Tuesday, July 14

Entrepreneurial Spark

David Heinemeier Hansson (DHH) — "It's AI accessible, but not agent hysteric... As our entire industry is getting swept up in agent hysteria (and I love AI as much as anyone!), we thought it better to focus on the human communication that's the cornerstone of Basecamp. The new site just speaks plainly to that mission and shows you the software right at the top."
Source

Chris: We all now have the ability to create even more noise than we possibly could. When social media came out, it was initially meant for connecting with friends on a more regular basis, right? When the algorithm came up and companies decided that they could use it to push their products and this and that and the other, we always want to make things simpler and easier, and so we try to automate stuff. The idea that we can basically just push out features to systems by writing a couple of sentences is very tempting to make products do anything and everything you want them to do. At the end of the day, are those automations just noise and actually not allowing us to do the type of work that we need to do?

Right now, I agree with him. There's this huge push toward adding AI into everything as humanly possible, and that no longer is a selling feature. Wasn't about five minutes in time when you could say that you had AI built into your platform, but now pretty much everybody has that. I really like their stance: our tool is still the tool -- you still use it exactly the same way. It has a new UI and some different features to it, but a lot of it is very similar. If you want to interact, you want to pull down your stuff through an MCP, you can do that. There is a value to having all of your tasks being submitted into your system, into your chat, that makes life a lot easier, but laying into Basecamp, this idea of having everything be AI-itized is very much against their thesis.

An example of this and what I've been doing is that, historically, folks come to me for websites. One of their first things is, "We've outgrown our website," and a lot of folks are talking about their Wix site or their Squarespace site, or they're just very, very old WordPress site that just got bloated over the years. I envision that, over the next three or four or five years, the folks that are actively building their websites with AI completely right now are going to actually take the place of the people that used to come and say, "I need you to build me a better website because Squarespace doesn't cut it anymore." They're now going to say, "I want you to build me a website because the one that I DIY'd myself in AI doesn't work anymore."

The thing is, I'm using a lot of AI tools behind the scenes for things, and it's a lot of the stuff that nobody cares about. It's the content importing process. It's the ability for me to take your colors and your fonts and a little bit of content and throw together a very basic color palette wireframe to show you on call number one: "This is how your stuff will interact together in a consistent way and get quick wins." I can't tell you, historically speaking, our website process has taken no less than three months. From the discovery call, it's typically taken two or three weeks to get to a design, which then has to be coded out. Now, if we get enough information ahead of our discovery call, sometimes we're able to show a very basic color palette wireframe, and then by meeting number two, within a week or two, we're able to actually showcase a full-blown home page. Once we have a good understanding of the basics, a lot of the stuff that we have already designed works well across a number of different types of clients. We're able to then spend time not designing the same parts and pieces over and over again, but on the two or three elements that actually need some TLC, that need to be uber creative. We can focus on the texture and the parts of the site that actually make it feel human, more so than designing a really slick testimonials block when a standard one with their colors and their fonts will do.

Developed into a blog draft (2026-07-14): "We've Outgrown Our Website" — ChrisLaFay.com personal essay, crafted from this spark using the Matthew Dicks storytelling framework. Draft at blog-2026-07-14-weve-outgrown-our-website (Classic City/Strategy/content-queue/); mirrored in the Workbench (slug blog-2026-07-14-weve-outgrown-our-website, status review).

Meetings Processed

• 9:00 AM | Client | Georgia State Election Board | 2026-07-14-georgia-state-election-board | 2026-07-14 - Georgia State Election Board - Client | ✉️ follow-up owed → Hope Coan + Sherry Phillips

The Josh Hays Question

I have a loaded question. These days, how much does “custom”, “hands on”, “human driven”, “handcrafted”, etc matter to the people buying a website from Classic City?

What people are saying

I liken this to the adage from the last 5-10 years of "I'm tired of my website looking like a Wix website." They don't necessarily say those exact words, but that's the intent behind them. It may come across like:

• My website looks dated
• I can't edit my website
• All of our pages look like different people made them
• What our website says doesn't line up with who we are

You've heard all of these, too. There's nothing new under the sun. So, while the technology has changed, there is still a plethora of people out there saying the exact same things. And the thing is, they WANT something doesn't feel like a template (that is another line I hear often).

