When I asked Mark Westguard how he got into the web, he didn’t start with a business plan or a bootstrapped success story. He started with a memory, coding as a kid, discovering the web at university, and getting excited about the early days of Netscape.
In our interview, we traced his path from launching one of the UK’s fastest-growing web agencies in the ’90s (yes, the ‘what’s a website?’ era) to his unexpected move into product development. And that path included:
Creating an RSVP tool for his own wedding that turned into a licensing deal with Condé Nast
Running an agency in New Orleans where a team member introduced him to WordPress and WordCamps
Listening to his team’s pain points and deciding to solve one: forms
“Forms felt like putting an alien on the page,” they said.
So he built WS Form, a fully responsive, developer-focused plugin that handles just about anything you can imagine on the frontend, including some pretty creative uses of AI.
Why He Didn’t Replace Support with AI
Mark is thoughtful when it comes to customer support. He sees AI as a tool to assist support, not automate it completely. Repeat questions like “Why won’t my form send email?” could be triaged with AI suggestions, but the final response is still human.
“Support has been one of the strongest points of the product. I want to keep it that way.”
And that’s a philosophy I think more product builders should pay attention to. It’s easy to chase efficiency. It’s harder to stay human at scale.
AI Inside the Product
WS Form also includes AI in smarter ways. You can:
Ask it to build a form (“Make me a mortgage calculator”)
Use OpenAI endpoints to generate images, transcribe audio, or create content
Build a form that outputs a blog post and featured image, then publishes it in WordPress
This isn’t just novelty. It’s practical automation that stays in the user’s control.
Mark’s Advice for Builders
At the end of the interview, I asked what surprised him most about moving from agency work to product work. Without hesitation: support.
The shift isn’t just about building a plugin. It’s about maintaining relationships, documentation, and real trust with users over time.
That’s what makes WS Form work. And that’s what makes Mark someone worth listening to, especially if you’re trying to build a product of your own.
💬 Question for plugin devs and indie builders:
How are you thinking about AI in your product or support flow?
Let’s keep this one going—drop your thoughts on LinkedIn.
I recently had the chance to sit down with Weston Ruter, a longtime WordPress core contributor and someone whose story reminded me why I love talking to people about the open web.
Weston’s first computer wasn’t some fancy setup, it was a Radio Shack Tandy with no drives and a cassette player. From there, he graduated to a Gateway PC (complete with the cow print box) and slowly built his skills through trial and error, Netscape Composer, and eventually diving into raw HTML when the WYSIWYG editors crashed.
Before WordPress came into his life, Weston built his own CMS. All static HTML written to the file system. Basically a static site generator before anyone was calling it that.
WordPress + Open Web = Perfect Match
His first taste of WordPress came via a one-click install from his hosting provider in the early 2000s. But what really drew him in was the project’s commitment to open standards. Weston had been geeking out on W3C specs for fun (yes, really), and WordPress felt like home.
Fast forward a few years and Weston was working at an agency where he helped bring WordPress into client work. Eventually, he started contributing directly to WordPress core, starting with Customizer enhancements that brought widgets into a visual drag-and-drop experience.
He hasn’t looked back since.
Today’s Work: Making WordPress Admin Faster
In our conversation, Weston shared what he’s working on now: improving the WordPress admin experience by making use of the browser’s back/forward cache (BFCache). The goal? Instant page loads when navigating between admin screens, especially for logged-in users. It’s the kind of deep performance work that doesn’t always get the spotlight, but makes a real difference in day-to-day use.
He’s also helping lead efforts around client-side performance metrics, speculative loading, and other clever ways to speed things up without breaking things. It’s technical stuff, but it all comes back to something simple: making the web feel fast and easy.
Why I Loved This Interview
Weston’s story is a reminder that big things often start with small curiosities. A broken web editor. A random blog post. A Perl script sold for $25. That spark of “I wonder if I can build this…” is what pulls so many of us into tech in the first place.
If you’ve ever rolled your own tool, obsessed over site speed, or just wondered how people end up contributing to major open-source projects like WordPress, this one’s for you.
Thanks again to Weston for the conversation—and to everyone out there keeping the open web alive.
Roger Williams (00:00) At what point did content management systems become the thing and did you ever roll your own?
Weston Ruter (00:05) Oh, yes, of course. uh, I called it, uh, well, I was interested in, in HTML. There’s these front link relationships where you can have like parent or in like next and previous. And, and so I made a CMS that was, like family, something
Roger Williams (00:07) you
Weston Ruter (00:28) related to the family and you would create children, pages, and then they would have siblings and it was all HTML written to the file system. I guess static site generator before it was cool.
Roger Williams (00:42) Yeah,
nice. At what point did open source software and specifically WordPress come into your world view?
Weston Ruter (00:49) I had a blog that I made about linguistics back in, I think, 2003, maybe? was through my host. There was a one-click install and it was WordPress. So I think it was, I think it was WordPress at one point, something like super old. And I didn’t do a lot of hacking on it back then, but.
Roger Williams (00:59) Okay.
Weston Ruter (01:15) That was my first exposure to WordPress. then I’ve always really loved just open standards and open web technologies.
Roger Williams (01:27) Hey everybody, it’s Roger with Kinsta. I’m joined today by a new friend, Weston. Hey Weston, how you doing?
Weston Ruter (01:32) Doing well.
Roger Williams (01:32) Awesome. Hey, you know, we connected because you’re a contributor for the WordPress project and, um, you know, Kinsta is helping out support your work with that. And I wanted to just chat with you and, know, obviously we want to talk about the WordPress project. That’s, that’s the most exciting stuff, but I’m really curious to like, get to know you a little bit better and learn kind of your origin story. How did you like get into computers and WordPress and stuff? Take us back a bit. Like, do you remember like your first computer memory?
