Who can be the Netflix of ghost kitchens?

20.05, Monday 24 Jan 2022

I am sad that TikTok has parted ways with their global head of marketing for going rogue (news in the NY Post) because he was going rogue in delightful ways.

SPECIFICALLY: TikTok Kitchens.

December 2021: TikTok is opening 300 restaurants to deliver some of its most viral food trends like feta pasta and corn ribs across the US (Business Insider).

The idea in a nutshell is that:

  • New food trends get spotted on TikTok
  • They are farmed out to ghost kitchens, running inside existing restaurant kitchens. Established startup Virtual Dining Concepts takes care of training and roll-out
  • Consumer orders and delivery are run by Grubhub, under the TikTok Kitchen brand.

It was a neat concept! You can imagine the satisfaction loop getting even tighter: see some weird food in a tiktok, tap a button in-app and it arrives at your door 30 minutes later.

But possibly not happening now (the launch date was intended to be March this year).

What neat is that this is a new distribution pipe for meals, and I hope someone else comes along to make use of it.

Restaurants are like cable TV, a patchwork of customer relationships, physical infrastructure, and inconsistent local availability. It’s fine but firms are unfocused and limited in scale.

TikTok Kitchen could have been like Netflix! Instant global footprint, and everything apart from (a) audience and (b) content fully commoditised.

Remember when Netflix got into original content with House of Cards? It was a $100m investment for a new business pillar, but Netflix was able to use audience data to guarantee a win:

It already knew that a healthy share had streamed the work of Mr. Fincher, the director of “The Social Network,” from beginning to end. And films featuring Mr. Spacey had always done well, as had the British version of “House of Cards.”

And then a nationwide marketing campaign performs exponentially better than many smaller local campaigns.

So imagine TikTok (or whoever achieves this) doing the same: data-driven meals given simultaneous global rollout, with vast development budgets, and economies of scale for both marketing and also purchasing (you could centrally pre-purchase the world’s feta production to lock up the baked feta pasta supply for the next 12 months).

Ghost kitchens on food delivery apps shouldn’t be the lame knock-off version of established brands. This mechanism should be used to do something entirely new.

Netflix discovered a content/audience/subscription flywheel and now we have a TV renaissance, with Disney and Amazon and Apple joining the arms-race to acquire eyeballs.

SIMILARLY I’m into the idea of inhuman amounts of money going into food, flywheels of tastebud acquisition churning through binge-boxset meals, flagship high-production-values meals, weird niche audience meals (in the global village every niche is nation-sized), and all the rest, competing cinematic universes of cuisines being explored and transmitted via the ghost kitchen machine.

(Independent neighbourhood restaurants will surely be fine - there’s always a market for authenticity - and I won’t terribly miss mid-tier casual dining chains.)


There are some clear gaps in the online food marketing value chain, and primarily that comes down to this fact: you can’t lick a screen and taste anything except glass. With film and TV, you’re selling visual goods in a visual medium. No such luck here. Which means there’s always going to be friction between awareness and purchase.

Fixing this problem needs serious funding, and that’s why I’m into the idea of a ton of money going into the ecosystem.

Here’s the Norimaki Synthesizer by Homei Miyashita, a researcher at Meiji University in Tokyo:

You lick this gadget, and the rod-shaped device is able to simulate any flavour represented by the five universally accepted basic taste sensations: sweet, salty, sour, bitter and umami.

The gadget uses five gels made of dissolved electrolytes that, when electrically charged, provide controlled amounts of each of the five basic tastes to deliver a combination of tastes to the user’s tongue.

The research team liken the process to optical displays that produce many colours from lights of three basic colours (red, yellow and blue).

Each of the gels are made by dissolving five different electrolytes - sodium chloride, glycine, magnesium chloride, citric acid and glutamic sodium - in a small amount of water in separate solutions to create highly concentrated blends.

And so: By adjusting the sliders, Miyashita and his research team could change and transition between tastes, including going from a sweet taste like “gummy candy” to the salty and sour taste of sushi.

Online ads with flavour are the missing link. Measurable, targetable, optimisable advertising for breakfast, lunch, dinner, and snacks.

Look, it’s basic.

But it’s a start, right?

I bet the first computer displays didn’t look like much either. A decade or two of progress and investment will sort it out.

(I was complaining the lack of innovation in screens a while back. Lickable pixels would make up for that.)

It’s 2028. You’re catching up on TikTok and your favourite micro-influencer does their twist on this week’s latest meme meal. The recipe has already propagated across the ghost kitchen network; TikTok spotted the emerging exponential, pre-purchased ingredients, and has already pushed them to the edge so this is tasty, trendy, available right now, and also friendly to the wallet. You cautiously dab the screen with the tip of your tongue – not bad. A larger, wide-tongued doggy slurp, fully from the bottom to the top of your phone. The pixels fizz with flavour. You’re hungry. Buy it now, it’ll arrive at your door on a bike or a drone or robo-courier in 30 minutes or less, this is good, you want your dinner, one-lick purchase, boom.

More posts tagged:

If you enjoyed this post, please consider sharing it by email or on social media. Here’s the link. Thanks, —Matt.