Impressive Hosting

In this episode of Impressive Hosting, Jesse Friedman speaks with Aaron Ware from Linchpin and Joshua Goode from WP Cloud at Automattic about WordPress hosting for higher education. They discuss approaching university websites, managing peak traffic during enrollment and commencement, and the debate between WordPress Multisite versus single-site implementations. The conversation covers performance optimization, caching solutions for logged-in users, and the "noisy neighbor" problem in shared environments. Throughout, they emphasize making hosting decisions based on an institution's specific needs rather than defaulting to legacy approaches.

00:00 Teaser   
00:54 Intro
23:57 Scaling for Success in Higher Education
28:37 Making the Right Decision in Higher Education with Multi-site
30:16 Navigating Caching Challenges in Higher Education
33:16 Multi-site and Higher Education
38:15 Single Site versus Multi-site in Higher Education
44:26 "The Noisy Neighbor:" Performance Scaling Challenges on Multi-Site 
51:28 Post-Game Analysis

Creators and Guests

Host
Jesse Friedman
Jesse Friedman is a long-time innovator, author, speaker, and currently, the Head of WP.Cloud and VP of Innovation and Partnership Success on Jetpack at Automattic. He is the author of Web Designers Guide to WordPress and a member of the New Riders, “Voices that Matter” series. A former professor, and long-term WordPress community contributor.
Guest
Aaron Ware
Guest
Joshua Goode

What is Impressive Hosting?

Impressive Hosting is a podcast that explores the core tenets of great WordPress hosting, from performance and security to scalability and user experience. Hosted by Jesse Friedman of WP Cloud, each episode features in-depth conversations with industry experts, developers, and hosting professionals who share insights, best practices, and real-world challenges. Whether you’re managing enterprise-level WordPress infrastructure, optimizing hosting for higher education, or scaling for high-traffic events, Impressive Hosting dives into the strategies and technologies that power the modern web.

Teaser
Jesse Friedman: If we have to access all these points, because they're all important, what can we do to make it easier for a student who's just trying to find something, or a parent looking for information about financial aid? What we try to do is look for solutions that are wins for the clients.
When students are coming into the school for the first time in the first semester, the first week, there's an influx of traffic.
Jesse Friedman: There's people signing up for classes. There's a variety of things happening all at once.
How do you start to think about that? How do you start to think about scale for an institution like that?
Joshua Goode: When you have one site that can take advantage of bursting up to 110 workers, or if you decide to do a multisite and pack it with 100, 200, 500 sites, on our end, it's still treated as a single site. So it's going to have that same burst limit of 110 workers. How can we solve for some of the things they want to do?
Intro
Jesse Friedman: Welcome to Impressive Hosting, where we seek to uncover the core tenets of great WordPress hosting. I am your host, Jesse Friedman. And with me today is Aaron Ware from Linchpin and Josh Goode from WP Cloud at Automattic. Welcome guys.
Aaron Ware: Yeah, thanks for having us.
Jesse Friedman: Thank you, Aaron. Why don't we start with you and do a brief introduction. Tell us a little about yourself, where you live, what you're interested in, what you're working on.
Aaron Ware: Sure. My name is Aaron Ware. I own a smaller digital agency called Linchpin. We're 100 percent remote, but the company is based out of Little Rhode Island, where I live and actually Jesse, you live as well. Representing Little Rhodey today. We're pretty boutique and diverse in the clients we work with. Everything from Fortune 100s to smaller mom and pop stores, brick and mortar stores to startups. And even with this conversation, higher education at all different sizes of universities and institutions.
Jesse Friedman: Nice. Thanks. Joshua, why don't you jump in?
Joshua Goode: Yeah, I serve as the technical solutions engineer for WP Cloud. That's a bundle of partner success, technical account management, sales engineering, and integration engineering with our partners. Across the board, I help people integrate and get the most out of the WP Cloud platform. I'm from Tennessee, the southeast corner of Tennessee - any further south you're in Georgia, any further east you're in North Carolina. It's a beautiful area. I'm very grateful that I get to do all this from here. The big Automattic dreams right here from rural Tennessee.
Jesse Friedman: But you don't spend all your time online writing code and doing technical things, right? You also manage a farm?
Joshua Goode: We do a little bit of farm work out here. It is a rural area, so there's a lot of opportunity for that. We've pulled back on some of that operation. But yeah, farming, and a bit of craft beer brewing. That's actually one of my past careers. In between when a tech company I worked for sold out, I had an opportunity to join a brewery. I still do brewing on the side, just a little bit. Non-professionally. So I get away from the computer when I can, but the draw of the things we do with Automattic keeps me at that computer a lot.