Weston Ruter (02:01) Yeah, we got a Gateway PC back in I don’t know 1995 or something It came in the cow box the spots and It was just before Windows 95 came out. So we got Windows 3.1 1 1 for work groups, right? and So I that was my first computer
Roger Williams (02:10) Yeah.
Yep. ⁓
Yes, yeah.
Weston Ruter (02:29) Actually, we had a Radio Shack Tandy before that.
Roger Williams (02:32) Yeah.
One drive or two drives?
Weston Ruter (02:35) It had no drives, it had the cassette player that you played the programs into it via. So I remember, I think it was called Star Quest and it was a game that was 2D black and white game with the monitor, you know, molded into the keyboard. That was my video games I had as a kid. And then we had a Pong game too that my mom had.
Roger Williams (02:38) ⁓ okay.
yeah.
Sure, sure.
Weston Ruter (03:00) We were,
my mom, my parents were not all on board with the video games for our household.
Roger Williams (03:06) in a similar boat so with the gateway 95 I mean the internet was right there so it like a ol prodigy how were you getting online right
Weston Ruter (03:14) we had Juno, our email service to dial up to get the emails. And eventually we got GTE. They had, they had a like five megabyte, free website hosting service. So that was my first, I guess, attempt at getting a webpage made. but I had to.
Roger Williams (03:16) Okay. Yep. Yeah.
⁓
Okay.
Weston Ruter (03:36) learn or add to use HTML at that time to be able to do that. And I didn’t know HTML. So I kind of, that was kind of a dead end. then, then I was using Netscape at the time, of course, and Netscape 4 had a composer, application that came with it, that you could do whizzy wig, webpage building. And so that’s how I started learning how to.
Roger Williams (03:46) Yep.
Okay.
Weston Ruter (03:57) how to make web pages and then, but it was kind of buggy. So, and because I was using these heavily table-based layouts, as you did back then, was the nesting was so complicated that it would crash Netscape Composer. So that was what forced me then to dive into the HTML source so that I could make the change I needed to manually because,
Roger Williams (04:05) Yes.
Weston Ruter (04:18) it would crash otherwise if I tried to do it with the user interface.
Roger Williams (04:21) Yeah, it was always fascinating how much those WYSIWYG’s just fell apart as soon as you started just trying to make some slight changes to them. Alright, so then fast forwarding a bit, did you go to university for computer science or anything like that?
Weston Ruter (04:33) Yep. I have a bachelor’s degree in computer science and also minors in linguistics and Spanish. I, I was wanting to study like computational linguistics and I was interested in that intersection, but I did, I found some crossovers, but I still appreciated that experience. I, I use, I still speak Spanish and love learning languages, learning.
Roger Williams (04:35) Okay.
Weston Ruter (04:55) American Sign Language right now as well.
Roger Williams (04:58) Okay, all right, all right.
When you’re in your comp sci, I took one semester of comp sci and the thing that kind of drove me a little crazy in there was nothing was about the web or the internet. was all just, we were doing like 3D graphics and stuff like that. Was that kind of your similar experience or did your program kind of delve into the internet a little bit?
Weston Ruter (05:20) No, it was mostly C++ and just like operating systems and more lower level things. So most of what I’ve learned or what I learned from the web on the web about the web was from getting like my first computer book was Pearl for Dummies. that from getting that from Barnes and Noble and reading it cover to cover and
Roger Williams (05:35) Okay.
Weston Ruter (05:39) I just ate it up and then yeah from there on it was a lot of the dummies books.
Roger Williams (05:44) Sure,
sure. So the old CGI bin directory and building stuff. At what point did content management systems become the thing and did you ever roll your own?
Weston Ruter (05:54) Oh, yes, of course. uh, I called it, uh, well, I was interested in, in HTML. There’s these front link relationships where you can have like parent or in like next and previous. And, and so I made a CMS that was, I think it was, can’t even remember the name. think it was like family, um, something
Roger Williams (05:56) you
Weston Ruter (06:20) related to the family and you would create children, pages, and then they would have siblings and it was all HTML written to the file system. I guess static site generator before it was cool.
Roger Williams (06:34) Yeah,
nice, nice. At what point did open source software and specifically WordPress come into your world view?
Weston Ruter (06:42) I had a blog that I made about linguistics back in, I think, 2003, maybe? was through my host. There was a one-click install and it was WordPress. So I think it was, I think it was WordPress at one point, something like super old. And I didn’t do a lot of hacking on it back then, but.
Roger Williams (06:52) Okay.
Weston Ruter (07:08) That was my first exposure to WordPress. then I’ve always really loved just open standards and open web technologies.
all the W3C specs, I would like read them and like just study all the different aspects of them. And because WordPress was like had web standards as a, as a priority and like, you know, those XHTML valid, buttons and all that, the emphasis on adherence to
Roger Williams (07:37) Sure.
Weston Ruter (07:42) The web standards I think was attractive to me about WordPress, but then I…
Yeah, I started to use it in other projects and I like to do things online and open and not close. So actually there was one closed source project I had back in 1999 called Router Search, which was a Perl program for doing like a site search. I modeled it after Yahoo at the time.
Roger Williams (08:15) Okay.
Weston Ruter (08:15) with all
the search operators before Google was, I guess, on my radar so much. But I sold copies of the licenses for like 25 bucks and I think I made 200 bucks from that. For me at the time that was like living large. But yeah, ever since then everything has been open source. I haven’t sold any software that I remember.
Roger Williams (08:27) Nice.
Absolutely.
