On-demand drivers, no payroll: how Pizza Prince handles delivery without employing a single driver

Running delivery from a restaurant sounds straightforward until you actually try to do it yourself.

You need drivers. Which means recruitment, contracts, insurance, vehicles or mileage allowances, and a rota that somehow lines up with order demand that is impossible to predict hour by hour. On a quiet Tuesday night you have two drivers sitting in your car park eating your margins. On a busy Friday you have none available and your customers are waiting forty-five minutes.

Most independent operators who have been through this eventually land on one of two conclusions. Either they hand everything to Just Eat or Deliveroo and accept the commission hit, or they quietly stop offering delivery altogether.

Godwin Irie of Pizza Prince in Temple Bar, Dublin, found a third option.

"As soon as a customer orders, it goes directly to that company and they send us a driver," he told us on the Flipdish Takeaways podcast. "I don't have my own drivers. As soon as a customer orders, a driver comes in."

No payroll. No idle time. No vehicles on your books. Just delivery that shows up when you need it.

How the on-demand delivery model actually works

The mechanics are simpler than most operators expect.

When a customer places an order through Pizza Prince's online ordering system, two things happen simultaneously. The order lands on the kitchen system, exactly as it would for any collection or walk-in. At the same moment, it is automatically forwarded to a last-mile delivery partner, who dispatches a driver to collect and fulfil it.

There is no manual step in the middle. No one calling a dispatch number. No manager watching a board and trying to work out when to call a driver. The integration handles it.

Flipdish connects with a range of last-mile delivery partners including Stuart, Nash, Ondway and others, all of which work on the same on-demand principle. The restaurant defines its delivery area on a map, sets its delivery parameters, and from there the system runs the process automatically.

"It works with Flipdish, the till, so yeah," Godwin said. "You have everything in one place."

That last part matters more than it might first appear.

Why integration into a single system changes everything

The operational headache with delivery, historically, has not just been the drivers. It has been the fragmentation.

Orders coming in through three or four different channels. Tablets from Deliveroo, Just Eat and Uber Eats all pinging simultaneously. A separate system for your own website orders. A driver dispatch process that sits outside all of it. Every one of those is a point where things can go wrong, where orders get missed, where timing breaks down, where the kitchen loses track of what is going where.

When delivery dispatch is integrated into the same POS system that handles every other order, that fragmentation disappears. The kitchen sees one flow of orders. The delivery goes out automatically. No one is manually copying an address into a separate app.

For a small team running a busy service, that is not a small thing.

Godwin has four chefs, an operations manager and himself. That is a lean operation for a site doing the kind of volume Pizza Prince handles in Temple Bar, where 70% of orders are walk-ins and they are open late to catch the post-pub crowd. Adding delivery coordination on top of that, with a separate driver fleet to manage, would put real pressure on a team that size. The on-demand model removes that pressure entirely.

The honest economics

There is a cost to on-demand delivery, and it is worth being clear about what it looks like.

Third-party last-mile partners typically charge a per-delivery fee, which varies depending on distance, provider and volume. That fee comes out of the delivery order margin. Unlike aggregator commission, which sits somewhere between 25% and 35% of the order value, the per-delivery fee on a last-mile integration is a flat amount rather than a percentage, which tends to work significantly better as order values increase.

A £25 order on Deliveroo at 30% commission costs you £7.50 in fees. The same £25 order delivered through your own online ordering system with a last-mile partner at a flat delivery fee might cost you £3 to £4. You also keep the customer data, which you do not get from any aggregator.

As Flipdish's own delivery ebook puts it, the on-demand model lets you take back control of your customer experience in a way that drives revenue rather than giving it away.

None of this means aggregators have no role. Godwin uses Just Eat, Uber Eats and Deliveroo as well, and is pragmatic about it. "They have their percentage. Everyone has to eat at the end of the day." The point is not that third-party marketplaces are wrong. It is that relying on them exclusively means your delivery economics are always set by someone else.

