DEV Community

MapToSite
MapToSite

Posted on

One chair, four bookings a day: the interval math a one-person salon's booking page actually needs

A one-chair nail salon I built a sample page for opens 10:00 and closes 19:00, six days a week. Ninety minutes for a gel set including prep and cleanup, fifteen minutes to sweep and reset the table, and the last appointment has to be done before 19:00, not started before it. That is 540 minutes of open time and, on a good day, five starts. One late client, one walk-in, one chat at the door, and it is four.

So the shop is selling roughly four units a day. Everything the website does should be judged against that number, because it is the smallest budget a booking product will ever work with.

(This is the business I run, so discount accordingly. The sample page, the calculator and the order page are linked at the end.)

The arithmetic of one chair

Model the chair, not the calendar. The only quantity that matters is minutes of chair time, and every service consumes slightly more of it than its name suggests.

Service Chair minutes Buffer Total
Gel, one colour 90 15 105
Hand care 45 15 60
Gel removal 30 15 45
Pedicure 75 15 90

A full day of gel sets: 10:00, 11:45, 13:30, 15:15, 17:00. The 17:00 booking finishes working at 18:30, the buffer runs to 18:45, and the shop still closes on time. Five is the theoretical ceiling for one service type; mixed bookings usually land at four, sometimes three.

Four gel sets at 45,000 KRW each is 180,000 KRW a day. The interesting part is what a lost slot costs: not the price of the software, but a fifth of the day.

Where the money actually leaks

Two different problems get sold the same product.

No-shows. One no-show a week is four a month — 180,000 KRW at 45,000 a set, about 2.2M KRW a year. A $39/month scheduler with SMS reminders is $468 a year, roughly 630,000 KRW at 1,350 to the dollar, so it breaks even if it prevents about a third of those forty-eight appointments. If no-shows are the problem, that subscription is worth its price, and I would tell a shop with a bad no-show rate to buy it before buying a page.

"I couldn't see the prices, so I didn't call." This is the other problem, and a reminder system does nothing for it. A client deciding between two salons on a Saturday morning wants a price list, today's opening hours, and a way to ask one question without a phone call. That is a page, not a scheduler.

Most one-person shops I have talked to buy the first product to fix the second problem, then stop paying for it four months later. The order is usually backwards.

What the page has to hold

Three things, in this order:

  1. Prices as data. Not a JPEG of a price list. An array of records, each with a name, a duration and a price, rendered into the page. When a price changes, one line changes and the printed menu quotes the same numbers.
  2. Hours that answer "open now". A pure function of the current time and a weekly schedule, computed in the browser. If the shop is closed, the page says when it opens next, not "welcome".
  3. A request, not a reservation. A one-chair shop has no database to hold a lock, and nothing is worse than a page that says a slot is yours and then double-books it. A request form with date, time and a note, plus a clear line that the owner will confirm, is the honest version — and the owner confirms from the phone anyway.

The interval math, in about forty lines

The only genuinely fiddly part is producing start times that respect durations, buffers and the closing time. Minutes-since-midnight keeps it trivial.

const OPEN = 10 * 60, CLOSE = 19 * 60, BUFFER = 15, GRID = 15;
const SERVICES = {
  gel:  { min: 90, label: 'Gel, one colour' },
  hand: { min: 45, label: 'Hand care' },
  pedi: { min: 75, label: 'Pedicure' },
};

// a booking is [startMin, endMin] including its buffer
const clash = (a, b) => a[0] < b[1] && b[0] < a[1];

function startTimes(service, booked, horizon = 6) {
  const dur = SERVICES[service].min;
  const out = [];
  for (let t = OPEN; t + dur + BUFFER <= CLOSE && out.length < horizon; t += GRID) {
    const slot = [t, t + dur + BUFFER];
    if (!booked.some(b => clash(slot, b))) out.push(t);
  }
  return out.map(m => `${String(Math.floor(m / 60)).padStart(2, '0')}:${String(m % 60).padStart(2, '0')}`);
}
Enter fullscreen mode Exit fullscreen mode

Three rules that keep this from producing nonsense:

  • The buffer belongs to the booking, not the service. Cleaning after a pedicure is not part of "pedicure" in the price list, but it is part of the day. Put it in the interval and the last slot of the day stops looking free when it isn't.
  • The last start is CLOSE - duration - buffer, not CLOSE - duration. This single line is where most hand-written booking pages let the shop run over.
  • The grid is a shop decision. Fifteen minutes suits nails; ten suits a barber. It belongs in one constant, next to the opening hours, so the owner's answer to "can you do 10:20?" has a single place to live.

The function above is deliberately dumb: it never looks at the past, never merges adjacent services, and never guesses. Its job is to answer "what could start today" in a file that has no backend at all — and a static page can afford that because the owner confirms each request by hand.

What I left out, on purpose

  • Calendar sync. Two-way sync needs OAuth per shop, a token refresh path, and a support burden the moment a client edits an event. For four bookings a day, the owner reads the request and writes it in the paper diary.
  • Deposits. Taking a deposit means a payment provider, refunds, and a policy that has to be written down. That is a different product with a different price.
  • SMS reminders. They work, and they cost money per message. If the shop wants them, they belong with the scheduler subscription, not with the page.
  • A login for the owner. The whole point is that the shop never makes an account. The page is published with an edit link; whoever holds the link can change prices and hours, and nothing else.

Three years, laid out

The table everyone should see before signing up for anything. Assumptions: one chair, four bookings a day, no staff, prices in USD for the software and KRW for the shop's own sales.

Option Cost over 36 months What it includes
Scheduling SaaS at $39/mo $1,404 Calendar, reminders, deposits, more settings than a single chair uses
Squarespace at $19/mo $684 Site plus scheduling, priced per month forever
Wix at $17/mo $612 Same shape, different defaults
Static page + printed kit, one payment $49 Page, price list, request form, 15 printed pieces
Hand-built, no vendor $0 Your evenings, and a page nobody updates when the price changes

The subscription column is not wrong — it is a different job. The question is which job the shop actually has. If the answer is "people can't see what I charge and when I'm open", paying monthly for a calendar is paying for the part that isn't broken.

The Korean version of this kit is 39,000 KRW once for the setup, with three months of the basic plan included, then 4,900 KRW a month if the shop wants to keep it hosted — no contract, no auto-charge, and the three-year total lands at 200,700 KRW, printed on the quote so nobody has to multiply.

See it rather than read it

  • The one-chair sample page all of the above was modelled on, with the price list, the open-now badge and the request form: nail room sample
  • The three-year cost calculator, with the assumptions on screen: cost calculator
  • The kit itself, $49 / $89 / $149, one payment, delivered within 24 hours: order the kit

If you run a one-person shop and disagree with the four-a-day arithmetic — or you have a better answer for "can you do 10:20?" — I would like to hear it. That question is the one that decides whether a request form or a real calendar is the right build next.

Top comments (0)