Weston Ruter (08:37) But yeah, so WordPress, open source, web technology, open web technology, open web standards, it’s all like perfectly aligned with my interests and started using WordPress at a agency I was working at in 2007 and kind of got them on board with using WordPress for building out their sites for clients.
Roger Williams (09:01) Okay.
Weston Ruter (09:02) That was kind of the launch pad for me to then go fully in on using WordPress for client projects.
Roger Williams (09:09) Okay, okay. At what point did you start contributing to WordPress?
Weston Ruter (09:13) Well, let’s see.
I read a blog post about this that dives into the history and now I think it was 2000-
Nine is my first core props But I started contributing heavily because
At the time I was at XWP and we had a client who we wanted to use the customizer for so that they could have this live preview experience to be able to see the changes that they’re making before they go live. And there were features that were lacking which, which were needed for them. And so I started to
The biggest feature I started working on was the, in the beginning was adding widgets into the customizer. And so instead of, so it was basically like block builder 1.0 in WordPress, where you could have visual drag and drop of blocks, but in the customizer and not using blocks, but using widgets and having the live preview experience there. so we use that for clients to help them build out.
Roger Williams (09:53) Okay.
Weston Ruter (10:12) their landing pages and index pages and things like that.
Roger Williams (10:16) Very cool and from then and then that was it. As soon as you started contributing you were hooked and you haven’t looked back.
Weston Ruter (10:22) Yeah, the work that I did on the adding widgets to the customizer got me heavily involved in the customizer component. then because of that work, I was given commit access to WordPress core back in 2015. And so I’ve been, and then I was a co-lead of the 4.9 release and
Roger Williams (10:40) Okay.
Weston Ruter (10:50) I’ve just been consistently involved in every major release since 3.2, I think, or 3.5. I can’t remember. My blog posted a little.
Roger Williams (10:58) Wow, awesome. Absolutely,
absolutely. So besides the customizer, are there any projects that are now baked in that really stand out to you that you were really proud to be a part of?
Weston Ruter (11:11) Well, the past couple of years, I’ve been heavily involved in the core performance team. And in that team, we have a plugin called Performance Lab, which is like a…
collection of performance feature plugins that are incubating for eventual merge into core potentially pending feedback from the community. And in that project, there’s one called, there’s a plugin called optimization detective, which uses
is a framework for collecting client side metrics for how a page is actually laid out to be able to then apply optimizations more accurately than what WordPress can do on its own. Because say if you have a page that’s laid out and you have an image which is the
what’s called the largest contentful paint. WordPress is gonna guess what that image is gonna be, but it doesn’t know if it’s actually in the viewport or not in the initial viewport. And so you could be that the first image in the post is actually way outside the viewport because there may be paragraphs of text or maybe it’s visible on mobile but not desktop or…
Or vice versa. so this project will, will let visitors to the site contribute those metrics for how the page is laid out to then be able to then better optimize those pages for future visitors so that the actual important elements get prioritized and how they get loaded. if there’s like a, if there’s like a embed that, you know, well,
change its height when it loads, like a tweet or something from Blue Sky or WordPress post embeds, it’ll measure the height that those embeds have after they’ve resized and then reserve the space for that so that when they do load, they don’t cause that layout shift. so a lot of optimizations are possible when you have those client-side metrics that WordPress doesn’t have.
Roger Williams (12:55) Okay.
Weston Ruter (13:02) by default because it’s a server-side, PHP-based environment.
Roger Williams (13:06) Okay, okay, so that’s using all JavaScript and client side operations.
Weston Ruter (13:10) Yeah, well, it’s using, it’s using the, it’s using a web vitals library, from the Google Chrome team, as well as, I mean, that’s an abstraction on top of this underlying performance observer on the web platform. And so it uses that to collect those metrics about the, measurements about where elements are on the page.
which are in the viewport, which aren’t on the viewport. And then with that data saved to the database, then all of those optimizations can be done server-side with PHP, so there’s no JavaScript involved at the actual serve time.
Roger Williams (13:43) Oh, okay.
Excellent. Excellent. All right. So very cool. Very above my pay grade here. I’m swimming in trying to figure out the words here. But sounds really cool. So, you know, today we’re at 6.82 I think is coming out like today or really soon.
Weston Ruter (13:59) Just yeah.
Roger Williams (14:01) just came out. Six nine is on the horizon. We’re kind of getting roadmaps in a little early to start speculating too much on that. But it sounds like you’ve got a project that you’re hoping to get in there. Can you tell us more about that?
Weston Ruter (14:14) Yeah, there’s one of the new features in 6.8 was this thing called speculative loading, which allows you to navigate more quickly when you to other pages on the site by if you hover over the link, the browser will start to or actually by default in WordPress, the it’s the conservative eagerness so that as soon as you click, as soon as you mouse down or touch
Roger Williams (14:19) Yeah.
Weston Ruter (14:35) down on the link, it’ll then start making the request for the page before you release the mouse click or release the finger. But you can have it do more aggressive pre-rendering of those pages so that the navigations appear to be instant because the browser has loaded the page that you are going to already in the background. So it’s
basically the same as like switching tabs in the browser. Well there’s another much older ability related to that and it’s called the back forward cache or BF cache and if you navigate around a random WordPress site you’ll notice and you’re not logged in that navigating I mean most of the time unless there are certain
scenarios where this BF cache is blocked from working. But if the BF cache is working, if you navigate from page A, B, C, D, E, and then you hit the back button, then those back button navigations will be instant. The previous page will show up instantly. And then if you go forward, the next page will show up instantly. And that’s the BF cache. But if there is a unload event hit listener or
the page uses WebSockets or the page is served with this no-store cache control response header, then the browser won’t put it in the BFcache, and so then the browser will basically have to rebuild the page from scratch. Whereas when it’s in BFcache, the entire snapshot of the page
All the JavaScript, all the DOM, everything is stored there in memory so that it’s immediately there, available to you to go back to. But if, yeah, bfcache is not available, then you’ll notice that there’s, it feels sluggish to then go back and forward. And that’s the experience you get now in the WordPress admin because all pages are served in, when you’re logged in, all pages are served with this no store.
directive for the cache control header. And that was done in part to prevent those pages from being cached by like a proxy, like varnish or something. But also for privacy, because if you are working on something super sensitive and then you log out, well, if the pages you were working on aren’t bfcached, then you could just hit the back button to go back.
