This is the latest in LSA’s Skate To Where the Puck is Going series. Running semi-weekly, it examines the moves and motivations of tech giants as leading indicators for where markets are moving. Check out the entire series here.

Signs are emerging that local food delivery could be Amazon’s next conquest — a scary prospect depending on where you sit. We examined this last week and, in doing so, introduced the concept of cloud kitchens. But what are they exactly? We’re back for a deeper dive on this latest local phenomenon.

In short, cloud kitchens are purpose-built for food delivery. They’re bare-bones and optimally placed for a delivery network. They also don’t require the stuff of typical restaurants like wait staff, dining room, fancy sign and premium real estate in high-traffic areas (insert “location, location, location” motto).

Altogether, this brings three main benefits. The first is cost. When you eat in a restaurant, you pay for all of the above overhead. Cloud Kitchens trim that down, meaning savings to pass along to consumers (enables price competition) or margin to pass around to the kitchen, delivery and on-demand network.

The second benefit is logistics. Now that you no longer require high-traffic placement, you can cluster cloud kitchens in optimal (and cheaper) locations for a delivery network. Think: warehouse districts instead of commercial districts. In logistics parlance, they essentially become fulfillment centers for food.

The third benefit is capacity. Any given restaurant’s kitchen is built to handle the capacity of its counterpart dining room when full. That has yield optimization challenges in terms of an idle kitchen when the restaurant isn’t full, yet it turns down business when the dining room hits capacity.

Cloud kitchens can conversely be built to scale up and down to variable demand (order) levels. And the more restaurants they work with, the greater they can build capacity into their operational planning. That can engender economies of scale, furthering cost benefits and unit economics mentioned above.

The capacity point also ties cloud kitchens back to their moniker. Think of it like cloud computing, where restaurants or on-demand networks outsource food preparation/fulfillment just like startups outsource variable compute needs to the Azures and AWS of the world. We’re calling it “kitchen-as-a-service.”

And just like all those tech startups don’t have to invest in on-site servers and other fixed costs, the above stakeholders in the food world can eschew traditional kitchens for the cash-flow friendly variable pricing of kitchen-as-a-service. There’s a “software is eating the world” joke in there somewhere.

But cloud kitchens aren’t a silver bullet. The whole point is that they’re built for delivery. So they don’t translate to sit-down restaurants which will continue to exist as a cultural staple. Cloud kitchens could also have marketing disadvantages, given the branding and trust typically carried by real restaurants.

But a purpose-built fulfillment system for prepared food delivery still makes sense given the on-demand era rise in ordered food as a share of consumption. Specifically, $85 billion is currently spent on food delivery, according to UBS, projected to grow to $365 billion over the next 10 years (see image above).

“Delivery is the fastest growing market in restaurants,” said Chef Eric Greenspan in an episode of the web series HNGRY (embedded below). “What started out as 10% of your sales is now 30% of your sales, and it will be 50% to 60% of a quick-serve restaurant’s sales within the next three to five years.”

Back to Amazon, all of the above fits the profile for the logistics powerhouse.  Growing market… fulfillment centers… cloud computing: All these factors scream Amazon. As explored in Part I, it’s already halfway there with delivery trucks, real estate (Whole Foods) and its $575M Deliveroo investment.

We’ll continue watching Amazon’s moves closely — along with the overall phenomenon of cloud kitchens and “kitchen-as-a-service” — for their strategic implications to local commerce and SMB operations.