Two very different businesses, asked the same questions.
A bounce house company and a private chef have almost nothing in common. That is the point of showing both. Read the left column as what a platform can honestly promise, and the two on the right as what has already been written for a business that needed something a template had never heard of.
A note on the third column. "David's" is a stand in name. It is a real restaurant, a real signed client, and a real build in progress, but which restaurant is the owner's to tell people, not mine to put in a sales document. Everything described in that column is genuinely what is being built. You would get the same courtesy: I ask before your name appears anywhere, and no is a complete answer.
Availability
A shopping cart does not know that a unit is out on the 14th, or that the pork ran out at seven. Booking apps bolt a calendar on top, and the calendar and the cart keep separate records, so the last word belongs to whichever one the customer refreshed most recently.
A double booking is refused in the database itself, by a rule Postgres enforces before the row is allowed to exist. The calendar greys out taken dates as a courtesy. The guarantee sits one layer below that, where a stale browser tab cannot reach it.
A kitchen does not sell dates, it sells tonight. So ordering is a switch he flips when the plancha is hot, tonight's location is his to set, and a dish goes to sold out the moment the last one leaves, from the same phone he is cooking with.
The one rule only your business has
Off the shelf inventory counts one product at a time. The rule that is actually specific to how you work is, almost by definition, outside the model.
There is exactly one misting fan, and two inflatables can go out the same day. A second rule blocks the fan on any overlapping booking, whichever unit asked for it. Written in an afternoon, because the rule was ours to write.
One cook, one plancha, one location a night. So the menu is not a catalog, it is tonight's menu, and everything on it can be turned off in a tap. The software was shaped around one man's Saturday, not the other way round.
What the money costs to collect
Checkout runs on the processor the platform supports, typically around 2.9 percent plus 30 cents, and the cheaper plan tiers add a cut of their own on top. You cannot shop it for a better rate, because the store and the checkout are sold to you as one thing.
Customers are steered to Zelle, which is free, with Venmo business at 1.9 percent plus $0.10 as the fallback. On a $267.50 rental that is roughly $8 a booking that stays in the account. A hundred bookings a year is about $800, which is most of what the site costs to run.
The processor is chosen to fit the business rather than the reverse. If he already takes Square in person, the site takes Square, so it is one set of books and one deposit instead of two systems to reconcile at midnight.
Sales tax
One tax rate for the whole store, applied wherever the order came from.
Georgia sources rental tax to the delivery address, and Braselton straddles four counties that do not all charge the same rate. The rate comes off the customer's ZIP, and every order stores the rate it was actually charged, so a change next year never rewrites last year's receipt.
Prepared food sold at one location is not taxed like equipment delivered to a driveway, so none of that machinery is here. Which is the whole argument in one cell: your rules get written, not inherited.
The record that protects you
A separate waiver or contract service, a separate login, a separate monthly bill. The signature lands in that company's dashboard rather than beside the order it belongs to, and reconciling the two is your job.
The liability waiver is signed on the phone inside checkout. The signing time comes from the server clock, not the phone. Each clause is stored with the exact wording shown. Once signed it is locked against editing by anyone, including the owner. That is what turns a typed name into evidence.
A catering job is where a kitchen gets hurt, so every request moves through pending, quoted, confirmed, declined, with the quote attached to the request. An office lunch for forty is never a text thread that two people remember differently.
The mistake it will not let you make
A cart will cheerfully take an order you cannot fill. Nothing in a template knows what a wasted trip, a late delivery, or an unpaid print costs you, because nothing in a template knows your business.
A wet setup is blocked until the customer confirms an outdoor spigot within 100 feet, because a water slide with no water is a truck run for nothing. Printed merch cannot reach the printer until it is marked paid, so the business is never out of pocket for something nobody bought, and the lead time is guarded in four separate places, because a banner arriving after the party is the one unacceptable outcome.
Ordering closes when he closes it, and a sold out dish leaves the menu the same second, so nobody pays for food that is already gone. The refund that never has to happen is the cheapest one there is.
Your side of the screen
You work out of the platform's generic order list, on a laptop, and fill the gaps with a notebook and your memory.
Pending and upcoming, approve or decline, blocked dates, deposit and balance tracked separately, a payment memo code per booking, and setup photos taken from the driveway on delivery day, stored where only the business can see them.
A kitchen rail, not an admin panel. Live orders move start cooking, mark ready, hand off, with completed orders and a sales view behind them. It is built to be read at arm's length with flour on your hands.
Language
A translate plugin will do the marketing pages. It stops at the checkout, mangles a menu, and never touches the staff dashboard, which is the one screen your kitchen actually lives in.