Jesse Friedman: Yeah. Very cool. So, a lot of people don't know this, but Aaron and I go way back. We met in 2005, so that's almost 20 years ago now. I joined BZ Results in Rhode Island. It was my first real job after college. And at that time, Aaron, what was your role?
Aaron Ware: I think my official title was Head of R&D, but primarily that was Head of Research and Development for Front End. In that ancient world, our lives revolved around Adobe Flash and Macromedia Flash before that. And wow, that's 20 years. That's crazy.
Jesse Friedman: It's a long time. You know what was interesting about that company is that it was run very much like a factory. In the sense that we were pumping out code, designing sites for auto dealerships. But it was the only time I've ever worked at a tech company where you promptly started at your desk at 8am, had a one-hour break, and then were done at 5.
And you just pushed all day long. There wasn't a whole lot of opportunity to relax and hang out with colleagues. There was no coffee room or coffee break area, but we did have Call of Duty. You remember?
Joshua Goode: Nice.
Jesse Friedman: We used to play hours of Call of Duty during our lunch breaks and after our days were over.
Aaron Ware: Yeah, that company had a lot of rigid rules in some aspects and then was very lenient in others. I learned a lot. 2005 is when you started, and 2006 is when I started Linchpin. That was when I was just about to transition out, and I was starting to lay the groundwork of resigning. It's really funny because earlier today, I told the team that I was going to be doing this podcast, and they started sharing "Oh, I met you when I listened to this podcast." And they sent me a podcast that I did for a local Rhode Island podcast. And then another team member was like, "I met you when you spoke at this WordCamp." And then another team member was like, "I met you when you spoke at the keynote of this other WordCamp." And I realized that a lot of the team, though I don't do this often, has found me through these different avenues. I just find that really interesting that I said today I'm doing a podcast.
Jesse Friedman: I was just going to say, to segue into - are you hiring right now?
Do Clients Care About Distributed Teams?
Aaron Ware: Yeah, we're pretty nimble, we're around eight to ten people total as a company. If there's an individual that can join the team and really flourish - sometimes we have to think about it from a budget standpoint and what we can afford with healthcare and salaries and those benefits. We're always looking, especially at this time. When we talk about hobbies and things I'm into, the block interactivity API is a huge part of what we've been doing. So if you're interested in that and you're looking to join a smaller agency that does a lot of WordPress work, I wouldn't mind someone applying for sure.
Jesse Friedman: Let's get into the meat of the podcast here and talk about some of the work you're working on, Aaron. A good segue might be: when you are approaching a new client, whether it's higher education or not, do they care at all if you're distributed? Does it matter to them? Does the conversation even come up?
Aaron Ware: Honestly, most of our clients only worry about us being distributed when they ask, "Hey, who's gonna answer the phone if something's wrong? Who's gonna answer an email?" With the team being distributed, we have a team member on mountain time, and we've had team members in Italy, Hawaii, or California.
Jesse Friedman: You can make it an advantage rather than a disadvantage.
Aaron Ware: Yeah, sometimes there's a stigma that distributed means farmed out or offshored. So some of that is educating a prospect or client that the team they're paying is doing the work. It's not a vendor - we're not going out to someone else to do that work. They understand because they actually meet the team. It's not just a PM; they speak to the developers or designers or myself, and we're pretty personable. So that helps. It also helps that we're pretty small.
It's one of those things where a lot of faces touch a project, and usually pretty much everybody touches a project at some point. If you're a larger agency, that's just more value for you. If your team's in the UK, we have people in the UK. If you're in Australia, you're in Singapore, it doesn't matter. And that's a positive thing.
Getting Clients in Higher Education
Jesse Friedman: Let's talk about higher education. If you had a university coming to you today to send you an RFP or something like that, what would you do to start scoping out the size of that project? How would you get a better understanding of what it is that they actually think they need?
Aaron Ware: With higher education, typically they are more prepared than, say, just a smaller business with maybe a marketing manager. With higher education, especially if they have state funding at any level, they're going to have an RFP and it will go out to bid. Now they might have a vendor preference, they may really like the agency that's working with them today, but it still will have to go out to bid because of certain rules and regulations.
For example, in the state of Connecticut and Rhode Island and Massachusetts, there are portals that every agency or freelancer can sign up for, and in that they'll have an RFP. They'll have their set things that they really want in your proposal or your statement of work. Sometimes that's a non-starter because maybe that organization already has a huge investment in Drupal. And that's okay. For those we say, alright, it's not for us, but maybe there's an agency that does that.