What you are not paying for

Perhaps the clearest way to understand the value of the on-demand model is to look at what it removes from your cost base.

No driver recruitment. Finding reliable delivery drivers is genuinely difficult, particularly since the pandemic reshaped attitudes to hospitality work. The time and energy that goes into hiring, training and managing a delivery team is real. The on-demand model outsources all of it.

No idle driver cost. A salaried or hourly driver who is not delivering is still costing you money. With on-demand, you pay per delivery. When there are no orders, there is no cost. The economics follow the demand.

No vehicle overheads. Insurance, fuel reimbursement, maintenance, the administrative load of running a vehicle fleet even at small scale. None of that applies.

No rostering complexity. Trying to staff a delivery rota around unpredictable order patterns is one of the harder operational puzzles in food delivery. On-demand removes the puzzle entirely.

For an independent operator like Godwin, who is also planning his second site and doing the work of managing a kitchen, a front of house team and a growing brand simultaneously, removing that entire category of complexity has a value that does not always show up neatly on a spreadsheet.

The trade-offs worth knowing about

Fair coverage means being honest about where the model has limits.

You are dependent on your last-mile partner's driver availability. In a very busy period, particularly if multiple restaurants in your area are using the same provider, there can be a wait for a driver. That is manageable, but it is a real variable.

You also have less direct control over the delivery experience than you would with your own drivers. A team member you have trained knows how to hand over an order, represents your brand and understands how you want the customer interaction to go. A third-party driver may not.

Some operators address this through packaging. A well-branded bag, a handwritten note, a QR code pointing to your loyalty programme. The delivery moment can still feel like yours even when the driver is not. But it requires thinking about deliberately rather than happening by default.

And finally, the flat per-delivery fee model works better at higher order values. For a pizza operation where average order values are healthy, the maths is comfortable. For a lower ticket product, the per-delivery fee takes a proportionally larger bite.

Why this matters for operators thinking about their second site

Godwin made a point during the podcast that connects the delivery model to something bigger.

When he opens his second location, the system he has already built will replicate cleanly. The Flipdish POS, the online ordering setup, the last-mile delivery integration. He will not need to build a separate delivery operation for the new site. He will just connect it to the same infrastructure.

That scalability is one of the underappreciated advantages of the integrated model. A delivery operation built on a patchwork of separate systems does not scale well. Every new site multiplies the complexity. A delivery operation built inside a single platform scales as simply as adding a new location to your account.

"It requires just having a proper system in one place," Godwin said. "Then you can duplicate it."

That is a simple sentence, but it describes something that a lot of growing operators get wrong. The delivery model you choose when you have one site should be one you could still live with at five.

What this looks like for an operator considering it

If you are running a restaurant or takeaway and currently either not offering delivery or handling it entirely through marketplaces, the on-demand last-mile model is worth a serious look.

The starting point is your online ordering setup. If your ordering system already integrates with last-mile delivery partners, the technical piece is straightforward. Define your delivery area, connect to a partner, test a few orders. The infrastructure already exists.

If you are not yet on a system that supports this, the Flipdish delivery page outlines the partner network and how the integration works in practice, including partners that cover Dublin, London and other key markets across the UK and Ireland.

The broader argument Godwin makes is not just about delivery logistics. It is about keeping complexity out of your business so you can focus on the thing that actually brings people back.

For Pizza Prince, that thing is the pizza. The Roman-style slices made with Italian ingredients, the pizzamisù, the late-night walk-in trade from Temple Bar. The delivery just has to work without getting in the way.

The on-demand model, as it turns out, is very good at not getting in the way.

Listen to the full episode

This conversation is from Season 3 of the Flipdish Takeaways podcast. Godwin covers everything from pivoting away from Neapolitan pizza after six months, to the invention of the pizzamisù, to his advice for anyone thinking about opening their own restaurant.

Listen to the Flipdish Takeaways podcast