Roger Williams (16:27) Okay.
Weston Ruter (16:40) and see what those are, what was on the page, even though you wouldn’t be able to make changes because you had been logged out. So for those reasons, the new store directive was added, but there are alternative ways to preserve the privacy by this page show event listener and other mechanisms that I’m working on.
Roger Williams (16:58) Yeah.
Weston Ruter (16:59) but, and then there’s this, no, private directive for the cash control response header, which is what actually is intended for preventing those pages from being cached in by proxies. So that’s already there in WordPress. And so what, what I’m working on is, is there’s a track ticket and then I have a plugin that’s pending for the.org directory. It’s called no cash BF cache.
and it will allow you to when you’re in the login screen there’s with a checkbox that says the checkbox that says remember me along with that there’s this new little sparkle button and it will say if you click on that it opens this popover that says now new feature if you remember me it’ll also enable bfcache and so when you have
Roger Williams (17:28) Yeah.
Weston Ruter (17:41) then logged in, it will then omit that no store directive from the response header and then also include JavaScript on the pages to make sure that after you’ve logged out that those pages in bfcache get evicted so that the privacy is preserved.
Roger Williams (17:57) Okay, alright, so you’re going to single-handedly improve the speed of WP-admin.
Weston Ruter (18:03) Well, that is the hope that, I mean, in my own use of the WordPress admin, was feeling it was painful for me to navigate around. So I wanted to improve it for myself. And so, and actually a couple of years ago, was, I think it was WordPress 6.5. There was a track ticket for the use of unload event.
Roger Williams (18:14) Sure.
Weston Ruter (18:24) listeners in WordPress admin, which are another reason why BFcache is blocked. And so I was thinking, I was excited because I got that committed and I was saying, okay, now we’re going to have BFcache. But then I didn’t realize that at the same time or shortly before this other commit came in that added no store. So all that works, I removed, unload was invalidated by the no store. So it’s revisiting that.
Roger Williams (18:44) you
Weston Ruter (18:50) that ticket from 6.5 and hopefully will allow for a much faster navigation experience in admin. And also on the front end too, because if you’re logged in, this no store directive is sent to every request while you’re logged in. let’s say if you have a WooCommerce store where you’re logged in as a customer or BB Press or Buddy Press or membership sites, anything like that.
you’ll be able to benefit from this BF cache to speed up navigations.
Roger Williams (19:18) Very cool. I’m looking forward to this happening because I too would like to see things sped up a little bit in the admin. As far as working on this goes, if other contributors are interested in working on it with you, what does that look like in this instance?
Weston Ruter (19:32) There’s a track ticket that is on the WordPress core track. And then I have a repo on GitHub that is open and available for anybody to contribute to.
Roger Williams (19:41) Excellent. All right. All right. We’ll grab a link for that and throw it into the notes here. The time has flown by I don’t want to take up too much more of your time, but I’d be remiss to not get at least some of your excuse me. Get some of your thoughts and insights on the hype of the day AI. Are you a vibe coder? Do you utilize these tools when you’re working?
Weston Ruter (20:00) I don’t vibe code. I haven’t tried it, but I do use some of the auto completion features in PHP storm. I have, I use, actually I enabled a Gemini code assist for GitHub, which I’ve been really happy with because it basically adds a AI coder viewer to your, all your projects. And actually I learned about it from Adam Silverstein.
Roger Williams (20:13) Okay.
Weston Ruter (20:23) because he had a code for his. And so whenever I open a pull request and Gemini code assist will give me a summary of the changes in the pull request and then it will give me inline code review comments about areas of concern or, and then it’s caught things that I missed. So really useful.
Roger Williams (20:42) Okay, so almost like a grammar leaf for coding.
Weston Ruter (20:45) Yeah, I don’t use Grammarly, but yeah.
Roger Williams (20:46) Okay, all right.
Am I am the ultimate of laziness so I use all of these tools although I don’t code But if I did I would definitely use them all too So that that’s great. I Appreciate the insight on you know using AI, but you know maybe being a little bit skeptical of it Maybe not you know both feet jumping into the deep end. Is that pretty accurate?
Weston Ruter (21:08) Well, I mean, I will use Gemini, I have Gemini subscription, like Gemini Pro, whatever it’s called. So I will use it to brainstorm ideas and get code reviews even like in the Gemini web app. But in terms of the IDE, I’ve still been exploring the best like, I use PHP Storm, I’m trying to figure out the best.
Plug in and there’s various options, like all competing, Gemini, chat GBT, and then JetBrains is own plug-in. So I’m still like waiting to see who’s gonna win me over.
Roger Williams (21:41) Okay, okay, all
right, all right. So you’re being wooed, I love it. This has been great, Weston. I really do appreciate your time and giving us some insights into like your history and how you approached computers and the WordPress project. If people wanna get in touch with you, ask you questions or pick your brain and whatnot, what’s the best way for them to reach out?