And it's a small community around here of agencies and smaller shops. So usually we'll extend an olive branch and be like, "Hey, have you seen this?"
Jesse Friedman: Do higher education institutions typically favor open source, not necessarily just WordPress, but open technologies like Drupal, WordPress, things like that? Or do you see them also using closed networks as well?
Aaron Ware: Yes, I would say that more often than not they're favoring an open source solution. Mainly that's because the talent acquisition for them is going to be easier than if you use something like Craft or tried to go towards another hosted solution. And we can speak to some of those decisions that at least we've seen from our experiences. But other things we're going to look at for higher education: What is their current infrastructure look like? Do they actually have IT staff that are familiar with web?
Because sometimes they don't. Sometimes it's, hey, we make sure email's working, we make sure that the VPN is working, we make sure that the WiFi access points are all rocking and rolling. However, when it gets to the web, sometimes IT may be 100 percent hands off outside of DNS entries.
Jesse Friedman: But are they still making the decision about who to host with?
Aaron Ware: Sometimes. Sometimes they do. Sometimes it's up to marketing. And we've found that sometimes it's trying to help higher ed understand governance. So governance is a big thing in higher education, and it's really for those who are listening that aren't really familiar with that: it's trying to make sure that all stakeholders at whatever level, whether it's people in a specific department, whether it's marketing or IT/engineering, whether it's administration, whether it's board members - every one of those roles is really important and every one of those roles should be respected and heard.
And really, it's also about who's responsible - one for the decision making, but two for the implementation. And what we have found is sometimes it's really hard for a university to say "who should own this thing and who should be responsible for it?" Because you think about it, traditionally most websites are marketing pieces, right? Especially they could be as simple as brochure ware for a small business that's just trying to get leads. But ultimately sometimes it's a bit of a challenge for a university, especially smaller ones, to understand, "Hey, who are the stakeholders?" Because there could be a lot of them, and "who is responsible again for the hosting decisions for even ancillary services?" Because there's a lot of services that are needed in higher education. And some of them are 100 percent out of marketing's control or IT's control.
Jesse Friedman: That's a great point. It's something that we have to do anytime that we're working in business development or partnerships as well. You want to identify the folks who are going to be making those decisions and have the power to do that. How about the actual project itself? Do hosting companies come with their own ideas on how to solve for things? Or do they just speak to the solutions that they need, or rather the end result, and let you decide?
Do Hosting Companies Come With Their Own Ideas?
Aaron Ware: One of the things that we sometimes come across, even if it's a new bid, is legacy debt. Sometimes we don't actually get a choice initially. Say we're getting our foot in the door, and this has happened historically. Let's say another agency has built a solution, and that solution isn't necessarily what the client wanted, and we're there to help. And we haven't even bid on this - they've come to us for this. Sometimes we have a challenge where we're inheriting where the site is hosted and maybe that hosting solution isn't performing well. But is the hosting the issue? Is how it was built being the issue? So sometimes there's that technical debt and that legacy debt. I would say if it's a brand new bid, the first thing we try to look at when we're determining how we host is, with higher education, there's sometimes peaks and valleys.
Enrollment is a very important thing. So when that's happening, there's a few things we need to take into account. When we think about the catalog of courses and programs that most educational facilities have, we want to look at hey, when do enrollments happen? Is it annually? Are there trimesters? What's the matriculation look like? And also, what does commencement look like? What do events look like? Because those can really impact that traffic. And sometimes, if it's hosting and maybe, again, IT is involved and they're paying for the hosting because it's budget, they may say, "oh my God, enrollment is like this. So we need this crazy tech stack because enrollment is going to be crazy." But the rest of the year, they're not scaling down for the valleys, right? So what we try to do is look at that from historical knowledge, from analytics. And that would be in the question.
Those are things that we would look at as much as they can share, of course, because sometimes they're not going to share all the intimate details like traffic and those types of things. But that's definitely something we're going to ask and that way we can make the right decision when we're recommending a host because ultimately most of the time we don't select the host for them, we're giving them the best information in order to make an informed decision.
Joshua Goode: Because ultimately we don't pay the bill, they do, so it's not really our choice, it's their choice. If they're coming to you with an issue and something that's already been created and it comes down to is it due to how it's been built or due to the hosting? How common is it that the main solution there is to get a better host?
Aaron Ware: I would say less often than it's just the host in this day and age.
Joshua Goode: Got it.
Aaron Ware: Most higher eds are making at least a slightly informed decision. There are big names in the space now, comparatively to, I would say, underperforming shared hosting. I think that happens a lot less now than it did, say, five years ago. And that's just a more educated - I'll say consumer in air quotes - there's more people in the space and with WordPress being so ubiquitous, makes it a little bit easier to host, especially when you can even see commercials for hosts, like when you're watching YouTube.