A trial from years ago

Probably 4-5 years ago, I marketed a "30 Day Website." That's exactly what I called it. It was $5K and we knocked everything out in 30 days. I had a WordPress template that we had built that had a handful of blocks, and we'd just write VERY basic CSS on top of it and be done. This worked well for folks who already had a website (and literally just needed it ported over) or folks that needed a "one page landing page" for their new business.

However, I got all the "cheap'o's" knocking on my door. I mean... there isn't really any surprise there. These are folks who simply wanted the easy route and didn't want to think about it, and they obviously complained a lot.

The product in-and-of itself wasn't BAD, it wasn't messaged correctly nor pitched correctly.

These people were obviously sold a system/template/structure that they knew they were getting into. It wasn't custom and it wasn't meant to be.

The permission-less demo

One thing my main business coach keeps coming back to is this concept of a "permission-less demo." That was a new phrase to me, so I'll define it:

When someone uses your website (or ingests your own marketing materials), those "items" they consume ARE your product. They got to experience YOUR PRODUCT by partaking in your marketing.

So for me, that's having a dope website that drives people to a call to action. I can say to fresh leads: "Look, you saw my website and how it worked. We're gonna do the same for you!"

So, to answer a part of your question a bit more directly now: my website SHOWS that what we do is custom without having to put it on blast using our words. In fact, a client told me yesterday during a call: "When Kevin told me we needed to work with you, I asked him to send me your website. I was on it for about 10 seconds and immediately thought: 'This is who I want to work with. Sign the contract.'" (for reference, this was a cold client who didn't know me).

That guy NEVER said: "I want a custom/hand-crafted website." He just wants something that feels hand-crafted.

Not pushing AI

I don't talk about AI on our website at all and that is very intentional (I'm sure I have some blog posts that mention it, but it is NO WHERE on our primary pages). The reason is simple: I'm not selling a "cheap-o $5K starter website." If I were to say my website build process was powered by AI behind the scenes, people would flee.

They wouldn't say "This isn't custom! It's not worth $15K+!" but they would feel like its lesser quality.

So now, when I get on the phone with people, I focus in on: "The reason we can do ALL of this for $15K is because we've done the legwork on designing the building blocks you don't want to pay us for (aka. testimonials, image cards, etc...). That allows us to focus on the 3-4 special things that will make your website pop." And THAT is what is interesting to people.

I also tell people that if they send me over design inspiration and their brand packets ahead of time, I can have some basic comps done during our initial meeting. Not as final designs (and I overly emphasize this), but as a simple directional marker to see if we're heading down the right path. And, 2 out of 6 people I've done that with have said: "I actually really like this as it is; can we just expand on that?" (which was the plan anyways).

Summary

Yes, people want "custom," but they aren't saying it. They want something "better and easier to edit." Yes, the words "Custom WordPress" are on our site, and I think that is important. BUT... they don't care what parts and pieces are custom. They don't care how much of the "crappy work" is done by AI. They just see that they get a robust site that is WAAAAAY better than the crap they see online now, AND its consistent, AND its easy to edit, AND they like my team.

Developed into content (2026-07-14)

Crafted from this answer using the Matthew Dicks storytelling framework:

• Blog cornerstone (ChrisLaFay.com): "Nobody Asks for a Custom Website" — anchored on the 10-second cold-client "sign the contract" story + the permission-less demo. Draft at blog-2026-07-14-nobody-asks-for-a-custom-website; Workbench slug blog-2026-07-14-nobody-asks-for-a-custom-website, status review. Companion to blog-2026-07-14-weve-outgrown-our-website.
• LinkedIn post (Problem / story): the 10-second signature cut. Draft at 2026-07-14-ten-second-signature-custom (content-queue/finals/); Workbench slug li-2026-07-14-ten-second-signature, status review.
• Still available to mine: the "30 Day Website" failure (its own LinkedIn story), and the sharper AI-invisible lines ("they'd FEEL it's worth less" + the $15K reframe) to finish the existing Not Leading with AI.md stub.