Weston Ruter (21:59) Probably blue sky or I mean I’m on blue sky Twitter LinkedIn You can send me a contact form message on my my blog. So yeah, it was very very nice
Roger Williams (22:11) Excellent, excellent. All right, well, we’ll have links for all of that in the notes for people to reach out and ask you all types of questions. Really great talking to you. I look forward to speaking with you again soon. Have a great day. What’s that? WordCamp US, we’re, six weeks out. So I’ll see you there and I’ll hopefully see everybody else there as well. All right. All right, Weston, have a great day,
When Brad Williams talks about building websites, he doesn’t start with plugins or page speed. He starts with a Commodore 64 and a number-guessing game he wrote when he was eight.
That moment, typing BASIC code from a book into a machine his family bought off a cousin ,ignited a career that would take him from Marine Corps programmer to co-founder of WebDevStudios. Along the way, he taught himself HTML on PageMill, got made fun of for using Macs in the ’90s, and learned that formal schooling wasn’t where he’d thrive.
So he joined the Marines. They trained him in Visual Basic, taught him to think in arrays and conditionals, and gave him a security clearance to work on military intranet systems he still can’t show in a portfolio. It was structured, intense, and formative.
After leaving the military, Brad entered the web professionally, and eventually stumbled across WordPress while at a conference in 2006. That moment shifted everything. He became a tinkerer, a plugin hacker, and then a WordPress agency founder. By 2010, he and his team went all in on WordPress.
Now, Brad’s latest project, ThemeSwitcher Pro, is another bridge. It’s designed for people stuck in the Classic Editor, helping content teams migrate one page at a time to block-based editing, no dev time required, no redesign budget needed.
It’s a clever plugin. But more than that, it’s a continuation of Brad’s whole philosophy: solve real problems with practical tools, and make the modern web a little more accessible for everyone.
I met Rob Harr at the Agency Builders Conference in Florida. He wasn’t trying to impress anyone—but he impressed the hell out of me.
He spoke about agency operations with this calm clarity, like someone who’d seen behind the curtain, broken the machine down to parts, and built it back up better. His message was simple, almost annoyingly so:
“Most people already know what the right thing to do is. They just don’t do it consistently.”
That stuck with me. Because I’ve seen it too. Whether you’re running paid ads, launching client sites, or just trying to follow up on leads—you already know the playbook. But ideas aren’t the problem. Execution is. More specifically: consistent execution.
Rob calls it the accidental agency owner problem. You start as a designer or a developer. You’re great at your craft. Suddenly, you have clients. Then a team. Then payroll. But no one taught you how to run a business. You’re still thinking like a technician, not an operator.
That’s where Rob shines. He helps agencies zoom out. Build systems. Break goals into repeatable actions. Treat their agency like the product.
We also talked about AI, enterprise work, and Rob’s origin story as a second-generation software engineer. He got his start writing code at 12. Landed at LexisNexis. Survived the banking meltdown. And eventually co-founded Sparkbox, where he still works today. He also runs an ops consultancy called Upwell.
You can watch the full interview below. But if you take away nothing else, take this:
“Goals are good. Systems are how you reach them. And consistency is everything.”
I recently sat down with Aaron Jorbin, core committer, bow tie enthusiast, and all-around WordPress sage, for a live bug scrub walkthrough. Spoiler: I survived.
Before we get into it, a quick confession. Despite working in tech for decades, I’ve never consistently contributed directly to WordPress. Sure, I’ve filed a bug or two, but the inner workings of Trac always felt a bit… intimidating. Like trying to do yoga in a crowded elevator.
The good news is that open source isn’t just for the people who write code in their sleep. It’s for all of us. And thanks to a bit of mentorship from Aaron, I finally dipped a toe (okay, maybe a foot) into the waters of contribution, and didn’t drown.
So, What’s a Bug Scrub?
A bug scrub is basically a group effort to triage open tickets in WordPress. You don’t need to be a developer to participate. In our session, we focused on UI copy, little things like button labels or notification text that impact user experience.
Aaron and I reviewed a few tickets:
One where a dropdown said “Not Set” instead of “Default” (spoiler: it’s already been fixed in recent releases).
Another about unclear email notification wording.
And a third where failed plugin updates just awkwardly… hang.
Tag issues with workflow keywords like close or needs-patch
Not break anything (success!)
The Power of Mentorship
Aaron’s guidance wasn’t just helpful, it was permission-giving. There’s something powerful about a seasoned contributor saying, “You don’t have to be perfect. Just be helpful.” That ethos, of collaboration, experimentation, and learning in the open, is what keeps WordPress strong.
In open source, progress happens in increments. You don’t need to write a patch. Sometimes just leaving a clarifying comment or verifying behavior in the latest release moves a ticket forward.
As Aaron put it:
“Each of the ones we looked at today, moving it forward meant something different.”
Advice for the Curious
If you’re like me, interested but unsure where to start, here are a few tips:
Use the Playground to test things quickly without setting up a full environment.
If you’re unsure whether a ticket should be closed, use the close keyword instead of marking it resolved. Let experienced folks review.
Be kind, curious, and open to learning.
Also: don’t underestimate how helpful it is just to verify whether an issue still exists in the latest version. That alone makes a difference.
You Can Do This (Yes, Even You)
I’m pushing 50, I mix up keyboard shortcuts daily, and I still managed to contribute in under an hour. So yeah, you can totally do this.
The WordPress project runs on volunteers. You don’t need a permission slip. You just need to show up with good intent, a bit of time, and a willingness to learn.
If you’re curious where help is needed right now, Aaron recommends checking out:
Kevin Leary is one of those rare developers who’s just as comfortable diving into analytics and design as he is refactoring code. I caught up with Kevin for a fast-moving conversation that covered AI, web development, and how a career that started with graphic design and assembly code led him to running his own successful consultancy.