English only, because that is what the business needed. Worth saying out loud: you pay for the features your business has, not for a checklist someone else's business needed.
Bilingual on both sides, by design rather than by translation. Customers order in English or Spanish. The kitchen board runs in Spanish while the site is maintained in English, because the man cooking and the man maintaining do not share a first language. Decided on day one, since building it in costs nothing and retrofitting it is paid rework.
Marketing that runs itself
A newsletter block and a social feed embed. Both of them wait for you to have the energy to post.
A free invitation designer with every booking. It draws the party invitation in the customer's own browser, and every invitation carries a small credit line. A parent texts it to twenty guests, and twenty families see who supplied the bounce house.
A notify list, because the hard part of a pop up is telling people it is happening. One tap when ordering opens and the people who already love the food know the plancha is hot. No posting, no boosting, no ad spend.
Everyday changes same answer for both
Yours to make, every time, forever. Every price, photo, and package edit becomes your evening, and the site quietly drifts out of date once the novelty wears off.
The things that change hour to hour are yours, in a tap: ordering open or closed, tonight's location, a dish sold out, a date blocked. Everything slower than that you simply send to me. Prices, photos, packages, menu wording, service area, and any bug that turns up are included in the monthly at no charge. No ticket number, no upsell. The split is deliberate: you should never wait on me for something happening tonight, and you should never spend a Sunday on something that is my job.
When the ground moves under you same answer for both
Fees rise in an email. The editor gets redesigned the week of your busy season. A feature is deprecated. The app developer is acquired and sunsets the thing your bookings run through. You get no vote, and usually no notice.
Nothing changes unless you ask for it. When something underneath genuinely has to change, a payment provider reworking their system or a security update that cannot wait, that is my job and it is already inside the monthly. You find out because everything kept working, not because you got an email telling you to migrate by the 30th.
When something breaks same answer for both
A support queue staffed by someone who has never heard of your business and cannot see your data. They will suggest you contact the app developer. The app developer will suggest you contact the platform.
The direct number of the person who wrote it. Same or next business day as the standard, and faster than that when a weekend of bookings or a cook night is on the line. One person, one system, nobody to hand you off to.
The failures you would never think to test for are handled before they happen. Confirmation emails leave from a verified domain with DKIM, SPF, and MX set up properly, which is the entire difference between the inbox and the spam folder, and the invitation mailer is locked so it can only ever email the person signed in, because an open endpoint sending from your domain is a spam relay waiting to be found, and a burned domain takes every confirmation email down with it. On a platform, when your confirmations start landing in spam, there is nobody to call.
The idea you have next year same answer for both
"The app does not support that." Your options are to wait for a company that has never met you, or to change how you work so it fits the software.
Gift cards, a second location, text alerts on delivery day, a loyalty punch card, a standing weekly order, a route sheet for the driver. Quoted at $95 per hour, always priced before any work starts, so you decide with the number in front of you.
What you own at the end same answer for both
Your content lives inside somebody else's template engine, and your orders and signatures live inside two other companies. Leaving means rebuilding from scratch.
Your domain, your database, your customer list, exportable any day you ask for it. The site is a file, the data is a table, and both of them are yours. Staying is a choice you keep making, not a wall you are standing behind.
Cheaper is the part people get wrong.
The platform sticker price is real. It is also the smallest line on the bill, because a store plan alone cannot take a dated booking, hold a signature, or run a kitchen rail in two languages. Here is what the two paths cost once the site does the job you hired it to do.
Path one: platform, plus apps
| Store plan, the tier required to sell | $36 |
| Booking or ordering app that handles real availability | $50 to $100 |
| Digital waiver or contract service | $30 |
| Forms, email, or automation add on | $15 |
| Per month | $131 to $181 |
Across four bills, four logins, and four support desks that each blame the other three. Then the card fees on top, on every order, at a rate you do not get to negotiate. Then your own time: roughly 40 hours to build it, and a couple of hours most weeks to keep it honest. That time never appears on an invoice. You pay it anyway.
Path two: built for you
| Care and hosting, everything below included | $99 to $199 |
| Hosting, database, logins, email, daily backups, SSL, uptime monitoring | included |
| Price, photo, menu, and package changes, plus any bug fix | included |
| A direct line to the person who built it | included |
| Per month | $99 to $199 |
One bill, one login, one phone number, and the cheapest way to take a payment that your business can actually use. The build is a one time cost, half to start and half at launch, and your exact figure is on your proposal. The monthly is month to month, cancel with 30 days notice, because a client who stays by choice is the only kind worth having.