Joshua Goode: Yeah.
Aaron Ware: Which I get inundated with, based on how I browse the web. It happens a lot though in smaller business. When we're working with small and medium businesses, hosting is almost always going to be a solution when it comes to performance. Because it's usually a quick turnaround versus if we're investigating a hodgepodge of plugins, and not necessarily performing in terms of queries or other technical kind of engineering details that I won't get into.
Joshua Goode: Right.
Jesse Friedman: So one of the things that I think about when I think about higher education is that it's not as simple as just building a one-time website. Even if you have a complex e-commerce platform, or you're trying to do something crazy with like auction sites or whatever, you're still building a website. But with higher education, you could have the marketing website, you could have a different website for admissions, you might have a network of sites where each student can sign up and build a website. You can get into internal intranet type sites for classes and colleges with inside the university. So when you typically see a new higher education opportunity, are you thinking about it from that angle? Trying to take on everything at once? Are you trying to come in and solve for one problem at a time?
Aaron Ware: I'm pretty pragmatic about the things that we can accomplish, especially when it comes to budgets. I would love, just like any business owner, I would love everything. Give us every problem you could possibly have and we can come up with a solution for it, but if a smaller university is coming to us with, say, a $65,000 budget for their main marketing website for their edu, it's going to be pretty much impossible for us to say "Let's take on admissions, right?" Because they probably do have a solution for that. Now what we try to do though, is we still like to plan out: how can the website take advantage of these other platforms? Is there an API available so we can aggregate some of that, maybe again, I talk about the catalog, right? Can we aggregate the actual master's programs? Can we get the calendar of X, Y, Z thing, right?
So we try to make sure that the website itself, the marketing site, is the main hub for that, at least the communication and the marketing aspect of it. Now, if that means they still need to go off to an admissions website, those things are going to happen. That's pretty common in the space. However what we do like to do, though, is say okay, if we have to go to all these points, because they're all important, what can we do to alleviate that for a student that's just trying to find something, or a parent that is looking for information about financial aid? What we try to do is say how can we make that easier? So things like single sign-on. That's a really important thing. And that's a pretty universal thing that most higher ed will want to try to do for governance, right? We want to keep track of who's logging in, where they're logging in, what they have control over. And that's something that, you know, is one of those things that in a $65k budget, which is just a random number that's relatively low from that standpoint. That's something we try to look for - we try and look for those solutions that are wins for the clients.
Jesse Friedman: Yeah, that's a great way to go about it. In our last episode, we had Jim Groom on from Reclaim Hosting. He also focuses on higher education. You have a diverse portfolio, but one of the things that we learned about is that right now, for example, it's November.
How Do You Scale for Success in Higher Education?
Jesse Friedman: So a lot of managers of websites or hosting companies are thinking about Black Friday and the influx of traffic that e-commerce websites are going to get. In that episode, we talked about when students are coming into school for the first time at the first semester, that first week, there's an influx of traffic. There's people signing up for classes, there's a variety of things happening all at once. And you just talked about the mountains and valleys and how higher education institutions should be thinking about that. So when you think about scale within their own user base, especially since a lot of these people are going to be logging in, not maybe on the edge or cached as much as a typical user on the outside of a website might be - how do you start to think about that? How do you start to think about scale for an institution like that?
Aaron Ware: The first thing that we look at, even before it gets down to hardware and how it's hosted, is we look at the actual application level itself. That's really our starting point. How cacheable the site is - that's something that's really important. When we look at WooCommerce, for example, when we look at being logged into the WordPress admin or being a user that has a session, right? Those are all really big challenges.
So when we think about having uncached hits - let's say there's no Cloudflare, there's no Web Application Firewall, there's no Object Cache, no Redis, right? We try to look at what is making the site not perform well. The easy ones are there's a ton of plugins that are doing a lot of things, and maybe those don't really need to be there actively. There's a plugin I love, Query Monitor, but not every site needs to have that on in the admin every time because it's doing a lot of stuff and you're getting a lot of useful information, but maybe it's not needed.
Joshua Goode: Mm-hmm.
Aaron Ware: Maybe from a governance standpoint, do we really need 30 admins, or are we really looking at two admins and the rest are editors, right? So what we try to look at is not only from a plugin standpoint but what is the site's functionality? And are we actually doing an effective job with static assets, expensive queries, and how we can utilize cache for that? Whether it's WP cache or transients, those are all the things that are going to be a baseline of any project that we inherit.