On Learning to Learn
Kevin studied both computer science and graphic design at Champlain College and got his start working for agencies in the Boston area. What stood out to me was his mindset: he wasn’t afraid to pivot or jump across disciplines. In fact, that curiosity helped him go independent.
“I like to jump from design to dev to analytics. At first I thought that wasn’t a good thing, but I realized that’s what drives me.”
That ability to synthesize across disciplines is what makes Kevin’s work so effective today—he’s not just building sites, he’s helping clients improve performance and user experience in measurable ways.
How Kevin Thinks About AI
We had a great moment in the interview where Kevin described how he approaches conversations about AI with clients:
“I talk to clients about AI almost the same way I talk about any technology. First, what do you want to do? What are your goals? Then we figure out if AI helps with that.”
It’s refreshing to hear that kind of clarity. Kevin uses AI daily in his own coding workflow but is careful not to overhype it. He points out that without a deep knowledge of the codebase, AI-generated code can actually slow you down. That honesty is something a lot of developers (and clients) can relate to.
WordPress, Spam, and Smart Use Cases
One AI use case Kevin mentioned was especially clever: he used OpenAI’s API to score contact form submissions for spam instead of relying on captchas or Cloudflare Turnstile. By scoring messages and reviewing their accuracy before putting the filter in place, he created a custom, client-friendly solution to a very real problem.
Hosting That Doesn’t Waste Your Time
We also talked a bit about Kinsta, and Kevin didn’t hold back:
“I always say this to everybody: the amount of time you’ll save by spending an extra $200–300 a year is so astronomically higher than the amount of time it’ll take if you go with the cheaper option.”
He’s been a fan since 2016 and still recommends Kinsta today because, as he puts it, it just works. We’ll take it!
Who Kevin Works With
If you’re running a custom-built WordPress site that originated with an agency and now needs ongoing performance, analytics, and UX improvements—Kevin’s your guy. He’s happiest working with companies who already value good design and want to keep improving.
Full Transcript
Kevin Leary (00:00)
I talk to clients about AI almost the same way I talk about any technology. First, what do you want to do? What are your goals? And then once you determine what we’re setting out to do, maybe AI is good to do it. But the red flag for me is if someone comes in and says, we have to use AI because that’s what’s going on. Everyone else is doing it because there are a lot of cases where it doesn’t really make sense. But for me, I’d say I use it day to day just because I think that is where
coding is going. So you kind of have to learn how to work with it. You definitely can improve your workflow, but it’s also a little bit eye opening to see that you still really do need a deep knowledge of the code base that you’re working in. Because sometimes one or two things are off, and if you can’t find that, it may end up taking more time than if you had just written something yourself.
Roger Williams (00:38)
Okay.
Hey everybody, it’s Roger with Kinsta. Today I’m joined with a new friend, Kevin. Hey Kevin, how are you?
Kevin Leary (00:52)
Good, how are you Roger?
Roger Williams (00:53)
You know, I’m doing really well. It’s the middle of the week. Things are cruising along. So I’m pretty happy with stuff. I’m excited to be on this call with you and learn more about you from looking at your website. You you’ve got a huge variety of technologies that you’re involved in. So, you know, to like get us started here, can you kind of tell us how you got into the web and big data and AI stuff?
Kevin Leary (01:16)
Yeah, definitely. So, I mean, I’ve always been a programmer, more or less. I went to school for graphic design and computer science, did a little bit of work during college at an agency. And then I worked at a couple of agencies in the Boston area. They were very good, very great. I liked working there, but I found that like, I like to jump around from role to role too much for an agency. And at first I thought, well, maybe this isn’t good. Maybe I should just focus on development. But then I realized that’s kind of what
drives me almost, like the ability to jump from design to dev to analytics. And that’s where I sort of decided to go off on my own and pursue all of those things. So I’d say to answer how I got into WordPress, way, way back, I just was poking around with CMSes. And I saw it, and I was hooked, basically. I could just see that it’s very interesting to see how the themes are set up, to dissect them, things like that. And then with
Roger Williams (02:02)
Okay.
Kevin Leary (02:09)
BigQuery and data warehousing stuff. I got into that by way of originally working with Segment, which I believe now they’re owned by Twilio. But it was before the new GA had the event-based stuff. It was just an approach that made sense to me to be able to track a business event when this happens, another one when this happens, another one when this happens. So I set that up, connected it to BigQuery way, way back. I think it was like 2014.
maybe, 2012, 2013. And that was really valuable. And so that kind of steered me towards using analytics to help measure how effective that design and dev work was. I think that, is that a good explanation? I that covers it.
Roger Williams (02:52)
Yeah, no,
no, so I want to go back and dig into things here a little bit. So you said you’ve always been a programmer. Did you go to college or school for any computer science or anything?
Kevin Leary (03:01)
I did, I went to Champlain College in Brentwood, Tidumon.
Roger Williams (03:05)
Okay, and you studied comp sci and got all into assembly code and everything?
Kevin Leary (03:10)
Yeah, was kind of it was like a hybrid, almost self created program where it was half graphic design and then it was half computer science.
Roger Williams (03:17)
Okay, very cool. Anything from that experience that you still kind of pull from today?
Kevin Leary (03:23)
I mean, to be honest, maybe not at college, but one thing I learned then was like to really know how to do things well, you’ve learned on your own. I do a lot of reading, any new topic I’ll dive in and try to understand it first before I work with it, which is sometimes hard because often you have to dive right into it. But yeah, learning to learn, guess you would say, which is a little bit cheesy, but it’s really ultimately what, what I came out with.
