of the modular web and you talk about all these things that we’re all kind of aware of, but they’ve come together in a way that creates the need for modularity. Can you talk a little bit about some of the key inputs into that?
Sam:
Yeah, I mean I think there’s almost like a Fellowship of the Ring kind of moment here where you just get a lot of different groups that coming together that are aligned in a lot of different veins. For example, the rise of kind of JavaScript on the web, especially the JavaScript ecosystem, the importance of web performance, the SEO update that Google did a year and a half ago I think is maybe the best example of that, mobile especially and the need for performance on mobile specifically. There’s a lot of design trends that have come along that that’s sort of taken a while to shake out with mobile that are pretty important. The cloud and SaaS in general. The proliferation of all these kind of services for search and authentication and of course have the CMSs. And just the proliferation of all these different services.
Sam:
So there’s a number of different threads. I mean then the gradual steady rise of the ecommerce over the last decade being another piece. And so there are all these really different pieces of the puzzle that when you put them all together have created this environment where it’s now possible to assemble a system for running your website and your online presence from all of these different pieces where that was much more difficult to do five or seven years ago.
Larry:
Right. And all those things you mentioned, just cloud and SaaS provides an infrastructure for a lot of this stitching together. It’s so interesting to me because JavaScript was around for 20 years before it really took off. Was it the React framework do you think that really… I don’t follow that world very closely.
Sam:
There were a couple things. So the 2000s were kind of a lost decade for JavaScript in a lot of ways. But starting near the end of that decade, there was just a bunch of different things that happened at the same time. There was an open source kind of standards committee that kind of got their act together after a long time and agreed on some sort of syntax improvements that really helped move the developer experience forward. There were things like Node, so the server side run time and NPM as a package management system that also originated around that time. And there’s been a lot of things since the React and Vue and similar kind of component based frameworks that came around the middle of the decade that kind of accelerated that trend. But I’d say that, really, the movement and maturity of JavaScript started around the beginning of the 2010s.
Larry:
Which is also the same time that mobile is becoming a real issue to deal with. And if I recall correctly, that’s around the time that headless CMSs kind of became a thing as well. I wonder if there’s a chicken and egg thing there. Did these things emerge because they were needed because of each other?
Sam:
You never really know. But you’re right though, I mean cloud and SaaS happened at that time, JavaScript started to get good, and mobile happened at the same time. In some ways those trends just stack on top of each other in terms of driving momentum for this modular world that we’re in now.
Larry:
Yeah, and that’s the point. Regardless of the exact sequence and relationship between all those things, it’s clear that there’s a need for a modular architecture for the modern web. Can you talk a little bit about what that looks like? A lot of my listeners I think are come out of publishing and journalism in places that probably just have a WordPress site or Drupal or maybe an enterprise site and they may have seen a Jamstack site or something like that. But can you talk a little bit about that whole environment of the modular web that you’re talking about?
Sam:
Yeah, definitely. And I think it’s different for different kinds of sites. So it’s going to look different for an ecommerce site, it’s going to look different for a more content driven site, it’s going to look different for a news site. But I think there are some things that are fairly universal across different industries and use cases. I think one of the terms and phrases that I use in the book is this idea of a solar system, I was actually sort of talking with the Contentful CMO at the time just about how this architecture was unfolding. And I’m over here as an engineer, kind of turned founder, and she’s there as experienced marketer, been a marketer for 25 years and we’re just kind of talking about this and how should we talk about it? And I’m like, well, it’s like a federated architecture. You’ve got all these different pieces of the puzzle that are kind of connected to each other. And I’ve start drawing with my hands in the air and she’s like, well maybe it’s a solar system. You’ve got a few different planets and then each of the planets has moons, and the moons orbit the planets, but the planets are all independent sort of systems.
Sam:
So I put that analogy in the book and I think that’s a really good way to think about it, the solar system idea. In terms of how the architecture is constructed, you might have an ecommerce system, and let’s just say you’re a Fortune 500 company or something, and you might have an ecommerce system that’s powering a store and you might have different types of review and discount and promotion and shipping and logistics types of systems that feed into your ecommerce. You might have your headless CMS. But then you might have other types of systems like internationalization systems that feed into that CMS. You might have your development framework, which is pretty closely related to the hosting platform that you’ll be using, so that might be a Gatsby. And then React, which is your develop… And your various components as well, if you’re using maybe a Storybook, and I’m throwing out a lot of these sort of frameworks and technologies and tools.
