All articles
  • consulting
  • case-study

Case study: a one-page website for a golden-era hi-fi shop

A semi-retired vintage audio enthusiast in Erwin, TN needed exactly one thing from the internet: a way for people to find him and set up a visit. Here's what we built, what we deliberately left out, and why small is a feature.

Carter Sheppard3 min read

Most of what I write about is AI systems. This one is about a website — the smallest thing we've shipped — and I'm writing it up because it makes a point about consulting that the bigger projects hide: the job is fitting the tool to the business, and sometimes the right tool is deliberately tiny.

The client

Retro Audio Revival is a vintage hi-fi shop in Erwin, Tennessee, run by a semi-retired enthusiast who believes — correctly, if you ask him — that the equipment built between 1968 and 1979 sounds better than anything you can buy new. The business runs a simple loop: buy quality golden-era gear (Marantz, Sansui, Pioneer, Kenwood), restore it in-house — recap, alignment, lamp replacement — then modernize it with Bluetooth and streaming inputs and sell finished systems out of a showroom that's open by appointment only.

That last part matters. This is a passion that became a side business, run on the owner's time. He doesn't want walk-in traffic, a booking system pinging his phone, or an e-commerce cart. He wants the right people to find him, understand what he does, and reach out to set up a visit.

What he actually needed

When you say "the business needs a website," the default path in 2026 is a site-builder subscription: a monthly fee forever, a template that looks like everyone else's, and forty features he'd never touch wrapped around the three he needs. The other default — a custom agency build — has historically been priced for businesses much bigger than a one-man shop.

The actual requirements list was short:

  • Tell the story, because the story is the sales pitch — this is a person, not a chain.
  • Show how the business works: buy, restore and integrate, sell. One loop.
  • Make "showroom by appointment" impossible to miss, and make contact effortless — call, email, or Instagram DM, whichever the customer already uses.
  • Look like the product. If the site feels like a SaaS landing page, it's lying about what's inside.

Notice what's not on the list: no CMS, no inventory database, no scheduling software, no login. Every one of those would have added cost, maintenance, and a thing to break, while moving him further from how he actually operates. The appointment "system" is a phone number, and for this business that's not a limitation — it's the correct architecture.

What we built

One page. A hero with the shop's name in a warm amber dial-glow over a woodgrain texture — the design pulls straight from the walnut cabinets and backlit dials of the equipment itself, down to the typography. The story section, in his voice. The buy → integrate → sell cycle drawn as exactly that, a cycle. A strip of the four brands he lives for. And a contact section with three tap-to-act cards — phone, email, Instagram — under a headline that does the qualifying for him: looking to buy, looking to sell, or want to visit the showroom.

It's fast because there's nothing to be slow. It works on a phone because that's where his customers are. And it's fully custom — built with the same AI-accelerated development approach we use on much larger software projects, which is the economic unlock here. AI compresses the build timeline enough that a fully bespoke site becomes viable for a business this size, where it used to be agency-budget territory. He owns the result outright. No subscription meter running.

The part that keeps paying

Because it's custom code and not a template, the site can grow in whatever direction the business does — an inventory page when he wants one, a restoration gallery, nothing at all if the phone keeps ringing. The v1 shipped small on purpose; the flexibility was the point of doing it custom.

That's the consulting lesson I'd generalize. Fit is worth more than features. A tool that matches how the operator actually works — appointment-only, personal, no ceremony — beats a more capable tool that fights him. It's the same principle we apply when we put AI automations inside a business: start from the workflow that exists, build the smallest durable thing that serves it, and leave room to compound.

If you run a business in the region and the web presence question has been sitting on your list for years — it's probably a smaller project than you think. Tell us what you're working on.

ShareLinkedInX

Article Alerts.

Writing on AI use cases that actually ship, and on buying and operating regional businesses. No cadence promises — just sharing what I'm seeing in the worlds of AI, regional businesses, and broader market and venture trends.

Keep reading

Something here land? Tell me what you're working on.

Based in the Smoky Mountain corridor · Working with teams everywhere