Roger Williams (03:47)
No, absolutely. I think a lot of times people go to higher education and they have like a direct goal in mind. And maybe sometimes they get disenchanted because it is really about learning how to learn. At least that was my experience as well. As you were starting to work with agencies, what kind of work were you on? Were these like enterprise level sites, small medium sized businesses?
Kevin Leary (04:00)
Sorry.
I’d say medium, not too much enterprise right away, but medium to large businesses at some of them. There were some design sites that I did from scratch with a brand that was like, I’d say smaller, more or less. And then my dev skills were a little bit stronger. that was where I was put. That’s where I did well. ⁓ But yeah, at the time it was like Symphony apps. There was really no WordPress at the agency.
Roger Williams (04:28)
Thanks a lot.
Kevin Leary (04:34)
It was like jQuery was new, not to point out how old I am, guess, but yeah, that was the case. Actually, the first agency job I had, I remember I went in and I was like, gosh, I got this thing jQuery. And I remember the developer there was like, no, I’m using Scriptaculous. It’s way better. And anytime I mention that, people are like, what’s Scriptaculous? Nobody knows what it is.
Roger Williams (04:34)
Okay.
There’s a giveaway.
you
⁓
Exactly, All right, very cool. then fast forwarding, it sounds like you’ve got some squirrel syndrome, which I suffer from as well. I like to go after the shiny new things and try all the things. As you started your own business, how did you start finding clients initially and have you niched down as it were?
Kevin Leary (05:14)
Yeah, too. I guess in terms of the squirrely scenario, too, it’s more, I guess, that I find that if you want, let’s say you’re trying to increase leads on a site. You can do it with design, but it’s only a piece of it. You can do it with analytics, but it’s only a piece of it. If you know how to code, you can look in and see maybe people on an Android phone are having issues with this. It’s sort of all dove into all of it to gear it towards like,
making those things better and when you can do all of them, great. But it’s all encompassed in that too. It’s not necessarily some of the things are not new. They’re just in that space. But yeah, sorry, what was the other part of the question? Second one?
Roger Williams (05:52)
Well, so then
as you figured out that you wanted to go out on your own, what were some of the first steps to get that really started and cemented?
Kevin Leary (05:57)
yeah.
So initially, did it outside of sort of pulled back my hours at agencies and worked a lot of night and weekend hours to build it up because it would be very hard to do a hard switch to it. But initially, a lot of it was SEO. So it’s pretty good at it. Could rank pretty well locally for things like Boston WordPress Developer and other stuff like that. And then the real, probably the larger clients that
have been great to work with in their more long term. A lot of those came from that original agency work 10 years down the road. Like they just reach out and they’re like, you know, we worked together on this. I don’t know if you remember, if you’re available and that that’s probably the best source of work referrals to be honest. Yeah.
Roger Williams (06:40)
Okay, absolutely.
Yeah, when people have worked with you and have an idea of what you can do and how you operate, that’s always a big bonus. Fast forward to today. I’d be remiss not to bring up AI. It’s everybody’s favorite subject. I take it at this point, you’re well over the hurdle of should we be using AI? Where are you?
Kevin Leary (06:53)
Yeah.
Roger Williams (07:01)
focusing your efforts and how do you talk to clients about AI?
Kevin Leary (07:05)
I talk to clients about AI almost the same way I talk about any technology. First, what do you want to do? What are your goals? And then once you determine what we’re setting out to do, maybe AI is good to do it. But the red flag for me is if someone comes in and says, we have to use AI because that’s what’s going on. Everyone else is doing it because there are a lot of cases where it doesn’t really make sense. But for me, I’d say I use it day to day just because I think that is where
coding is going. So you kind of have to learn how to work with it. You definitely can improve your workflow, but it’s also a little bit eye opening to see that you still really do need a deep knowledge of the code base that you’re working in. Because sometimes one or two things are off, and if you can’t find that, it may end up taking more time than if you had just written something yourself. ⁓
Roger Williams (07:44)
Okay.
That makes sense.
Absolutely. I run into the same thing where I’ll get it to write something and it just doesn’t feel right. And I’m talking about text, not not code even. And I’ll just end up using it as like a blueprint instead of what I’m actually going to end up with. So that that’s really cool. When it comes to like WordPress and AI. Are you seeing any solutions yet that are
Kevin Leary (08:08)
Yeah.
Roger Williams (08:17)
that are either client facing or customer facing that you’re implementing.
Kevin Leary (08:20)
Yeah, there’s one that I wrote about recently that’s kind of interesting. I won’t mention who because it’d be best to do that. But spam with gravity forms is a big problem. Anyone who has a form on any WordPress site, it’s rampant. It’s almost impossible to block it. But because it’s a low-volume site, there’s not too many submissions going into this contact form. We used OpenAI APIs to basically just say, hey, here’s the message. Does it look like spam to you?
give it a score of 1 to 100. And then based on that, what we did initially was it just put the score at the top so the internal team could review it and see, like, is this effective? We found it once. It was really good. So now we replaced, what is it, Cloudflare turnstile with this AI approach. It works well so far. I guess the thing is, like, you know, it works well until the AI bots figure out how to beat the AI spam filters.
Roger Williams (09:13)
you
Kevin Leary (09:13)
And then I
guess you’re in this big loop, right, where they just go back and forth. But for now, yeah, it does work pretty well. Also, lot of content creation, too. This is pretty common.
Roger Williams (09:20)
Excellent, excellent.
Definitely. As you’re implementing WordPress sites, I’d be remiss to not bring up and mention Kinsta for hosting. Is there anything specific about working with Kinsta that you really appreciate?
Kevin Leary (09:34)
Yeah, definitely. First, I’ll say it. I have no incentive to say this, right? But I really like Kinsta I started using it way back. I don’t know when it was founded. I could be wrong about this, but I think it was 2016 or 17 for a site that’s still online today. It’s proto.life. And they don’t publish articles anymore, but it was great, to be honest. At that time, a lot of people, was WP Engine was the only default. And there’s still both.