Sam:
But you can see that, well, the development tools might be one planet and the ecommerce tools might be another planet and the content tools might be another planet and there might be integrations between these various things, but each of these different puzzles in some ways you have to figure them out one at a time and then you pull them all together to create your overall architecture and your overall technology set up.
Larry:
Yeah. A few minutes ago you used the word federated. I’ve also heard you talk in the past about the content mesh. And then, I don’t know if you used the term, but I think often of the role of orchestration of that you solar system you just mentioned. What’s your current thinking on that? How you orchestrate, how you pull all this together?
Sam:
Yeah, I mean the short version of the story is that there’s a lot of different workflows that you’ll want to think about. This is one of the things I spend time talking about because a lot of folks who are going into this role have the same question, how do all these pieces work together? I kind of get what they are individually, but how do they work together? And so there’s a lot of these pairwise integrations that help the system hang together.
Sam:
So just to give it a couple of examples, often you’ll want to feed your products and your ecommerce system into your CMS if you want to store auxiliary information inside your CMS. And so CMSs and ecommerce systems often have integrations with each other to enable that. You often want to be able to see… Inside your CMS you want to preview the content and what it looks like on the site before you push it live. And so, we spent some time building this, some hosting platforms have preview functionality that integrates with the CMS to provide that sort of experience and workflow for content users. And you can go down the list, but there’s often an integration between your component library and your site, your ecommerce data also gets pulled into your site. And so we have these kind of reference architecture diagrams and I have a couple in the book of, hey, here are all your different systems and then here are the integrations between your systems that are the most important ones that you’ll want to be thinking about.
Larry:
Yeah, I’m in a co-working place so I have access to a printer, I actually printed out your solar system pages ’cause I wanted to have it handy as we were talking. And so I Scotch taped it together, so that’s one way you can orchestrate it. But one thing as you were talking just now about the way you stitch together like a CMS with your ecommerce system or the ability to preview, each of those like the preview function, that could be just a fairly a simple small unit of functionality. I guess one of the things I’ve encountered or heard about in the modular systems is this kind of need for “just rightness,” how you could atomize everything and decouple everything and have these little tiny modules or you could have a giant monolithic CMS. We’re obviously somewhere in between those. What’s the mechanism for just rightness?
Sam:
Oh man, that’s a such a philosophical question, but a good one. I think I’ve seen it kind of emerge over time a little bit. People just get a sense of how they want their systems to play together. Some of the integrations between CMSs and ecommerce vendors for example, happened within the last couple years as the CMSs realized that was something that their customer base desired. We were pretty early on with preview again two, three years ago, but that’s become more of a popular workflow as well. So I think over time people get a feel for it, but it’s definitely not something that… The systems don’t emerge fully fledged into the world with that just rightness.
Larry:
No, and you’re reminding me too that it’s a craft. Most content people I think would associate themselves with some content craft like writing or design or engineering or strategy around their content practice. And so I see that kind of emerging role of the orchestrator of this. Which gets me to one of the real big things I wanted to ask you about because you’re out in this world so much, especially for content practitioners. What does the future hold? Will there be new jobs like content orchestrator or something like that? But how do you see content roles unfolding in the future?
Sam:
I don’t know. I mean, that’s a good question. I think one of the things we see in terms of technical trends for the headless CMS world is that people are viewing content in two ways, they’re viewing it as data and they’re viewing it as code. And they’re viewing it as data in the sense of there’s a growing sense of, well, we need to make sure that our content sort of pairs well with our ecommerce system because we might need to supplement our ecommerce system or maybe our content models are related to our products in some way. And that sort of needs to be solidified, it needs to be piped between the systems. It’s kind of data in that way, and it’s structured.
Sam:
But the flip side is that it’s also kind of code. And so the content world is starting to think about, well, how do you merge in content branches to each other and how do we have more fully fledged review processes? Maybe in five years we’ll have some sort of Git-like pull request and review system is going to become more prevalent. I think it’s certainly possible. Right now, I don’t think the technology is there but I think that these kinds of workflows are likely to become more common over time, especially in larger organizations.
