Every groomer has met the client who thinks a full groom is just a bath.
They look at the price, then at the dog, and ask what the rest of the money is for. From where they're standing it's a fair question. They can't see the two hours, the coat that was matted to the skin, the dog that wouldn't stand for its back feet, the second pair of hands, or the fifteen years of knowing which blade.
"I could build that in a weekend" is the same sentence, pointed the other way.
You know how the first one feels. Here's the work the second one can't see.
We sell booking software and make money if you choose ours, so weigh what follows with that in mind. The test further down works on us too.
It goes on the list with the power and the gates
Some things in your business you'd never choose on features. The gate between the salon floor and the front door. The switchboard. The hot water. The dryer, if you've ever had one die with a wet doodle on the table. (The kit-and-price version of this is You Didn't Buy the Cheapest Clippers.)
The questions you ask about a gate are whether it'll hold a dog that has decided to leave, every day, for years, and who comes out when the latch goes. And before you trust one with a room full of other people's dogs, you want to know it has held somewhere like yours for a long time.
Your booking system is on that list, payments and client messages included. It's infrastructure. When it's right you stop noticing it, the way you stop noticing the power. If it stops on a Saturday morning you should notice just as fast, and you won't, because a broken booking page is quiet. Nobody rings to say they couldn't book; they just book somewhere else.
So you wouldn't build your own gate. And you'd want to see where one has held before it holds yours.
What's on the other side of the counter
A booking page takes about an afternoon now. Describe it to an AI and there it is, on a real address, taking a pet and a service and sending a confirmation. That part is real.
Here's what the job looks like once real clients turn up. None of it shows in a demo.
- A client taps Book twice because the first tap looked like it did nothing. One groom, or two?
- Two clients want the last Saturday slot and tap in the same second.
- A groom booked in August for a Saturday in October, after the clocks go forward. Is it still at nine?
- A deposit taken in April against a groom in July. A card that declines at four in the morning and needs another go. A partial refund because the dog wouldn't let you near its nails. A chargeback with a deadline.
- Where the card numbers live, and the card industry's security rules that come with them if the answer is your own database.
- The reminder that actually arrives, from a number the client recognises, and the client who texts STOP and must never hear from any part of the system again, including the part you add next year.
- The groomer who left in February and is still logged in on the salon iPad.
- Who moved that booking, when, and from which login, when the client swears it wasn't them.
- The client who asks you to delete everything you hold on them.
- A backup somebody has actually restored from, to prove it works.
- The client who books with a screen reader, or with the text turned right up.
- Something watching all of it, so a job that has stopped tells you before a no-show does.
Getting working code out of a model is the first afternoon of that job. The rest runs for as long as your doors are open, because none of it stays done. Stripe retires a feature your checkout was built on. Apple changes what Safari does with a date field. From 1 October 2026 Australian businesses can't pass card fees on as a separate surcharge, so any checkout that adds one has to change before then (what the surcharge ban changes). Build your own and the reading and the code change are both yours. Nobody sends you a memo.
And there's the change you ask for yourself. You move the cancellation window from 24 hours to 48, and the model rewrites, with total confidence, the function that also decides who gets charged a fee, because the two were tangled together in a way nobody wrote down. There are no tests. You find out on the first cancellation, or the tenth.
Whoever is surest they could build it in a weekend is usually the person who hasn't yet had a client tap Book twice. It's hard to worry about a problem you haven't met.
All of it needs an owner, too. Software you depend on has to be somebody's job, and not their fifth priority, which is the polite way of saying nobody's. In a one-van business that's the groomer, who has six dogs booked today. In a business with three sites it's whoever got asked to "have a look at the website" between shifts.
Better at what?
If you're ever told somebody has built a better booking system in a weekend, or a month, you don't need a view on the person saying it.
A weekend is genuinely enough to build the demo, and a month is enough to build a very good one.
"Better" is a comparison, though, and a comparison needs a basis. Better at which line of that list? Measured how, and against which version of what? A demo can't answer any of that, because a demo is the happy path, and the happy path was always the easy part.
Then there's the calendar. Software gets good by being wrong in front of people who say so, over and over, for years. Something a month old hasn't had the chance to be wrong in front of anyone yet. That's arithmetic about time, and it applied to us as well: our first version went out to four businesses in Sydney's Eastern Suburbs in 2022, and plenty of what's right in Petboost now is right because one of them told us it was wrong.
The evidence worth asking for is boring on purpose: how many businesses run it, how long it has been running, who answers at seven on a Saturday, and what they shipped the last time the law changed.
What it looks like when it's right
Here's what a good day looks like at each end of the market.