Platform figures are typical published prices in mid 2026 and are worth checking on the day you compare them. The exact numbers are not the argument. The argument is that four subscriptions which do not talk to each other cost more than one system that does, and they still leave the reconciling to you.
The work you will never see, and would never buy on purpose.
Nobody shops for these. They are the four places where a website silently loses you money or trust, and they are the clearest picture of what you are actually paying for.
How fast it loads
Grace & Joy is one file of plain HTML, CSS, and JavaScript. No framework, nothing to download before the calendar appears. A platform page ships a large bundle of scripts before it renders a single word.
The person deciding whether to book you is in a parking lot on one bar of signal, with your competitor in the next tab. Speed is not a technical detail. It is whether she waits.
Your customers' privacy
Setup photos are pictures of where a family lives. So that storage is private, every view is a link that expires in an hour, location data is stripped when the photo is resized, and a photo can only be deleted inside 24 hours: long enough to bin a blurry shot, short enough that a record cannot be scrubbed after an incident.
A platform's file upload is a public link by default. Parents notice how you treat pictures of their house.
Room to grow
Printed merch, the invitation designer, setup photos, and an upgraded waiver record were all added to a site that was already live and taking bookings. No downtime, nothing redone, no plan upgrade. On a platform, growth usually means a new app, a higher tier, or starting over.
Version two is not a migration. It is a Tuesday.
Paperwork a lawyer can read
The liability waiver exists as a clean reviewer's copy with numbered questions written for an attorney, and none of those internal notes are ever served to a customer. A platform hands you a generic template and no path to getting it checked.
Your liability document is a document, not a checkbox somebody else wrote.
When Squarespace is the right answer.
- You need five good looking pages, a photo gallery, your hours, and a contact form.
- Nothing on the site has to check a calendar, hold a signature, price a delivery, close ordering at nine, or tell your crew where to be on Saturday.
- You would rather pay less and do the work yourself, and you know yourself well enough to know that you actually will.
If that is you, go use Squarespace, and I will help you pick the template. It is a good product and it is the right call more often than people in my line of work admit. What I build is for the other case: when the website is not describing the business, it is running it.
Not a portfolio. Two real builds.
Grace & Joy Party Rentals
Live / Braselton and Jackson County, GeorgiaA customer picks dates on a live calendar, sees the weather forecast for those dates, chooses grass or driveway, adds a wet setup only on the unit that supports one, confirms there is an outdoor spigot within 100 feet, signs the liability waiver on their phone, and submits a request held as pending until it is approved. Sales tax comes off the delivery ZIP. Five branded emails go out along the way, and every booking gets a payment memo code so money can be matched to a name.
The demo runs the booking system under a made up brand, Sunnyside Party Rentals, so you can take it apart without dropping a fake order onto a real company's calendar. Pick your dates, choose grass or driveway, add the wet setup, read the weather for the day you picked, and sign the waiver. The second link is the live business, taking real bookings today.
David's, arepas and catering
In build for a real restaurant client / name withheldTwo doors from one system: tonight's ordering on one side, catering enquiries on the other. He opens ordering when the plancha is hot, sets tonight's location, and flips a dish to sold out as it runs out. Orders arrive on a kitchen rail he moves from cooking to ready to handed off. Catering requests hold their own pipeline with the quote attached. Every screen on both sides works in English and in Spanish, so he runs his business in his language while I maintain it in mine.
On the name. "David's" is a stand in. The client is a real restaurant currently in build, and the work below is real, but which restaurant is theirs to announce and not mine to advertise. Your name would get the same protection.
Do this one thing: order something in English, then hit the EN and ES switch and order it again in Spanish. Same screens, same system, both languages, and the kitchen side works the same way. That is the part no platform will sell you.
Three steps, and you can stop after the first one.
A conversation
Free, and not a pitch. Twenty minutes on what your week actually looks like and where the website is failing you. The first thing I will tell you is whether you even need me.
A proposal in writing
A fixed build price, a fixed monthly, and a plain list of what is included and what is not. No hourly surprises and no starting from. If the answer is no, we shake hands and you keep the proposal.
The build
Half to start, half at launch. The monthly begins the day it goes live, month to month, cancel with 30 days notice. Anything new later is $95 an hour, always quoted before a minute of work happens.
You will never get an invoice you did not see coming. That is the entire policy, and it is the reason the referrals keep arriving.
If your website has to do the job and not just describe it, call me. If it does not, I will tell you that too.
Matt Lacey
The Luke Group, LLC
matt.lacey1207@gmail.com
Reference build, live: graceandjoypartyrentals.com