CMS or DIY? Which is better for SEO?

CMS or DIY? Which is better for SEO?

I’m totally guilty of not following my own advice.

And not in a cute “haha, look at me” way. More like: I know the sensible thing to do, I’ve told other people the sensible thing to do, and then I go home and do the opposite because apparently I enjoy manufacturing my own problems.

I started my career building custom PHP and Perl sites for local businesses. Back then, “launching a website” meant you were one bad Apache config away from spending Friday night learning new swear words. Then WYSIWYGs showed up (heck yes, I loved Dreamweaver), and later WordPress and Drupal made it possible for normal humans to publish online without needing to understand why a missing semicolon can ruin your day.

Today we’ve got Wix, Squarespace, Shopify, Webflow, and even Replit. You can ship a decent-looking site in minutes.

Which means, logically, you’d have to be a little insane to build your website by hand.

All of my personal sites are hand-written in Python. Every freaking one.

I don’t know why I do this to myself. Actually, I do know: control. I like knowing exactly what HTML gets shipped, what headers get set, what caching is happening, and what’s going to break when I touch it. It’s the same reason some startups insist on building their own internal tools instead of buying them: they’re convinced their edge is “custom,” when the real edge is usually “shipping and talking to customers.”

Even better, I have two different blogging platforms - one on sethserver and one here on the Black SEO Analyzer site. Why two? Because I built one as a general-purpose blog, and the other as a place to post SEO experiments tied to the tool. Different audiences, different content shapes, different levels of “I might rewrite this whole thing tomorrow because I had a shower thought.”

Meanwhile, Medium exists and is sitting there like, “Hi. I solved this in 2012.”

Does CMS vs DIY actually matter for SEO?

Generally, no; not really.

Search engines mostly care whether:
- You’re actually answering the question the searcher asked (content that helps, not content that performs SEO theater)
- Your site is crawlable and organized (so bots can find the good stuff without getting lost in your nav…IKEA)
- Your pages load on mobile without feeling like dial-up
- Other sites treat you like you’re worth referencing (links and mentions)

A modern CMS helps you avoid a bunch of dumb mistakes. A DIY site lets you control every detail, including the dumb mistakes.

And yes, I’ve seen both in the wild. I’ve had clients pay real money for “custom” sites that were basically a slow CMS with fewer features. I’ve also seen founders slap together a Shopify site in a weekend and start selling while the “custom platform” team was still debating folder structure. Startups love to debate. Customers love when things work.

CMSs have an advantage early on (mostly because they’ve already stepped on the rakes)

Google is generally happy to crawl a fresh Shopify site. And it’s in Shopify’s best interest to make sure your site is technically clean out of the gate, because “my store doesn’t show up on Google” is not the support ticket they want.

A decent CMS stack usually gives you sane defaults like:
- reasonable HTML structure
- an autogenerated sitemap
- canonical tags that are mostly correct
- robots controls you can find without spelunking
- metadata defaults that won’t embarrass you too badly

It also reduces technical footguns-meaning self-inflicted wounds like accidentally blocking Google from your entire site with a single line in robots.txt, or shipping a redesign that turns every page into a 302 loop carnival.

Fewer, not zero.

DIY can work perfectly fine (if you do the basics and don’t get “creative”)

I’ve never had trouble getting a hand-spun site indexed quickly. Google will show up if you make it easy for them and you’re publishing something worth indexing.

My DIY checklist is simple:
1) Your /robots.txt allows crawling (and doesn’t block your whole site because you copied the “staging” version)
2) You have a sitemap.xml
3) Important pages return 200 (not redirect chains, not 500s, not “works on my laptop”)
4) Titles and headings are sane (humans first, robots second)
5) Pages load fast enough that humans don’t leave out of boredom

Google’s SEO starter guide is actually pretty good, and it’s less annoying than most SEO advice on the internet:
https://developers.google.com/search/docs/fundamentals/seo-starter-guide

My recommendation (even though I ignore it)

If you’re starting a new site, especially for a business: use the CMS du jour, or at least the one you already know. The biggest SEO wins early are usually about publishing useful content consistently and learning what your audience actually wants-not about perfecting your handcrafted HTML like it’s going to win a science fair.

Now, the part where I admit the hypocrisy again: I still hand-code my own stuff because I like control, I like tinkering, and I have a long history of choosing the harder path for no measurable business benefit. This is not a virtue. It’s just a personality “quirk”. Hey, I’m quirky, right? …right?

Does DIY secretly help SEO? I only have conspiracy theories. My favorite one is that people who hand-code sites tend to obsess over performance and structure, and then they attribute the results to “custom code” instead of “I spent 40 hours shaving 200ms off load time because I can’t relax.” Another theory: CMS sites all look the same to Google, so the weird little DIY site feels “fresh.” This is probably nonsense, but it’s my nonsense.

If you decide to hand-code everything anyway, just promise me you won’t accidentally noindex your whole site and then blame Google for “shadowbanning” you. Google has enough problems. You do, too.

-Sethers