One groomer, one van. A client books at nine at night from the link in your Instagram bio, and the only suburbs on offer are the ones you actually drive to. Three days out their card is checked, so a dead card turns up while there's still time to sort it. The reminder goes the day before, at the time you set. You're twenty minutes behind at the second house, so you send the running-late message from the next appointment and keep going. At the end of the groom the invoice is written, and half an hour later the card is charged and the receipt goes out. Their next one is already in the diary. You get home with nobody to chase.
Three sites, one owner. Each site runs its own calendar, team, services and clients, and you move between them from one login. The daycare yard fills at seven, and every dog in it was checked against its vaccination record when it was booked, so nobody has that conversation at the gate. The yard stops taking bookings at the number of dogs it holds safely. A card that fails its pre-authorisation flags itself on the appointment three days out, while the front desk can still fix it. Billing sits with the organisation, SMS credits and card volume included, and at the end of the week you can check every site's numbers without ringing anybody.
Nothing in either day is clever. You can count it, though: the bookings that came in overnight, the invoices nobody chased, the reminders nobody sent by hand and the no-shows that didn't happen.
The test, for whoever supplies it
Put these to anyone who wants to run your bookings. That includes us, and it includes you with a laptop and a free weekend. Somebody who has done the work can answer each one in a sentence.
Do they know the industry?
- Can it refuse a booking because a vaccination has lapsed? Ask to see it.
- Is capacity a person or a place? A groom needs a station and a stay needs a kennel.
- One owner, three dogs, one invoice. Show me.
Do they know what it is to run a business?
- What happens to my money when a card declines at four in the morning?
- Can I export my clients and pets today, without asking anyone?
- Who answers at seven on a Saturday, and have you put that in writing?
- Can I talk to two businesses that run it who aren't in your marketing?
Can they show the engineering?
- What happens when a client taps Book twice? Show me.
- When a scheduled job stops, who finds out, and how?
- Where do the card numbers live?
- When did you last restore from a backup?
- What did you ship the last time the law changed? Ask for the specific thing.
- How long has it been running, and for how many businesses?
Anyone can write the first two on a website, us included. The third only shows up in the answers. The pet-specific part has its own write-up, and the buyer's guide goes through the product criterion by criterion.
Our answers, since we sell the alternative
We sell a booking system, and we make money when a business chooses ours instead of building one. Here are our answers to the same test, and where to check them.
The industry. Petboost began at Hound Health Bondi, our co-founder Annika's daycare and hydrotherapy business, which opened in 2018 and is open today. Every business type since was designed inside real businesses: body-area grooming notes from established salons to boutique one-person studios, play-area capacity on daycare floors, kennel capacity and 24-hour billing inside boarding facilities, and GPS-verified check-ins with walkers. We're still inside Australian pet businesses every week.
Running a business. We've owned businesses as well as built software for them. Hound Health is one. Mo, our CTO, founded Call of Service, which Hipages acquired in 2019. Dan, who runs customer success, co-founded and sold RosterCloud, a timeclock and payroll startup. Our story has the rest. What we've put in writing about support is on the contact page: an Australian team in Australian time, a free 1800 line, live chat, and email answered within 24 hours in business hours. You can export your clients and pets as a CSV whenever you like, and our pricing is public, so put your own numbers through it rather than taking my word for it.
The engineering. Software is our first trade. A career designer and a career engineer build Petboost full time, and we're backed by the founder of TechnologyOne, Australia's largest enterprise software company. A second tap on Book while the first is still going is ignored. The job that sends reminders and takes payments runs every fifteen minutes and checks in with a monitor each time, so if it keeps failing or stops turning up, an alarm goes off on our side. Card details are never stored on our servers, because Stripe holds them. A client who texts STOP is opted out, whatever else the message said. And ahead of the surcharge ban, businesses on Petboost got a warning in their payment settings and a tool that builds the card fee into their prices.
Put the same questions to MoeGo, Gingr (whose owner, Togetherwork, bought PetExec in 2024) and Time To Pet. They're real products with real teams doing the unglamorous half, and none of them asks you to be the one fixing it at seven on a Saturday. If you're already on one and it works, the case for building your own is very thin. Where we'd argue for ourselves is the pet-shaped part, and being Australian and on the floor every week.
When building your own is the right call
Plenty of the time, and it's worth doing. The test I'd use is whether anybody other than you finds out when it breaks at two in the morning.
If nobody does, build it. A calculator for what a groom really costs once you count the drying time. A script that turns your roster into the shape your accountant keeps asking for. A one-off that reads a CSV and tells you which clients haven't been in since March. Something that renames four hundred photos properly. None of it touches a client or moves money, and if it falls over you fix it on Tuesday. That's where AI earns its keep in a pet business, and it's what our guides to AI in a pet business are about.
The line is clients and money. Once a stranger can reach it, or a dollar can move through it, it isn't a weekend project any more.
See how ours takes a booking · How to choose between them · Start a free trial