Larry:
The way you just said that, that’s different. You can roll back versions in Word or something like that. But what you’re saying is much more something like Git where there’s a system for managing versioning, but not just versioning, the other stuff you mentioned as well. Does that come out of… I know that some technical documentation folks talk about docs as code and document their work the same way they work with code. I guess it almost doesn’t matter there the origin of that, but the fact that you can think of content in the same way that you do code, that’s really interesting.
Sam:
Because at the end of the day content and code are both words that are multiplayer, collaborative aspects. Development has become much more collaborative over time. If you look at… Do you read Joel Spolsky’s blog posts at all?
Larry:
I do. I have off and on. But not lately.
Sam:
Did you read his blog posts of the 12 things to have… So he wrote a blog post around 2000 or so and he was like when you’re looking for a job as a developer, you should look for a job at a shop that has version control because it’s really rare, but it’s really important. You don’t want to just lose things. Fast forward 20 years or even if you fast forward 10, 15 or especially fast forward 20 years, gosh, it’s like version control, it’s like of course you’d use version control, why wouldn’t use version control? But that was something that we had to develop over the last couple of decades and it’s something that enables easier collaboration. We’ve seen some of those changes in the data science world. They came to the development world first coming to, they came to the data science world after that. I suspect they’re coming to the content world and we already see some of them in terms of some of the features that are coming out of various kinds of headless CMSs.
But I think it just takes time for these… Especially in a sort of post COVID world where it’s not as easy as just gathering around somebody’s desk and all sort of shouting out ideas together. You need a little bit more of a structured process in software to solve for it.
Larry:
You’re reminding me of this thing that’s fairly common the way the smallest ecommerce site is held to the same standards that Amazon is. The way that anybody working with words, you think, well heck, I can collaborate in my Google Docs why can’t I do that in my CMS? I know a few CMSs, I know Sanity has collaborative authoring, are there others that do that as well?
Sam:
We haven’t seen it as much. What was that? GatherContent, we were sort of familiar with. I think that’s a great example. I think the experience is going to move more towards is it that sort of collaborative multi-player authoring? Is it something more like Git? I don’t have an opinion on exactly which way it breaks, but I think it’ll break some way towards more multiplayer, more collaborative.
Larry:
And I think my hunch is that in a modular world the chance for advancing that kind of stuff is better than if you’re tied to somebody incorporating that as a feature of a monolith. If it’s something that can be developed independently, there’s better odds of it happening. Is that a reasonable hypothesis?
Sam:
I 100% agree about that. I think what you were just saying was one of the really underappreciated aspects of the cloud in general. So I think about something like Algolia, it’s really high quality search and it’s become a search primitive in many ways that just gets used across WordPress, across Drupal, across ecommerce, across really just a ton of different things. But well, how did that happen? Well, because you could amortize the cost of developing great search across the whole web as opposed to, oh, there’s only so many resources for a great search in Drupal, or pick your CMS. But if you can use it across the whole web then you can invest more resources in it. I think that’s powerful, right? That’s powerful.
Larry:
Exactly. I don’t know if it’s the exact right concept, but I’m thinking of best in breed or separation of concerns. Those two are related that if you just focus on search, you can do it really well, and part of doing it well is doing it in a way that’s modularize-able so that developers at WordPress and other places that can grab it.
Sam:
100% agree. And I think that the interesting, and in some ways tricky aspect from sort of an architectural perspective is that we’ve spent the last decade or so un-bundling the CMS and breaking it out into its constituent pieces in creating these best in breed solutions. But I suspect we’re going to spend the next decade re-bundling the CMS and getting all these integrations sort of really good and out of the box and even more intuitive to use over time than they currently are. Because you add complexity costs, there is some complexity costs and we need to reduce that complexity costs over time. That’s something I think a lot of folks are across the ecosystem are working on.
Larry:
So a little bit of a pendulum swinging back and forth there. Yeah, I think so. That makes sense.
Larry:
Hey, you mentioned in passing a couple of the benefits of modularizing. Things like, the need for speed arose from the Google search algorithms favoring sites that perform better. What else is pushing this? Are there factors that’ll push for even more modular solutions or things like search, like you mentioned Algolia that they really captured that part of the market. Are there other things do you see coming especially of interest to maybe content people in that family of optimizable things?