Typically, if it's a project where we didn't bid on it and someone is reaching out to us, 99 times out of 100, the performance of the site is bad. Now, most of the time it's not a hosting issue anymore for higher ed anyway. It usually is a thought process in how the site was developed. And we kind of work from there through that technical debt, that legacy debt.
We had a larger nonprofit in Massachusetts come to us because they had a smaller two-person shop that took on a project that was too big for them. It was originally a multisite where all the sites basically used the same parent theme, but they had different offshoot child themes. But ultimately, every design was really different - dramatically different. They were kind of inheriting stuff from a parent theme, all looked different anyway, so that functionality would probably better suit it in a plugin that could be toggled on or off depending on features.
And the site was going down constantly for them. The site was hosted on a single VPS that wasn't configured because it's a small two-person team - they have one designer, one developer. The developer is not a sysadmin, they're not in DevOps, they don't really understand how to configure Redis, let alone all the other things.
Jesse Friedman: Sure, and that falls on them anyway sometimes, right? Because they think just because you can build a website and fix my printer, it means you can also run the server.
How Do You Make the Right Decision in Higher Education with Multi-site?
Aaron Ware: Exactly. And they think that they're making the right decision in terms of going to a multisite, right? They're thinking, "Okay, we can manage all the updates and we can do all this from one WP admin network admin area." But come to find out, there's a lot of performance implications with multisite. There's a lot of configuration that you need to do to make it performant, let alone just understanding - hey, I have 120 plugins installed because I have five or six different sites on it and they all have different needs. Maybe two of them are WooCommerce, the other three or four aren't WooCommerce. We still have all this technical debt and these challenges of updating.
That one client moving from just having a VPS on a host that, quite frankly, I had never heard of until obtaining the client - switching them to one of the bigger name hosts immediately made a difference because it was configured. Honestly, just a basic configuration of any sort of scalable managed hosting is going to make an immediate impact on the performance of the site. And that's just going from a non-configured LAMP stack. That makes a huge difference.
Jesse Friedman: Sure, yeah.
Joshua Goode: Yeah, I...
Jesse Friedman: You're touching on something I'm super passionate about, Josh. I'm gonna let you ask that.
Joshua Goode: Yeah.
Jesse Friedman: Want to stay on this point around multisite.
Joshua Goode: Yeah.
Navigating Caching Challenges in Higher Education
Joshua Goode: What I was going to bring up is I really want to dig into the multisite a bit here. Before doing that - we don't really have to dig into the weeds too much or dig too much into the technical details - but you brought up a good point about cacheability. That's something that we see a lot when I'm working with hosts on WP Cloud. Navigating the caching challenges - we include a lot on WP Cloud by default, it's all built in. They still deal with large stores and even some education websites where caching is still a factor because they have unique situations. As you've mentioned, you've got all these students logging in, you have all these sessions going on.
So I'm just curious - again, we don't have to dig too deep - but how do you approach a situation where you've got all of these students logging into the site, a lot of uncacheable requests? What kind of things are you doing or navigating to bring caching where possible, or to just make the performance a little bit better, given that they're logged in?
Aaron Ware: Yeah, one of the things that we found when working with non-cache results were not so much the initial page hits and things like that - it's ancillary things that are happening behind the scenes, whether it be notifications, whether it be other statuses that are happening. And this happens with WooCommerce as well. Hosts have done a lot to work within the actual live cart and those types of things...
Joshua Goode: Yeah.
Aaron Ware: ...and the checkout process.
So we do rely on what is actually static, what is actually happening when that session happens, because there still are a lot of areas that, while they're not fully static, there are a ton of aspects of sites. There are larger universities that deal with that a little bit more than we do because our kind of bread and butter is smaller university and smaller higher ed. So we haven't necessarily needed to deal with that as much as maybe the other gentleman that Jesse mentioned...
Jesse Friedman: Jim Groom, yeah.
Aaron Ware: Yeah, Jim.
Jesse Friedman: Yeah.
Aaron Ware: Jim's in it and he's heavily into it. Most of our clientele, the biggest areas like admissions or whatnot, we've been more involved in the integration of it versus having to deal with the brunt of that. Where we find the most traffic that we have to deal with - yes, it's during enrollment, but a lot of it's commencement, especially once COVID happened...
Joshua Goode: Hmm.
Aaron Ware: ...and so many people weren't going to the ceremonies. They were trying to catch them remote. And we found that even now traffic is usually pretty high on commencement. And what we try and do is look at how can we make these landing pages that people are visiting - how can we make those highly cacheable? How can we eliminate someone even having a session on those types of things?
