◆ Ghost kitchens and virtual brands

Four brands, one line, and software that stops merging them. Zayos.

Quick answer

Ghost kitchen ordering software has to keep brands separate where diners look and joined where the kitchen works. Separate domain, look, customer list and reporting per brand. One ticket queue on one tablet. Zayos Concierge covers up to 5 brands per kitchen at $699 per location.

What is a virtual brand actually for?

A virtual brand exists to sell a second, third or fourth kind of food out of equipment, rent and labor you are already paying for. The hood runs whether you sell one concept or four. If a second menu can be cooked by the same people from mostly the same prep, its revenue arrives against fixed costs that are already covered, and that is the whole economic argument.

The second argument is discovery. A diner searching for one specific thing does not scroll a long menu looking for it, and a focused brand meets that search directly. Which is why concepts get narrow: one category, cooked well, named plainly.

Both arguments hold only while the brands stay believable. A diner who can see that four brands are one kitchen stops treating them as four choices, and that believability is a software job before it is a marketing one. It is where most stacks fail.

The recurring failure

Where does normal restaurant software collapse your brands into one?

Almost every restaurant platform is built on one assumption: one brand per kitchen. Nothing in the marketing says so, and everything in the data model does. These are the eight places it surfaces, in roughly the order operators run into them.

One customer list

A diner who ordered from your wing brand starts receiving email about your poke brand. The illusion breaks in one send, and the unsubscribe you earn is on both brands at once. Customer records have to belong to a brand, not to the kitchen.

One sender on the receipt

The order confirmation arrives from the parent company name, or from a domain the diner has never seen. This is the most common leak because transactional email is configured once and forgotten. Every brand needs its own name on the receipt.

One domain

Brands stuck on paths under a single site read as departments of one restaurant, and they compete with each other in search instead of ranking separately. Each brand should have its own host.

One theme

If the checkout is the same color for every brand, you are running one restaurant with four menus. Colors, logo and imagery are the cheapest part of the separation and the first thing a diner notices.

One number

Aggregate reporting is how a losing brand survives for a year inside a healthy total. You need revenue, orders, average ticket and refund rate per brand before you can decide what to kill.

One staff scope

The moment a brand has a partner, a licensee or a chef with a revenue share, everybody logging in can see everybody. Staff access has to be scoped to the brands a person actually works on.

One pause switch

At 9pm the line is buried. You want to mute the slowest brand for thirty minutes without taking the others offline. Software that only knows the kitchen, not the brands inside it, can only turn everything off.

One exit

Virtual brands get sold, licensed or shut down. If the customer list, the order history and the domain cannot leave with a single brand, the brand is not really an asset. Ask this before you sign, not during the sale.

How should tickets work when one line cooks four brands?

Separation belongs on the diner side. On the kitchen side, separation is the enemy. A cook watching four tablets will miss tickets on at least one of them, and the brand that gets missed is whichever one is quietest, which is usually the newest brand you are trying to prove out.

So the requirement is one queue, with the brand written on the ticket. Ask to see a real ticket in the demo: is the brand name at the top where a cook glances, is it legible on a tablet at arm length, and does the same label reach whatever goes on the bag. The handoff is where brand confusion becomes a refund, when a driver asks for an order under a brand your evening staff has never heard of and the bag on the shelf says something else.

Stock is the same story from the other direction. When something runs out it runs out for the kitchen, so the 86 has to reach every brand and every channel that sells it, immediately. A brand still selling an item the walk in does not have is how a good night turns into four refunds and two one star reviews.

What should you ask before you sign?

Eight questions. The pricing one and the exit one decide more than the feature list, because this format runs on launching brands and retiring them.

  1. 01 Does adding a brand cost extra, and what is the cost at the fifth one?
  2. 02 Does each brand get its own domain, or a path under yours?
  3. 03 Which brand name appears on the order confirmation the diner receives?
  4. 04 Can I pull revenue, orders and refund rate for one brand alone?
  5. 05 Can a staff member be given access to one brand and not the others?
  6. 06 Can I pause a single brand during a rush without pausing the kitchen?
  7. 07 If I sell or shut down a brand, what leaves with it, and in what format?
  8. 08 When the diner arrives or the driver walks in, what tells them which brand this bag is?