Sam:
I think… That’s a good question. Just generally one of the big things driving this over time has been ecommerce because more dollars over the internet just creates larger markets for a lot of this stuff. To your point about Google’s SEO change, and performance for ecommerce sites is always a huge driver of everything they do. So I don’t have a crystal ball what the next big change is going to be. COVID was a huge boost as well because, again, it was sort of boosted the shift towards ecommerce. I don’t know. I’ll be interested to see… What are your thoughts? Do you see anything over the horizon that you think is going to be…
Larry:
I don’t know. I see things that I hope will happen. We were talking before we went on the air a little bit about some of the buzzwords floating around like omnichannel strategy. A lot of people talk about that, but I don’t see a huge amount of it, actually real efforts to integrate content strategy efforts across an enterprise. There’s, of course, the branding people talk to the marketing people talk to the product people. So omnichannel I hope is a thing.
Larry:
Personalization is another thing that people talk about a lot and is sort of in its nascency but it doesn’t seem to really… Even the people who are doing it well like Amazon, it’s still laughable sometimes. Like, oh you bought a couch, you must be a couch enthusiast. So, I don’t know. Actually just those two examples ’cause I think those are two that people talk about a lot. Do you see opportunities for modular architecture’s increasing the odds of implementing omnichannel strategies or personalization?
Sam:
I like those examples. In some ways I think the promise is that the things that the big technology companies, the Amazons, the Googles of the world, take for granted which is having the resources to be able to do that with a quality that it provides a lift in whatever metrics they’re looking for. A lot of these modular tools are starting to provide that functionality for the rest of us. I think personalization is interesting in the sense of we have had personalization tools but… I love Deane Barker’s blog posts and I think he kind of captured the zeitgeist a few years ago. He was like, look, they’re not all their cracked up to be. They’re a lot more work than you thought and maybe they’re not as big of a reward as you thought. So I know it’s a buzzword, but maybe just think twice before you bet the whole hog on this… And I think he’s right, right? But there’s a chance that changes over time as the technology gets better.
Larry:
Hey Sam, I can’t believe it. We’re already coming up close to time. These always go way too quick and I could talk to you forever, and look forward to seeing you at Decoupled Days again next year. I hope we get to see each other again there. But, hey, is there anything last before we wrap up that just come up in the conversation or that you just want to make sure we talk about before we wrap up?
Sam:
I think one of the things that we’re excited about for the next few years is, one of the things that sets sort of Gatsby apart and that’s the thing that we spent a lot of time thinking about is the data layer and Gatsby has a data layer and it basically lets you integrate a number of different CMSs or data sources and treat them fairly similarly. And so for a while we’ve had folks saying, well, the data layer is one of my favorite parts of Gatsby as opposed to other frameworks. But, we actually just launched a product called Valhalla, which sort of lets people use this data integration layer with any kind of framework or mobile app or anything. So it’s still early days, but we see a lot of excitement around it.
Sam:
I think that comes back to your point about orchestration, which is that the interesting things that we’re likely to see over the next two, three years are not necessarily new point solutions, but they’re new integration and orchestration ways of pulling our existing pieces of the puzzle together.
Larry:
Nice. And that does feel like that right now. There’s been a lot of individual stuff done, really good search stuff, really high performance, ability to render pages and stuff. But now it’s like how come we really tease it together just right. I love that. And orchestration, I think that’s the right word for it.
Sam:
I think so too. It seems like a good word.
Larry:
Valhalla sounds… I’m really intrigued by that. I think we’re going to try to build my next site in Gatsby just so I can play with that and just build two or three… Because the whole point about Valhalla is that you can have multiple CMSs sourcing your one website or other experience, right?
Sam:
Multiple CMSs sourcing it and then pull that data into maybe, pulled into a Gatsby site but also pulled into a mobile app and do that. Be able to multi-channel.
Larry:
Now you just reminded me, I’m going to have to have you back on to talk about just the whole world of product content. Anyhow. One very last thing, Sam, what’s the best way for folks to stay in touch with you to follow you online and connect?
Sam:
Well, you can reach me on Twitter at @calcsam. That’s C-A-L-C like calculator and then Sam like my name. Or you can email sam@gatsbyjs.com as well.
Larry:
Great. Well thanks so much Sam. Really good talking with you.
Sam:
Thanks for having me on Larry, and it’s great chatting.
Leave a Reply