Joshua Goode: Yeah.
Aaron Ware: I think that's where we get a little bit more creative in solving that.
Joshua Goode: Great.
Multi-site and Higher Education
Jesse Friedman: Yeah, looking at it from that perspective. When we think about multisite, there's a lot that you get for free with it. We talked about this in the last episode as well, but we just touched on it. I love the idea of talking about multisite in 2024 because I started working on it 15 years ago. At the time, you got so much for free - you got plugin management, theme management, the ability to update everything with one hit, user management, things in terms of security because you could put up guardrails and prevent people from doing things. But now, 15 years later, we're looking at it from a different perspective with WP Cloud, for example.
The way in which we can roll out updates across an entire account allows for us to do that with symlink files, and then everything is essentially getting the same power that you'd get from multisite but you're doing it with single sites. We still run into some issues, right? Because multisite user management makes it really easy to assign individuals a single username and password into a variety of websites. But you can still solve for that today - there are SSO solutions, you mentioned single sign-on. There's other ways to compensate for that.
When you think about multisite and a large network, are you thinking about it from the perspective of "have that as an option, we're going to go with multisite," or maybe we don't and use something else? Or is it really just when you run into customers who are asking for these types of things, multisite is pretty much the default still?
Aaron Ware: I would say that over the last three years or so, I haven't had a single client come to me and say we need a multisite.
Joshua Goode: Wow.
Aaron Ware: Most of our clients, even if we said, "Oh, we're going to auto update everything. We're going to use machine learning, we're going to run the updates, do visual regression testing. If it fails, we'll roll it back" - even if we say we're going to do all those things, most of our clients, because we're a concierge service, we are the ones doing that. We are the ones rolling out those updates.
When it comes to clients of ours that we would propose multisite, the first thing I look at is what is the purpose of that? I can speak to a client that we had previously that had an over 1,000-site multisite. We didn't build it out - it was an inherited project that was having performance issues because, well, it was a 1,000-site multisite instance. And that challenge for them was they had officers that they had basically microsites of the main site.
And they wanted to give them the ability to have a few pages that they could customize, and they would have that sole role to do that. So for them it was governance, right? Hey, we have these loan officers that can only control their site. The template - everybody got the same exact template. So the branding of it was exactly the same. There were no child themes. And that's how they handled their governance.
So everybody got a login and they could, as an admin, manage everybody's stuff, managing the updates for plugins. If you need to do that and you need that level, I think that's a great solution. However, I think in this day and age, I would also think about it from a region standpoint, and how it's being cached and how it's being served up at the edge. I'd also be thinking - does this really need to be a multisite? Do we need individual domains for users? From the SEO standpoint, is there a lot of redundant content across these? Should this better be suited as a directory?
I'm just using this project as an example of how I would rethink it nowadays. More so with being able to roll out updates and being able to actually have a better deployment process for that. It makes it less and less of a viable solution for most clients, but there's always that scenario where it would break my heart if ever it was not a thing just because so many people worked on it and it's still a viable thing. But for us, we're not really proposing that too often anymore.
Single Site versus Multi-site in Higher Education
Jesse Friedman: Yeah, makes sense. I love Multisite. It was a very important tool and something that I used many times early in my career to help solve for a lot of problems. But I think the way I look at it now is that simple things like a must-use plugin, for example, can be leveraged to do something like - with higher education sites, you'll see that we want to put a banner up. You mentioned commencement, right? Everyone's graduating next week, you got to be out of the dorms. It's the type of message you would see at the top of every single website in a multisite network.
But a must-use plugin that's used across an entire network of single sites can do the same kind of thing. There's SSO tools to help with the management of users across many different sites, so you get the same tools. SSH and CLI can be used for Super admin powers, even though it's not like an official super admin within a network.
It's really important that we think about the power that you get from multisite and what you can replace that with and what you get for free and what other advantages you can take from single sites. Right out of the gate, single sites give you a performance upgrade. We talked a little bit about the complications with networks sharing a database, having concurrent users across an entire all at the same time, pulling the same data. Single sites get you a lot of stuff for free. You can allocate different resources. A student website, for example, may only need a single worker whereas the marketing website for the admissions team might need to be able to compensate for 50,000 visitors a month, something like that.
Aaron Ware: Yeah.
Joshua Goode: Yeah.
Aaron Ware: It really does come down to how much things can be processed at the same time and how many workers are available. And to your point, with a student site - we have one of our clients that has a dedicated site for their map, an interactive map because once enrollment happens and students, they may have gotten a tour, but a lot of times they don't really know where to go. But that's not really necessarily a really busy site per se. But it does need to be performant. Most users are going to be on their phone, looking around, and they're going to be trying to find, hey, I need to get to this office, I need to get to this class.