Roger Williams (09:47)
Okay, yeah.
Kevin Leary (10:03)
great, but Kinsta’s been, since then, very fast. Site hasn’t really went down the server in any way. And anytime there’s an issue, it’s easy to get a response from it. I would say the big thing I tell people, it’s actually what I recommend, to be totally honest, on Reddit. If somebody’s wondering about a host, it is who I recommend. A lot of the feedback you get is, could get a site for $5 or $10 elsewhere.
Roger Williams (10:21)
Excellent.
Kevin Leary (10:26)
I always say this to everybody, the amount of time you will save by just spending an extra $200, $300 a year is so astronomically higher than the amount of time it will take if you go with the cheaper option that it’s just, for me, it’s a no-brainer, to be honest. And that sounds very promotional, but it is not. It’s totally 100 % honest.
Roger Williams (10:44)
Ha
No, and I appreciate that you were not primed in any way. But I’m with you on that, managed aspect of having somebody there who’s watching the servers and if there is an issue, resolving it, having backups that actually work, it’s priceless. And then having support that can actually answer your questions and get things figured out for you. Kevin, I like to keep these short and sweet. This has been…
Really an enlightening conversation. I appreciate your time. If people want to get in touch with you, if they want to hire you, what are the best ways for them to reach out?
Kevin Leary (11:16)
Yeah, definitely. So kevinleary.net is the best place to go. You can submit the form there. And it doesn’t have AI spam blocking yet, but it might soon. And also just hello at kevinleary.net is another email that you can write a chat on.
Roger Williams (11:31)
Excellent. And as far as like ideal customers, who do you enjoy working with the most?
Kevin Leary (11:37)
I would say my sweet spot, so you mentioned the niche before, I didn’t really mention it, but to be honest, I inherit a lot of custom built agency sites. So maybe somebody pays a lot for a brand and they get the brand online, but then they want to maintain that and grow leads, try to get more business coming through, try to make it more effective at achieving those goals. That’s probably my sweet spot, I think, because I can set up analytics.
help with the design of certain flows, develop it out, build it, see it grow, and then iterate on it.
Roger Williams (12:10)
Excellent, excellent. All right, well, Kevin, I appreciate your time. I hope that at some point I can have you back and we can talk more about some SEO and some AI and all the fun stuff. Thanks. Yeah, yeah, you’re welcome. All right, Kevin, we’ll talk to you soon.
Kevin Leary (12:18)
Yeah, I’d to. Yeah, I’d love to. Thanks again for reaching out, too. It’s been great.
In this episode of Kinsta Talks, Aaron Jorbin shares how anyone who uses WordPress can start contributing — by testing the bugs targeted for the upcoming 6.8.2 release. Learn how to use the beta testing plugin, how to write helpful bug reports, and why your voice matters even if you’re not a developer. We wrap up with a short bonus on submitting to WordCamp US.
💡 Key Quotes:
“Now is not the time to fix your personal wishlist bug. Now is the time to fix the problems introduced in 6.8.”
“It’s impossible to test every plugin combination. That’s why we need the community.”
“There’s no wrong way to report a bug. There’s just incomplete or unkind ways.”
“A video can be worth thousands of words. If you can’t describe it, record it.”
One of the best parts of doing interviews for Kinsta Talks is getting to sit down with people I already know—but haven’t yet had the chance to really dig in with. That was the case with Kelsey Parks, a local friend, fellow WordPress Meetup organizer here in Durango, and the co-founder of Psyche Digital.
We recently sat down to talk about her journey into marketing, what she’s learned from years of working with startups, and how her agency helps founders build the kind of systems that don’t just survive—they scale.
From Startups to Systems
Kelsey’s path into agency life didn’t begin with marketing degrees or a traditional agency climb. It started with startups—and with chaos.
“I got completely addicted to the unemployable startup founder,” she told me. “The oddballs in a warehouse, making cool things happen from thin air.”
After stints in D.C. and China—where she helped Chinese companies position themselves for North American markets—Kelsey eventually returned to Southwest Colorado and teamed up with her childhood friend (and now business partner), Kate. The two had stayed connected since the third grade, and they launched Psyche Digital together to serve companies during one of the most intense phases of business: the early growth stage.
Why Systems Matter
Kelsey works with founders who are often at a critical juncture—what she calls the “we’re going to make it or we’re not going to be here in a year” phase. Rather than overwhelm them with trendy marketing tactics, Psyche Digital focuses on something more sustainable: building repeatable systems.
One line really stuck with me:
“Another thing that’s important to startup founders is knowing that the work we’re doing isn’t going to be lost when they bring in their own marketing team.”
That mindset—treating systems as something founders can keep—feels like the kind of long-term thinking a lot of fast-moving businesses forget to prioritize.
On AI and Interns
We also talked about AI, of course.
Kelsey doesn’t see AI as a threat. She sees it as a tool. Specifically:
“I think of AI as part of my workforce—more like an intern in my workforce.”
That framing resonated. Use AI to gather ideas, generate outlines, or even write drafts—but then let your experts take it from there. It’s not about replacing creativity. It’s about removing bottlenecks.
Meet Her at the Meetup
In addition to running an agency, Kelsey co-hosts the Durango WordPress Meetup with me. She’s been instrumental in building a welcoming space for folks working with WordPress across the Four Corners.
If you’re a local and haven’t joined us yet—now’s a good time. We meet monthly, and our conversations are always lively (and occasionally spicy—like Kelsey’s favorite Sichuan food from her days in Chengdu).