A platform that charges per brand is charging you for testing, which is the one activity that makes this model work. A platform that cannot release a brand cleanly is charging you at the end instead. Price the fifth brand and the closing of a brand before you price the first one.

How Zayos does it

How does Zayos keep brands separate in one kitchen?

A brand is a first class record, not a menu section. Each one carries its own public host, its own primary and accent colors and its own logo, and those colors are read live so they re skin both the storefront the diner sees and the operator screens your team works in. Customers and orders belong to the brand that produced them.

Staff access follows the same line. A membership can manage a specific set of brands, so a chef partner on one concept sees that concept, and the brand roster they open lists only what they are entitled to. Reporting is readable per brand, which is what lets you retire the concept that is not working instead of carrying it inside a healthy kitchen total.

The kitchen stays joined. Direct orders and marketplace orders from Uber Eats, DoorDash and Grubhub arrive through Otter into the same tablet, so the line works one queue. The 86 grid is built for that tablet during a rush, with large targets and a default 30 minute cooldown so an item comes back on its own rather than staying dark because nobody remembered it.

What it costs to run several brands

Concierge is $699 per month per location and covers up to 5 virtual brands per kitchen, so the fourth and fifth brand do not add a subscription. Operator is $499 and Operator plus Marketplace is $599, all per location per month, month to month, no setup fee. There is no percentage of sales. On direct orders the diner pays $0.99 pickup, $2.99 delivery, nothing on dine in and 10% on catering, and every brand keeps 100% of its food revenue and tips. Compare it against what a marketplace keeps on the pricing page and the commission calculator.

If you want the vertical view rather than the software evaluation, read Zayos for ghost kitchens. If you are adding a catering line to the same kitchen, catering ordering covers that flow.

Questions from multi brand operators

What is ghost kitchen ordering software supposed to do differently?
Keep several brands separate on the outside while keeping one operation on the inside. Every brand needs its own domain, look, customer list, receipts and numbers, while the kitchen needs one queue, one stock picture and one set of hours. Ordinary restaurant software is built on the assumption of one brand per kitchen, so it merges the outside layer, which is the layer diners see.
Do I really need a direct ordering site for a delivery only brand?
It is the only place a delivery only brand has any margin. A ghost kitchen has no dining room revenue to absorb marketplace commission, which runs 15% to 30% before extras and 25% to 35% blended once delivery, advertising and processing are included. A $40 order on a marketplace can land near $28 before the burner is lit. The same order on your own site keeps 100% of the food revenue, with a diner paid fee of $0.99 pickup or $2.99 delivery.
How many virtual brands can run on Zayos from one kitchen?
Up to five on the Concierge plan, which is $699 per month per location. Billing follows the kitchen, not the brand count, so the fifth brand does not add a subscription. Operator is $499 and Operator plus Marketplace is $599, all per location per month, month to month with no setup fee.
How does Zayos keep two brands in one kitchen from bleeding into each other?
Each brand is its own record with its own public host, its own colors and logo which cascade through both the operator screens and the storefront, and its own customer and order data. Staff access is scoped to the brands a membership can manage, so a partner on one brand does not see the others. Reporting can be read per brand rather than only as a kitchen total.
Does the kitchen have to watch a screen per brand?
No, and that is the point of the split. Direct orders and marketplace orders from Uber Eats, DoorDash and Grubhub arrive through Otter into one kitchen tablet, so the line works a single queue while the diner side stays separated by brand. Otter is the aggregator that pulls the delivery apps into one feed; more on that on the integrations page.
What happens to a brand if I close it?
It should close cleanly: the storefront comes down, the brand stops accepting orders, and the customer and order history for that brand stay exportable. Treat that as a buying requirement rather than an afterthought, because brand churn is normal in this format and a brand you cannot cleanly retire is a brand you will keep paying for.

Price your brands before you launch the next one.

The free AI report reads your current ordering setup and returns what the apps are keeping, in about 60 seconds. No card, no call.

Most operators are live in under 2 weeks. See how it works.