Jesse Friedman: There's no worse time for a site to load slowly when you're lost, right?
Aaron Ware: Yeah exactly. That's a thing we talk about - scalable hosting, that's something where that performance matters. Maybe they're lost and they're trying to find the nurse, trying to get some help.
Jesse Friedman: Yeah.
Aaron Ware: Those things do matter for sure.
Joshua Goode: Yeah, performance and scalability. As I work with partners and hosts who are really trying to build out specific solutions for their end customers for WP Cloud, it really comes down for me - our main discussion on whether or not multisite's a good solution. Even though we do support it, it comes down to scalability and performance.
There's so much built into WP Cloud in terms of the ability of scaling up your PHP workers as needed. We offer bursting that goes up to 110 workers and actually here soon, it looks like we're going to go well beyond that in our next generation of hardware. Navigating...
Jesse Friedman: We're getting... getting we're eh, a little...
Joshua Goode: A little, a drop in a little something, yeah.
Jesse Friedman: I...
Joshua Goode: Now when you have one site that can take advantage of that bursting up to 110, or if you decide you want to do a multisite and pack it with 100, 200, 500 sites, on our end, it's still treated as a single site. So it's going to have that same burst limit of 110 workers. So it's really trying to navigate, okay, how can we solve for some of the things they want to do?
And when it comes down to, "Oh we want all of these themes to be available on all of these, this segment of sites. We want these plugins to be available by default on these segment of sites" - we've worked really hard on WP Cloud to build in some of those features just for general day-to-day management, but they can be definitely used by multisite operations. Saying hey, you can have these on there by default. They actually don't take up any additional storage the way that we manage and symlink them to a file system - they at least don't take up storage on these individual sites, very similar to a multisite situation, and then how you manage updates and how you want to do that.
It's very controlled, and to the point to where those plugins also can't be edited. So if you've got someone who shouldn't have admin maybe somewhere and they just say, "Hey, I want to slip this into this plugin," we can actually work with you to lock that down. So I think we do have a lot of people with strong opinions come in. They love multisite - I don't blame them, but it's more of, okay, let's really navigate this to make sure it is a good fit for your situation. And let's also weigh these other benefits that we have with WP Cloud.
Jesse Friedman: You touched on something there that I keep going back to over and over again - but you have a thousand sites and one of them is absolutely crucial to your business and the core competency of what it is that you're offering, and then a bunch of them are ancillary things that if they do have a little bit of downtime here and there, it's not the end of the world. But the entire network is sharing the same amount of resources. And so if you get one of those sites slammed, absolutely slammed, and they're taking away those resources, every site on that network is suffering from that. And I believe that's still a true statement. Correct me if I'm wrong there.
Joshua Goode: Yeah, in the typical setup, yeah.
"The Noisy Neighbor:" Performance Scaling Challenges on Multi-Site
Jesse Friedman: Having these multisites broken up into different single sites allows for you to have granular control over the resource allocation for those things. It's something really important to think about - if you're going for multisite to be able to compensate for the management of thousands of websites and users and people doing a variety of different things, and you're thinking you're getting a lot for free with Multisite, which technically you do. You're still also setting yourself up to have to compensate for a lot of performance management and compensation for that.
Aaron Ware: Yeah, we call it the noisy neighbor, right?
Jesse Friedman: Yeah.
Aaron Ware: They're all connected. They're all in this neighborhood, right? And we have to deal with that. Sometimes even with hosting in general. So there's so many providers out there that a smaller agency, they can pick a package that can host sites - 25 sites, 100 sites, or whatever happens to be. But a lot of times those, again, are just on a specific docker container with whatever host it is. And then they're getting all of their individual sites. But they're still technically, most of the time, not on a clustered environment, right?
Joshua Goode: Yeah.
Aaron Ware: Most of the time it's just on, again, a VPS somewhere that wherever it's really hosted out of like Amazon or Google or Digital Ocean or whatever. And they're trying to sell as many sites onto that account to make that hosting revenue. And whether it's multisite and there's that one big multisite that's on there and it's consuming all of those resources, or it's an e-commerce site that maybe again is expensive, you have to deal with all those performance things.
With solutions like Pressable and WP Cloud, they're able to offset the idea of the "noisy neighbor."
Joshua Goode: Yeah, yeah.
Aaron Ware: And that's really intriguing to agencies like Linchpin. I always am thinking about that. Uptime's important, response time's really important. And again, application level impacts that. It's not to say that a site can't be incorrectly developed. But it's something we really think about especially that noisy neighbor because it does come up.
