The Bakery's Website Still Says “Closed for the Holidays”

There is a bakery near my old apartment with a website that still has a banner across the top: "Closed for the holidays — back January 2nd!" It has said that for at least three years. The phone number on the contact page belongs to a line they disconnected. The menu PDF opens to a 404. And yet the bakery is fine — the line is out the door on Saturday mornings, the owner knows half the neighborhood by name, and the croissants are excellent.

Nobody is going to fix that website. Not because the owner doesn't care, but because the person who built it moved away in 2021, the login is in an email account she no longer checks, and every agency that has cold-called her since quoted a full redesign for four thousand dollars. She doesn't want a redesign. She wants the holiday banner gone.

That gap — between "needs a rebuild" and "needs someone to spend twenty minutes on it" — is where a very unglamorous, very durable little business lives.

The idea: become the person who quietly keeps small local websites alive. Not a designer. Not an agency. A caretaker, on a small monthly fee.

Why this works right now

The obvious objection is that everyone has a website builder now, so who needs help? But that's exactly the problem. A decade of cheap templates means there are millions of small-business sites out there that were set up once, by someone, in some platform, and then abandoned. They didn't disappear. They aged. Plugins went out of date, SSL certificates lapsed, contact forms stopped delivering to an inbox anyone reads, and hours listed on the site drifted away from the hours on the door.

Meanwhile the cost of getting those details wrong went up. A customer who finds the wrong hours doesn't call to double-check; they pick the next result. A contact form that silently fails doesn't announce itself — the owner just concludes that "the website doesn't really bring us anything," which becomes a self-fulfilling prophecy.

The pricing reality supports the idea too. As of writing, industry guides put ordinary small-business website maintenance somewhere in the $50–$500 per month range, with most full-service plans for a simple site landing around $100–$200 a month. That's not a number that requires a hard sell. It's less than a small business spends on a single print ad, and unlike the ad, it compounds: every month the site stays accurate is a month the business isn't bleeding calls it never knew it lost.

And critically, the big agencies don't want this work. Their economics need project revenue. A shop that needs four hours a month is an annoyance to them and a perfectly good client for one person with a laptop.

What you're actually selling

This is the part people get wrong. You are not selling technical skill, because the technical skill required is genuinely modest — updating a plugin, swapping an image, editing a paragraph, checking that a form arrives. If you can navigate a content management system and read documentation calmly, you can do the work.

You are selling removal of a nagging item from someone's mental list. The bakery owner has thought about that banner maybe forty times. Each time it cost her thirty seconds of low-grade guilt and no action. Your product is that she stops thinking about it forever.

That reframing changes how you package it. Don't sell hours. Sell a small, named plan with a promise attached:

The report matters more than it should. A one-page monthly note — "site up 100%, 6 updates applied, hours corrected, 14 contact form submissions delivered" — is what stops the client from quietly wondering what they're paying for. It's also, conveniently, the cheapest thing on the list to produce.

How to start, concretely

Week one: pick a lane and pick a platform. Do not offer to maintain anything on any platform. Pick the one that dominates your area's small businesses — in most places that's WordPress, though local variations are real — and get genuinely fluent in it. A narrow specialist gets referred; a generalist gets compared on price.

Week two: build your audit. Write a checklist you run on any site in under fifteen minutes. Is it on HTTPS and is the certificate current? Does it load in under three seconds on a phone on cellular data? Are the listed hours correct? Does the contact form actually deliver — you test it yourself and see? Is the copyright year stale? Are there broken links or missing images? Is anything visibly out of date, like a holiday banner in March?

Week three: find ten businesses and audit them for free. Not a pitch — an audit. Walk in, buy something, and hand the owner a single printed page: here are four things wrong with your website, here's what each one costs you, here's what I'd charge to keep it from happening again. The conversion rate on this is far better than cold email, for a reason that has nothing to do with technology: you showed up, and you did the work before asking for money.

Free work that demonstrates a specific problem beats free work that demonstrates general competence.

Week four: land two clients and write everything down. Two clients at a hundred dollars a month is not a living. It is proof that the offer converts, which is the only thing you're testing at this stage. From there the growth is mostly referral — small business owners talk to each other constantly, and "she fixed my website and it's ninety bucks a month" is an easy sentence to repeat.

Keep a shared credentials document from day one, ideally in a password manager you control with client-specific access. The single most common way this business goes wrong is losing access to something and having to explain it.

The part nobody puts in the pitch

Here is the reality check, because the idea is good but not effortless.

Scope creep is the whole risk. "While you're in there, could you just—" is how a two-hour month becomes a nine-hour month at the same price. The defense is boring and it works: write down exactly what the plan covers, and when a request falls outside it, quote it cheerfully and in writing before you touch anything. Most clients are entirely reasonable about this. The ones who aren't will reveal themselves in month two, which is useful information.

The work is genuinely low-status and genuinely low-drama. You will not be designing anything beautiful. You'll be updating plugins and correcting phone numbers. If that bores you into quitting in month four, the business fails — not because it doesn't work, but because it requires the same unexciting attention every month, forever. That's also its moat.

Something will break on your watch. An update will take a site down for an hour. Have automated backups running before you accept a single client, know how to restore one, and practice the restore on a test site. The difference between a trust-destroying incident and a trust-building one is whether you noticed before the client did and said so first.

It scales awkwardly. Twenty clients is comfortable for one person. Fifty is not, and hiring changes the business into something else entirely — the margins get thinner and you become a manager. Decide early whether you want a solid one-person income or an agency, because the two require different choices at client number fifteen.

What this is really worth

Twenty clients at $150 a month is $36,000 a year for work that, once systematized, occupies maybe two focused days a week. That's not a venture-scale outcome and it was never going to be. It's a steady, boring, recurring income with near-zero startup cost, built on the least glamorous truth in small business: most problems aren't hard, they're just unowned.

The bakery's holiday banner isn't a technical challenge. It's a job nobody's name is on. Putting your name on a hundred small jobs like it, for a modest monthly fee, is a real business — and the kind you can start on a Tuesday with a laptop and a printed checklist.

If you try it, start with the shop you already buy coffee from. You already know the owner, you already know the site is out of date, and the conversation is easier than any cold email you'll ever write.