Hey y’all. You know how I’m always posting threads asking for conceptual advice? Well here we go again:
I have a few legacy sites that I built a few years ago as WordPress Multisite installations long before Trellis and even longer before I finally realized the errors of my ways and started USING Trellis. Basically these two clients have multiple product lines, and multiple domains, so I set up Multisite to address that need.
Each of these Multisite installations have 3 or 4 sites; they’re not huge networks. At the time, since I didn’t have good deployment or version control in place, it seemed easier to maintain them together as a single install, but I’m rethinking it now. Because it’s such a small number of sites, it might be easier to just run each of the sites on a separate droplet.
So here’s that conceptual question: when do you use Multisite? Do you have a number-of-sites-in-the-network threshold after which you use Multisite? Do you always use it? Never use it?
I think I’m pretty well settled on splitting these sites up, but I’m curious about your practices.
I personally only use it for clients that have multiple sites running off the same theme. Helps keep theme updates organised and billing hosting much simpler.
In my experience multi-site is too much of a bug bear to go putting sites together unless absolutely necessary.
Same as above. We’ve got one in the works now that has around 30 sites that all, except for one, are running the same theme. It’s the most compelling case for multi-site I’ve encountered yet.
I avoid it as well, it just seems sketchy to me. I don’t know why, considering wordpress.com is one gigantic multi-site network, or so I’ve been told.
I’ve seen some rather disastrous client multisite installs. I’ve never personally launched a company multisite for multiple products like has been described here. I personally don’t see much value in Multisite other than it’s original purpose which is basically like wordpress.com .
I suppose I could see value in multiple sites for a company, sort of like regional sites or franchises, as long as they all share the same plugins and theme(s). They would all need to be very similar though for me to see any value. I’ve seen company sites where they’ve had 5 or 10 company sites but each one had a separate theme, different plugins… pretty messy to be honest. Sounds like a good idea, until you get into the idea of actual deployments. Then you have to worry about 10 sites rather than 1.
I really think the best use case for Multisite is a service like wordpress.com , even if it’s not a “blog” per se, specialized websites might also be a good use case. Of course, then you also have to be careful the service doesn’t stray too far away from WP conventions. Once you get into app territory, I would suggest checking out an actual app framework like Laravel and staying far away from WordPress. That’s my personal opinion.
Even in the case where sites use the same theme or plugins, Bedrock’s composer dependency management can handle that.
Thanks for the insight. I’m definitely going to split these sites up as time allows.
Right, there is that. So that’s probably where you need to weigh the pro’s and con’s of both. Is it 2 or 3 sites? Then maybe you won’t mind deploying 2 or 3 times for a plugin or theme update. Is there 20, 30, 100 franchises that will all use the same theme? Suddenly any type of update to theme, plugins, WP, etc. starts to sound pretty good using Multisite.
That’s a rare case imo though.
If your main concern is managing themes and plugins across all your sites then I recommend using MainWP. I have been using this on all our sites (have it connected to over 100 between staging and live environments) and it never skips a beat for me. I paid the money for the lifetime subscription to get all their extra packages and I have been impressed with it. It’s also 100% open source so you can browse all the code for the dashboard and the child site plugin. Which is not the case for WP managers like InfiniteWP, which encrypts some of their code so who knows what they have going on there on your sites / dashboard.
I’ve considered using multisite for a couple of clients, but inevitably the costs and hassle seem to outweigh the benefits.
In my mind, there are a couple of reasons to go with multisite:
- The same ‘user’ is active on multiple domains and you only want one place to manage that user info. One example of this is some sort of online education site where each ‘course’ is it’s own domain, but the users (students) are the same across the various domains.
- Coordinate plugin / theme updates - for the reasons stated in other comments.
This is probably a separate conversation entirely, but do you just blindly update 100’s of plugins across so many sites all at once? I’ve always been scared some incompatibility issue might bob up and start breaking things en masse. It’s never happened to me of course … but it seems so likely.
Haha, I wish that was the case would make my days a lot easier.
I pretty much always push updates to staging server, they all have uptime monitoring so I’ll know if I break one and don’t notice it. Then I’ll visually check sites. Anything E-commerce is a whole different story when it comes to updates.
I use it more for the setup of sites, you can group plugins together so they can all be installed in one click. I use it a lot for the Backup Buddy integration. I like to see what plugins / themes need updated across all sites. I like the settings manager plugin. I like the client reporting, Broken links checker integration, branding settings for WP Admin across all sites, you can remove dashboard widgets, add your own, add menu items for client support.