We still to this day have a multisite that we basically transferred to every client of ours that was on this multisite for ease of use - we have transferred off into their own individual sites. And we called it the Hive - the Linchpin Hive. We had all these cells in the Hive and we just treated every site like that.
We had these plugins we want every client to have. The tried and true. We're gonna have the Yoast of the world on there. We're gonna have our Gravity Forms, we're gonna have Jetpack. We're gonna have all these plugins we trust. That's the suite that we're gonna give you.
And at first we found it to be easy. But then all of a sudden, again, you have a site that maybe has a lot of plugins it's using, but not everybody else is taking advantage of. They gotta go on their own, right? They should be on their own hosting solution. They should be on their own site, so that's an argument to maybe diversify and move them off. And so we've done that. We've moved everybody off of that. Except for one. There's literally one client that is in this multisite still to this day that we've reached out to and just - it's not necessarily a big deal. That database doesn't have all that overhead anymore, but technically speaking, that site is still technically a multisite. Because I forgot what ID they are in there, but we mapped the database to the WP whatever site.
Joshua Goode: Yeah.
Aaron Ware: At the time, right? I think it's just a transition to the world as it stands today and how scalability is just different. I find that to be really interesting that you mentioned it because actually, we just reached out to that client three weeks ago, just to hit them up because they're self-service. They just do their own thing. And it's a very small site.
Jesse Friedman: Yeah.
Joshua Goode: Nice.
Jesse Friedman: Maybe there's an opportunity for them with Pressable or Bluehost Cloud or another WP Cloud provider. But that'll be my only plug of the day. With that - this was a great episode, Aaron. I would love to have you back on in the future. Josh, you'll definitely be back on occasionally here and there.
Joshua Goode: Great, yeah.
Jesse Friedman: Let me ask you this Aaron - if folks want to reach out to you, if they want to apply for a job so they can say that they work for you and met you on this podcast, where do folks go? What URLs do you wanna share?
Aaron Ware: You can always go to linchpin.com for sure. Linchpin's with I's, no Y's in there. You can also hit me up on, man, I'm not really a fan of any social media, honestly. But you can hit me up on Twitter slash X at Aaron Ware. It's pretty easy to find me. If you just search the web and search for Aaron Ware, you can find me.
If you want to look at older podcasts, you can find me on a few podcasts on Spotify or wherever you get your podcasts, on wordpress.tv. You can see some of my sessions if you want to do that.
Jesse Friedman: You're like a pro, Aaron. Look at that.
Aaron Ware: In terms of celebrities, I'm like a Z-list. I'm like, really? I think that's as high as I'll go.
Jesse Friedman: You still got it down, though. Podcasts where you podcast and all that. That's great. Josh, how about you? Is there anything you want to share?
Joshua Goode: Yeah. First and foremost, if you're a host, you're interested in what we're doing with WP Cloud just jump over to wp.cloud. We've got some good information there and some ways to contact us and to start a discussion and we've got some good case studies. I think we've got one for Pressable up right now, working on some others.
Learn more about that. Or if you just want to reach out to me about something else, learn more about the farming or brewing and stuff we talked about, you can find me at goode.pro. That's good with the E dot pro.
Jesse Friedman: Nice. That's great. And as always, if anybody has any follow-up questions for the crew here, you can go to impressive.host and ask your questions there and we'll make sure to follow up. Aaron and Josh, thanks so much for your time.
Joshua Goode: Thank you.
Aaron Ware: Thanks for having me. I had a blast.
Jesse Friedman: Us too.
Post-Game Analysis
Jesse Friedman: Alright, that was another episode about higher education with Aaron Ware and Joshua Goode. It was great having them on. They taught us a lot. So let's recap. What did we learn today? We learned a little bit about the RFP process and going out and getting clients around higher education.
Some of the things that we also covered were around governance and the importance that higher education institutions place on the micromanagement of their websites and their network. That's super important when you're going after higher education clients. Giving them the ability to manage user roles and plugins and themes across their network is super, super important.
We also touched a lot on performance and scale and the debate about whether or not multisite is absolutely right for the job or not. There's a lot of value that comes with multisite. But it's also worth exploring what WP Cloud or other solutions out there can do to help you unpack a multisite. There's a lot you get for free from creating singular sites.
We talked about the granularity of resource allocation and how you can make sure that your noisy neighbors are not causing problems for the sites around that network.
Definitely take some time to consider your next multisite project and think about whether or not another solution might be good for you. Thanks, everybody, and looking forward to the next episode.